autoconf-1.7/ 40777 21570 1 0 5463274760 11371 5ustar djmdaemonautoconf-1.7/README100666 21570 13 3065 5423612413 12073 0ustar djmuserAutoconf -- generate automatic software configuration scripts from templates Autoconf is an extensible package of m4 macros that produce shell scripts to automatically configure software source code packages. These scripts can adapt the packages to many kinds of UNIX-like systems without manual user intervention. Autoconf creates a configuration script for a package from a template file that lists the operating system features that the package can use, in the form of m4 macro calls. Autoconf requires GNU m4. The configuration scripts produced by Autoconf are independent of Autoconf when they are run, so their users do not need to have Autoconf (or GNU m4). The file INSTALL can be distributed with packages that use Autoconf-generated configure scripts and Makefiles that conform to the GNU coding standards. The package's README can just give an overview of the package, where to report bugs, and a pointer to INSTALL for instructions on compilation and installation. This removes the need to maintain many similar sets of installation instructions. The file `acconfig.h' contains short descriptions of the C preprocessor variables that Autoconf can define, suitable for copying into other packages' configuration headers or Makefile.in files. You can use the program `autoheader' to automatically creates a configuration header from a `configure.in', based on the information in `acconfig.h'. Mail suggestions and bug reports for Autoconf to bug-gnu-utils@prep.ai.mit.edu. Please include the Autoconf version number, which you can get by running "autoconf --version". autoconf-1.7/Makefile.in100666 21570 13 10533 5457576645 13325 0ustar djmuser# Makefile for Autoconf. # Copyright (C) 1992, 1993 Free Software Foundation, Inc. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. #### Start of system configuration section. #### srcdir = @srcdir@ VPATH = @srcdir@ INSTALL = @INSTALL@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_DATA = @INSTALL_DATA@ MAKEINFO = makeinfo TEXI2DVI = texi2dvi M4 = @M4@ prefix = /usr/local exec_prefix = $(prefix) # Directory in which to install scripts. bindir = $(exec_prefix)/bin # Directory in which to install library files. datadir = $(prefix)/lib # Directory in which to install documentation info files. infodir = $(prefix)/info #### End of system configuration section. #### SHELL = /bin/sh DISTFILES = README Makefile.in INSTALL NEWS COPYING ChangeLog \ autoconf.texi autoconf.info* standards.texi make-stds.texi standards.info* \ texinfo.tex acconfig.h autoconf.sh acgeneral.m4 acspecific.m4 \ configure configure.in autoheader.sh mkinstalldirs editsh = sed -e 's,@''datadir''@,$(datadir),g' -e 's,@''M4''@,$(M4),g' all: autoconf autoheader autoconf.info standards.info autoconf: autoconf.sh rm -f $@ $(editsh) $(srcdir)/autoconf.sh > $@ && chmod +x $@ autoheader: autoheader.sh rm -f $@ $(editsh) $(srcdir)/autoheader.sh > $@ && chmod +x $@ info: autoconf.info standards.info autoconf.info: autoconf.texi $(MAKEINFO) -I$(srcdir) $(srcdir)/autoconf.texi --no-split --output=$@ standards.info: standards.texi make-stds.texi $(MAKEINFO) -I$(srcdir) $(srcdir)/standards.texi --no-split --output=$@ dvi: autoconf.dvi standards.dvi autoconf.dvi: autoconf.texi $(TEXI2DVI) $(srcdir)/autoconf.texi standards.dvi: standards.texi make-stds.texi $(TEXI2DVI) $(srcdir)/standards.texi check: installdirs: $(SHELL) ${srcdir}/mkinstalldirs $(bindir) $(infodir) $(datadir) # References to install-info have been removed until it's released. install: all acgeneral.m4 acspecific.m4 acconfig.h installdirs $(INSTALL_PROGRAM) autoconf $(bindir)/autoconf $(INSTALL_PROGRAM) autoheader $(bindir)/autoheader $(INSTALL_DATA) $(srcdir)/acgeneral.m4 $(datadir)/acgeneral.m4 $(INSTALL_DATA) $(srcdir)/acspecific.m4 $(datadir)/acspecific.m4 $(INSTALL_DATA) $(srcdir)/acconfig.h $(datadir)/acconfig.h test ! -r $(srcdir)/aclocal.m4 || $(INSTALL_DATA) $(srcdir)/aclocal.m4 $(datadir)/aclocal.m4 -if test -f autoconf.info; then d=.; else d=$(srcdir); fi; \ $(INSTALL_DATA) $$d/autoconf.info $(infodir)/autoconf.info -if test -f standards.info; then d=.; else d=$(srcdir); fi; \ $(INSTALL_DATA) $$d/standards.info $(infodir)/standards.info uninstall: rm -f $(bindir)/autoconf $(bindir)/autoheader cd $(datadir); rm -f acgeneral.m4 acspecific.m4 aclocal.m4 acconfig.h -rmdir $(datadir) cd $(infodir); rm -f autoconf.info standards.info Makefile: Makefile.in config.status $(SHELL) config.status config.status: configure $(SHELL) config.status --recheck configure: configure.in acgeneral.m4 acspecific.m4 cd $(srcdir); \ $(M4) acgeneral.m4 acspecific.m4 configure.in > configure && \ chmod +x configure clean: rm -f autoconf autoheader rm -f *.aux *.cp *.cps *.dvi *.fn *.fns *.ky *.kys *.log rm -f *.pg *.pgs *.toc *.tp *.tps *.vr *.vrs *.ma *.mas mostlyclean: clean distclean: clean rm -f Makefile config.status realclean: distclean rm -f TAGS *.info* TAGS: etags ${srcdir}/*.m4 ${srcdir}/*.sh ${srcdir}/[a-z]*.in ${srcdir}/*.texi dist: Makefile $(DISTFILES) echo autoconf-`sed -e '/define(AC_ACVERSION,/!d' -e 's/[^0-9.]*\([0-9.]*\).*/\1/' -e q acgeneral.m4` > .fname rm -rf `cat .fname` mkdir `cat .fname` for file in $(DISTFILES); do \ ln $$file `cat .fname` \ || { echo copying $$file instead; cp -p $$file `cat .fname`; }; \ done tar --gzip -chf `cat .fname`.tar.gz `cat .fname` rm -rf `cat .fname` .fname # Prevent GNU make v3 from overflowing arg limit on SysV. .NOEXPORT: autoconf-1.7/INSTALL100666 21570 13 12335 5445717646 12306 0ustar djmuser This is a generic INSTALL file for utilities distributions. If this package does not come with, e.g., installable documentation or data files, please ignore the references to them below. The `configure' shell script attempts to guess correct values for various system-dependent variables used during compilation, and creates the Makefile(s) (one in each subdirectory of the source directory). In some packages it creates a C header file containing system-dependent definitions. It also creates a file `config.status' that you can run in the future to recreate the current configuration. To compile this package: 1. Configure the package for your system. Normally, you just `cd' to the directory containing the package's source code and type `./configure'. If you're using `csh' on an old version of System V, you might need to type `sh configure' instead to prevent `csh' from trying to execute `configure' itself. Running `configure' takes a minute or two. While it is running, it prints some messages that tell what it is doing. If you don't want to see the messages, run `configure' with its standard output redirected to `/dev/null'; for example, `./configure >/dev/null'. To compile the package in a different directory from the one containing the source code, you must use a version of `make' that supports the `VPATH' variable, such as GNU `make'. `cd' to the directory where you want the object files and executables to go and run the `configure' script. `configure' automatically checks for the source code in the directory that `configure' is in and in `..'. If for some reason `configure' is not in the source code directory that you are configuring, then it will report that it can't find the source code. In that case, run `configure' with the option `--srcdir=DIR', where DIR is the directory that contains the source code. By default, `make install' will install the package's files in `/usr/local/bin', `/usr/local/man', etc. You can specify an installation prefix other than `/usr/local' by giving `configure' the option `--prefix=PATH'. Alternately, you can do so by consistently giving a value for the `prefix' variable when you run `make', e.g., make prefix=/usr/gnu make prefix=/usr/gnu install You can specify separate installation prefixes for architecture-specific files and architecture-independent files. If you give `configure' the option `--exec-prefix=PATH' or set the `make' variable `exec_prefix' to PATH, the package will use PATH as the prefix for installing programs and libraries. Data files and documentation will still use the regular prefix. Normally, all files are installed using the same prefix. Some packages pay attention to `--with-PACKAGE' options to `configure', where PACKAGE is something like `gnu-as' or `x' (for the X Window System). The README should mention any `--with-' options that the package recognizes. `configure' ignores any other arguments that you give it. On systems that require unusual options for compilation or linking that the package's `configure' script does not know about, you can give `configure' initial values for variables by setting them in the environment. In Bourne-compatible shells, you can do that on the command line like this: CC='gcc -traditional' LIBS=-lposix ./configure Here are the `make' variables that you might want to override with environment variables when running `configure'. For these variables, any value given in the environment overrides the value that `configure' would choose: - Variable: CC C compiler program. The default is `cc'. - Variable: INSTALL Program to use to install files. The default is `install' if you have it, `cp' otherwise. For these variables, any value given in the environment is added to the value that `configure' chooses: - Variable: DEFS Configuration options, in the form `-Dfoo -Dbar...'. Do not use this variable in packages that create a configuration header file. - Variable: LIBS Libraries to link with, in the form `-lfoo -lbar...'. If you need to do unusual things to compile the package, we encourage you to figure out how `configure' could check whether to do them, and mail diffs or instructions to the address given in the README so we can include them in the next release. 2. Type `make' to compile the package. If you want, you can override the `make' variables CFLAGS and LDFLAGS like this: make CFLAGS=-O2 LDFLAGS=-s 3. If the package comes with self-tests and you want to run them, type `make check'. If you're not sure whether there are any, try it; if `make' responds with something like make: *** No way to make target `check'. Stop. then the package does not come with self-tests. 4. Type `make install' to install programs, data files, and documentation. 5. You can remove the program binaries and object files from the source directory by typing `make clean'. To also remove the Makefile(s), the header file containing system-dependent definitions (if the package uses one), and `config.status' (all the files that `configure' created), type `make distclean'. The file `configure.in' is used to create `configure' by a program called `autoconf'. You only need it if you want to regenerate `configure' using a newer version of `autoconf'. autoconf-1.7/NEWS100666 21570 1 6510 5457607703 12167 0ustar djmdaemonMajor changes in release 1.7: * New macro AC_OBSOLETE. * Bugs in Makefile.in fixed. * AC_LONG_FILE_NAMES improved. Major changes in release 1.6: * New macro AC_LONG_64_BITS. * Multiple .h files can be created. * AC_FIND_X looks for X files directly if it doesn't find xmkmf. * AC_ALLOCA defines C_ALLOCA if using alloca.c. * --with-NAME can take a value, e.g., --with-targets=sun4,hp300bsd. * Unused --no-create option to configure removed. * autoheader doesn't change the timestamp of its output file if the file didn't change. * All macros that look for libraries now use AC_HAVE_LIBRARY. * config.status checks three optional environment variables to modify its behavior. * The usual bug fixes. Major changes in release 1.5: * New macros AC_FIND_X, AC_OFF_T, AC_STAT_MACROS_BROKEN, AC_REVISION. * autoconf and autoheader scripts have GNU standards conforming --version and --help options (they print their message and exit). * Many bug fixes. Major changes in release 1.4: * New macros AC_HAVE_POUNDBANG, AC_TIME_WITH_SYS_TIME, AC_LONG_DOUBLE, AC_GETGROUPS_T, AC_DEFINE_UNQUOTED. * autoconf and autoheader use the M4 environment variable to determine the path of the m4 program to use. * The --macrodir option to autoconf and autoheader specifies the directory in which acspecific.m4, acgeneral.m4, etc. reside if not the default. * autoconf and autoheader can take `-' as their file names, which means to read stdin as input. * Resulting configure scripts can take a --verbose option which causes them to print the results of their tests. * AC_DEFINE quotes its second argument in such a way that spaces, magic shell characters, etc. will be preserved during various stages of expansion done by the shell. If you don't want this, use AC_DEFINE_UNQUOTED instead. * Much textual processing done with external calls to tr and sed have been internalized with builtin m4 `patsubst' and `translit' calls. * AC_OUTPUT doesn't hardwire the filenames it outputs. Instead, you can set the shell variables `gen_files' and `gen_config' to the list of filenames to output. * AC_DECLARE_YYTEXT does an AC_SUBST of `LEX_OUTPUT_ROOT', which may be "lex.yy" or "lexyy", depending on the system. * AC_PROGRAMS_CHECK takes an optional third arg. If given, it is used as the default value. * If AC_ALLOCA chooses alloca.c, it also defines STACK_DIRECTION. * AC_CONST works much more reliably on more systems. * Many bug fixes. Major changes in release 1.3: configure no longer requires awk for packages that use a config.h. Support handling --with-PACKAGE options. New `autoheader' script to create `config.h.in' from `configure.in'. Ignore troublesome -lucb and -lPW when searching for alloca. Rename --exec_prefix to --exec-prefix for GNU standards conformance. Improve detection of STDC library. Add AC_HAVE_LIBRARY to check for non-default libraries. Function checking should work with future GNU libc releases. Major changes in release 1.2: The --srcdir option is now usually unnecessary. Add a file containing sample comments describing CPP macros. A comment in config.status tells which host it was configured on. Substituted variable values can now contain commas. Fix bugs in various feature checks. Major changes in release 1.1: Added AC_STRCOLL macro. Made AC_GETLOADAVG check for more things. AC_OUTPUT argument is now optional. Various bug fixes. autoconf-1.7/COPYING100666 21570 13 43076 5022247405 12274 0ustar djmuser GNU GENERAL PUBLIC LICENSE Version 2, June 1991 Copyright (C) 1989, 1991 Free Software Foundation, Inc. 675 Mass Ave, Cambridge, MA 02139, USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. Preamble The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation's software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Library General Public License instead.) You can apply it to your programs, too. When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things. To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it. For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights. We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software. Also, for each author's protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors' reputations. Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all. The precise terms and conditions for copying, distribution and modification follow. GNU GENERAL PUBLIC LICENSE TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 0. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term "modification".) Each licensee is addressed as "you". Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does. 1. You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program. You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee. 2. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions: a) You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change. b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.) These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program. In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. 3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following: a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable. If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code. 4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance. 5. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it. 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License. 7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program. If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances. It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice. This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License. 8. If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License. 9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. 10. If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally. NO WARRANTY 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. END OF TERMS AND CONDITIONS Appendix: How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms. To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found. Copyright (C) 19yy This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. Also add information on how to contact you by electronic and paper mail. If the program is interactive, make it output a short notice like this when it starts in an interactive mode: Gnomovision version 69, Copyright (C) 19yy name of author Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than `show w' and `show c'; they could even be mouse-clicks or menu items--whatever suits your program. You should also get your employer (if you work as a programmer) or your school, if any, to sign a "copyright disclaimer" for the program, if necessary. Here is a sample; alter the names: Yoyodyne, Inc., hereby disclaims all copyright interest in the program `Gnomovision' (which makes passes at compilers) written by James Hacker. , 1 April 1989 Ty Coon, President of Vice This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License. autoconf-1.7/ChangeLog100666 21570 1 166257 5463274574 13324 0ustar djmdaemonTue Oct 26 15:24:33 1993 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * Version 1.7. Tue Oct 19 23:49:50 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_TEST_PROGRAM): Don't remove conftest* before running $2 or $3 or $4; just once at the end. Mon Oct 18 01:38:00 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_PREPARE): Echo a newline into confdefs.h so it is never empty. Fri Oct 15 18:49:20 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_CONST): Added test of trivial use for broken Ultrix-32 V3.1 Rev 9 vcc. Fri Oct 15 15:44:39 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_OBSOLETE): New macro. * acspecific.m4 (AC_UNISTD_H, AC_USG, AC_MEMORY_H): Call it. * acspecific.m4 (AC_LONG_FILE_NAMES): Try to create files in ${prefix}/lib and ${exec_prefix}/lib instead of ${prefix} and ${exec_prefix}; they are more likely to be writable. * Makefile.in (clean): Remove *.ma and *.mas, the macro index files. Tue Oct 12 16:02:52 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_RETSIGTYPE): AC_PROVIDE self. Mon Oct 11 19:09:20 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * Makefile.in (editsh): Obfuscate @M4@ and @datadir@ references so configure doesn't edit them. Sun Oct 10 14:01:35 1993 Jim Meyering (meyering@comco.com) * autoconf.sh (--help): Exit successfully. Sat Oct 9 08:29:15 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * Version 1.6. * acconfig.h (inline): New entry. * acspecific.m4 (AC_DIR_HEADER_CHECK): Don't call opendir, in case the needed libraries (e.g., -ldir on Xenix) aren't in LIBS yet. From Jim Meyering (meyering@comco.com). * acspecific.m4 (AC_PROG_LEX): Fix typo. * acgeneral.m4 (AC_HEADER_EGREP, AC_PROGRAM_EGREP, AC_COMPILE_CHECK, AC_TEST_PROGRAM, AC_TEST_CPP): Remove any temporary files before doing the actions, in case they're nested tests. From gray@antaire.com (Gray Watson). * configure.in: Check for GNU m4 under several names. * Makefile.in: Use that value. From Franc,ois Pinard. * acspecific.m4 (AC_STRUCT_TM): Use a member of struct tm, to make sure the compiler complains if it's not defined. From Bruno Haible (haible@ma2s2.mathematik.uni-karlsruhe.de). * acspecific.m4 (AC_FIND_X_XMKMF): If libX11.a isn't in USRLIBDIR, check in LIBDIR. Filter out any make verbose messages. Tue Oct 05 19:21:29 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_LONG_DOUBLE): Announce that this feature is being checked even if the test is simply whether $CC is gcc. Tue Oct 5 14:23:28 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh: Produce HAVE_LIBfoo for AC_HAVE_LIBRARY. Sun Oct 3 15:41:36 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Write assignment for `extrasub'; in sed cmds, write "$extrasub" so configure.in can set it to do sed frobs. Take second arg and write it to config.status before `exit 0'. * acspecific.m4 (AC_CONST): Say `checking for lack of working const'. That is precisely accurate. Wed Sep 22 15:47:50 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4: If not using GNU m4, abort. * acgeneral.m4 (AC_PREPARE): Lose if we're not in the srcdir, not if we're in it. But disable the check for now. Mon Sep 20 15:32:30 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_PREPARE): Check for $srcdir being configured, diagnose and lose. * acgeneral.m4 (AC_QUOTE_SED): Quote @ and %. * acgeneral.m4 (AC_OUTPUT): Say "$file is unchanged" when it is. Sat Sep 18 14:32:04 1993 Ian Lance Taylor (ian@airs.com) * acgeneral.m4: Substitute for CONFIG_FILES and CONFIG_HEADERS before using them, in case they have multiple values. Fri Sep 17 14:40:20 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_WAIT3): wait3 works if ru_stime is nonzero, too. Thu Sep 16 15:39:53 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_FIND_X_XMKMF): Code moved from AC_FIND_X. (AC_FIND_X_DIRECT): New function, derived from code by Karl Berry and Rob Savoye. (AC_FIND_X): Call them. Wed Sep 15 19:06:46 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_PREPARE): Remove confdefs* on exit with trap 0. (AC_OUTPUT): Don't bother removing it. * acgeneral.m4: Remove --no-create option; not useful. Mon Sep 13 21:54:46 1993 Paul Eggert (eggert@twinsun.com) * autoheader.sh: Rename the temporary output to the real output if their contents differ, not if their contents are identical. This fixes bug introduced in Aug 30 change. Mon Sep 13 16:50:30 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Run config.status with CONFIG_SHELL if defined. Same for configure run from config.status. Rename gen_files to CONFIG_FILES and gen_config to CONFIG_HEADERS. * acgeneral.m4 (AC_PREPARE): Remove confdefs* in trap. Fri Sep 10 00:29:20 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_LONG_FILE_NAMES): Test /var/tmp as well. In loop, skip past nonexistent dirs. * acspecific.m4 (AC_CONST): Say "working", not "broken". We are checking for a working const as opposed to a broken or absent const, not for a broken const as opposed to a working one. Thu Sep 9 09:25:49 1993 Jim Meyering (meyering@comco.com) * acspecific.m4, acconfig.h (AC_LONG_64_BITS): New macro. Wed Sep 1 18:54:12 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_PROGRAM_CHECK): Use && instead of test -a. Tue Aug 31 19:21:35 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT_HEADER): Support generating multiple .h files. From gray@antaire.com (Gray Watson). * acspecific.m4 (AC_ALLOCA): If using alloca.o, define C_ALLOCA. * acgeneral.m4 (compile, AC_HEADER_EGREP, AC_PROGRAM_EGREP, AC_COMPILE_CHECK, AC_TEST_PROGRAM, AC_TEST_CPP): Remove $DEFS from cc and cpp command lines; include "confdefs.h" in test files. (AC_DEFINE): Append a #define to confdefs.h. Reduce duplicated code by introducing a temp variable, AC_VAL. Mon Aug 30 17:36:54 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh: Don't write output if it is the same as output file. Wed Aug 25 14:14:33 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_VFORK): Check for SunOS 5.2 bug with ignoring signal in parent before vfork. From eggert. Fri Aug 20 10:14:42 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_PARSEARGS): Support giving values to --with options. Go back to using sed for invalid test, but without using '*' in the regex. Thu Aug 19 14:53:29 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_LONG_FILE_NAMES): eval the args. * acgeneral.m4 (AC_PARSEARGS): Use case instead of sed and test to detect invalid package names. Remove =value from --with options until we support it. Wed Aug 11 18:52:41 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_FIND_X): Don't set x_includes if it's /usr/include or x_libraries if it's /lib or /usr/lib. Wed Aug 11 13:00:18 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_LONG_FILE_NAMES): If we cannot write $dir, echo a warning msg and continue the loop to skip that directory. * acgeneral.m4 (AC_REVISION): Also eat double quotes. Thu Aug 5 14:55:59 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acconfig.h: Add TIME_WITH_SYS_TIME. Mon Aug 2 14:55:16 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_DECLARE_YYTEXT): \-escape "s in rhs of AC_DEFINE_UNQUOTED. Remove gratuitous second arg to AC_SUBST. Sun Aug 1 19:13:08 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Define HAVE_GETLOADAVG if we find one and don't use our own getloadavg.c. * acconfig.h: Add HAVE_GETLOADAVG. Sat Jul 31 17:28:48 1993 Karl Berry (karl@cs.umb.edu) * acspecific.m4 (AC_PROG_INSTALL): Report results under -v. Fri Jul 30 18:08:30 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh (syms, headers, funcs, libs): Run values through sort|uniq to remove duplicates. Wed Jul 28 00:02:34 1993 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * Makefile.in (config.status): Run config.status --recheck, not configure. (install): Remove refs to install-info until it's released, because people are getting confused. * acgeneral.m4 (AC_OUTPUT): For config.status --recheck, echo the configure command line that we run. * acspecific.m4 (AC_PROG_FLEX): Use AC_HAVE_LIBRARY. Mon Jul 26 19:11:01 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Check that both -lutil and -lkvm exist before choosing them in hopes they will define getloadavg. * autoheader.sh (frob): Put $2 and $3 in the expansion of AC_HAVE_LIBRARY, so AC_DEFINE there is noticed. Mon Jul 26 14:21:33 1993 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (INT_16_BITS): Check the obvious way, so it doesn't pick up machines with 64 bit longs. Mon Jul 26 14:01:38 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Check for -lelf with AC_HAVE_LIBRARY instead of checking for with AC_HEADER_CHECK. Mon Jul 26 13:58:39 1993 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_SCO_INTL, AC_IRIX_SUN, AC_DYNIX_SEQ): Use AC_HAVE_LIBRARY. Mon Jul 26 13:55:17 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh (eval frob): Restore hairy sed use; we need it to handle multi-line macro invocations. Mon Jul 26 00:50:43 1993 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_FIND_X): Quote the Imakefile. Sun Jul 25 08:17:11 1993 Jim Meyering (meyering@comco.com) * acconfig.h (CRAY_STACKSEG_END): New #undef. Thu Jul 22 20:26:12 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * Version 1.5. * acspecific.m4 (AC_FIND_X): Let make substitute any variables in INCROOT and USRLIBDIR, instead of using sed. From wojo@veritas.com (Jack Woychowski). * acgeneral.m4 (AC_DEFINE): When printing value verbosely, use double quotes and AC_DEFINE_QUOTE, like we do when assigning the value, so shell variables get expanded the same way. * acgeneral.m4 (AC_REVISION): New macro. From wollman@uvm-gen.EMBA.UVM.EDU (Garrett Wollman). * acgeneral.m4 (AC_DEFINE): Add newline before open brace. Thu Jul 22 17:07:15 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_STAT_MACROS_BROKEN): New macro. * acconfig.h (STAT_MACROS_BROKEN): New #undef. Wed Jul 21 15:44:32 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_DECLARE_YYTEXT): Use AC_DEFINE_UNQUOTED so shell var is replaced in rhs. Wed Jul 21 13:31:38 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acconfig.h (size_t, mode_t, off_t): Added. * acspecific.m4 (AC_OFF_T): New macro. Tue Jul 20 15:39:44 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * autoheader.sh: Put header-file.in in comment at top. * acconfig.h (NDIR): Added. Mon Jul 19 22:10:49 1993 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * Makefile.in (info, dvi): New targets. Sun Jul 18 22:36:33 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh (frob): Use `#' as the first line of each definition. (eval frob): Totally simplify sed use to just handle "^@@@.*@@@$". Wed Jul 14 22:44:25 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acconfig.h: Restore blank lines between paragraphs. * autoheader.sh (libs): New variable and frob to set it from AC_HAVE_LIBRARY uses. Produce #undef HAVE_* for each $libs. Tue Jul 13 19:03:46 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acconfig.h: Sort the entries, like the comment says. * acspecific.m4 (AC_GETLOADAVG): Only check for the AIX library once, looking in both local and system dirs. Consolidate SVR4 and Solaris cases. Mon Jul 12 20:33:36 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): If we find sys/dg_sys_info.h, do AC_HAVE_LIBRARY on -ldgc. Sun Jul 11 00:43:51 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): BSD library is -lutil, not -lutils, and requires -lkvm too. Check for local AIX library using AC_HAVE_LIBRARY, not AC_COMPILE_CHECK. Un-nest some conditionals. Stop checking once we've found a way to get getloadavg. Thu Jul 8 20:21:28 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * Makefile.in: Remove rules for making *.conf; make Autoconf's configure script semi-normally. Wed Jul 7 14:37:35 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * autoheader.sh (--help): Print help message to stdout and exit 0. (--version): Exit after printing version number. * autoconf.sh (--version): Exit after printing version number. * acspecific.m4 (AC_LONG_DOUBLE): Make sure that long double isn't smaller than double, as in Ultrix 4.[23] cc. * acgeneral.m4 (AC_REPLACE_FUNCS): Include ctype.h in the test program to get stubs. * acspecific.m4 (AC_FIND_X): New macro. Tue Jul 6 19:15:17 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Try ls -L first, in case /dev/kmem is a symlink (as on Solaris). Wed Jun 30 22:08:22 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_MINUS_C_MINUS_O): Remove spurious `then'. Fri Jun 25 23:16:42 1993 Paul Eggert (eggert@twinsun.com) * acspecific.m4 (AC_CONST): Replace `p = ' with `ccp = '; the former wasn't ANSI C, and was causing working compilers to be rejected. Fri Jun 25 13:26:34 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_LONG_FILE_NAMES): Redirect rm's stderr to /dev/null. Thu Jun 24 15:58:04 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * autoconf.sh, autoheader.sh, acgeneral.m4 (AC_PREPARE): Undo change of Jun 16 1993. Only set `LANG' and `LC_ALL' to "C" if already set. Sat Jun 19 00:01:51 1993 Jim Meyering (meyering@comco.com) * acgeneral.m4: Undefine m4's `format' builtin. * acspecific.m4 (AC_HAVE_POUNDBANG): Make conftest executable, but not necessarily writable by group or other. Thu Jun 17 21:10:33 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_CPP): Put double quotes around ${CC-cc}, not single quotes. If --verbose option given, say what CPP is being set to. Wed Jun 16 17:50:00 1993 Jim Blandy (jimb@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_CPP): Make sure that `cc -E` doesn't run the program through the C compiler too. Bob Olson says it does on the NeXT. Wed Jun 16 16:17:05 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * autoconf.sh, autoheader.sh, acgeneral.m4 (AC_PREPARE): Always set `LANG' and `LC_ALL' environment variables to `C'. Fri Jun 11 14:29:31 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_MINUS_C_MINUS_O): Test that cc works at all, and only test it for -c -o if it does. Tue Jun 8 01:47:22 1993 Paul Eggert (eggert@twinsun.com) * acgeneral.m4 (AC_OUTPUT): The line DEFS="`echo \"$DEFS\" | sed 's%[&\\\]%\\\&%g'`" doesn't work in some shells, which don't allow nesting \"\" inside `` inside "", and which don't unescape \\\& in the expected (?) way. Also, some versions of echo interpret backslashes inside $DEFS. Put $DEFS into a temporary file to avoid these portability minefields. Mon Jun 7 20:11:50 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): In setting KMEM_GROUP, use new sed magic from friedman which should win with both meanings of ls -lg. Mon Jun 7 06:48:49 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * Makefile.in (dist): Change gzipped tar file extension to `.gz'. Use explicit --gzip option to tar to make sure tar uses the right compression program (or else exits from failure to understand the option). * acgeneral.m4 (AC_OUTPUT): Don't split sed expr for exec_prefix across two lines, since not all versions of sed understand that. * acspecific.m4 (AC_HAVE_POUNDBANG): Complete rewrite which doesn't depend on csh. Tue Jun 1 03:06:28 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * Version 1.4.1 (not announced to the general public, but a snapshot was put on the June '93 GNU CDROM). * Makefile.in (dist): If ln fails (e.g. because of cross-device links), mention on stdout that file is being copied. * acgeneral.m4 (AC_PREPARE): Use `[$]*' in assignment to configure_args to get shell positional args, rather than m4 args to AC_PREPARE. (AC_OUTPUT): Use `configure_args' in config.status when invoked with --recheck, rather than $*. Mon May 31 13:12:56 1993 Paul Eggert (eggert@twinsun.com) * acspecific.m4 (AC_LONG_FILE_NAMES): rm $dir/conftest*, not conftest*. Mon May 31 04:18:18 1993 Roland McGrath (friedman@nutrimat.gnu.ai.mit.edu) * acgeneral.m4 (AC_HAVE_LIBRARY): Quote libname in define. Sun May 30 19:52:24 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_SETVBUF_REVERSED): Pass (char *) main to setvbuf instead of zero. Thu May 27 20:30:53 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_PREPARE): Save $* in shell var `configure_args'. (AC_OUTPUT): Use $configure_args in place of $*. Wed May 26 16:19:51 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * autoconf.texi (AC_PROG_INSTALL): Doc fix. (Automatic Remaking): Put code fragment in @example ... @end example. Mon May 24 15:46:47 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh (frob): Redefine AC_CONFIG_HEADER to set shell variable `config_h'. (config_h): New variable, initialize to "config.h" before frobbing. (final output): Write ${config_h}.in. Sat May 22 17:45:19 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * Version 1.4 released. Thu May 20 20:25:45 1993 Jim Blandy (jimb@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_IDENTITY): New function. (AC_DEFINE_UNQUOTED): Use it to fix this; due to a misunderstanding of m4, this was using its first argument as the definition. Thu May 20 09:21:55 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_ALLOCA) [find_stack_direction]: Return the value from the recursive call. If it worked before, it was by luck. From Bruno Haible . Tue May 18 23:40:21 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_STDC_HEADERS): Require AC_PROG_CPP. Mon May 17 18:01:09 1993 Karl Berry (karl@hal.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Use variables gen_files and gen_config in the loop that generates the output (Make)files, instead of hardwiring the filenames. Sat May 15 17:23:19 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * autoconf.sh: Accept `-' to mean read stdin as input. * autoheader.sh: Likewise. Fri May 14 12:41:02 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * autoheader.sh, acspecific.m4 (AC_PREPARE): If `LANG' environment variable is set, reset its value to `C'. This is so `tr [...]' works more portably. Thu May 13 22:56:20 1993 Paul Eggert (eggert@twinsun.com) * acspecific.m4 (VOID_CLOSEDIR): Test closedir instead of assuming that it works. E.g. dynix closedir yields garbage, but has no prototype. Presumably Xenix closedir had the same problem, so stop special-casing it. Wed May 12 20:25:36 1993 Jim Meyering (meyering@comco.com) * acconfig.h: Add HAVE_LONG_DOUBLE. Wed May 12 15:07:36 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_DEFINE_UNQUOTED): New macro. * acgeneral.m4 (AC_FUNC_CHECK): Include ctype.h instead of stdio.h. We want it only to define __stub_* in glibc. Using stdio.h lost when it contained a conflicting prototype for $1; ctype.h has fewer prototypes. * acconfig.h: Add GETGROUPS_T. * acspecific.m4 (AC_PROG_RANLIB): Use : instead of @: for no-op. Some braindead make does bizarre magical things with @ in variables. Mon May 10 14:24:27 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_HAVE_POUNDBANG): New feature. * acgeneral.m4 (AC_OUTPUT): Add more backslashes to character class in DEFS filter (sigh). Sun May 9 14:04:31 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_DEFINE_QUOTE): No AC_QUOTE_SED (was innermost). (AC_HEADER_EGREP, AC_PROGRAM_EGREP, AC_TEST_CPP): Put a \ before $DEFS in string to be evalled. (AC_OUTPUT): Run DEFS through a sed filter that quotes things in it from sed (woo woo!) before writing it into config.status. * acspecific.m4 (AC_ALLOCA): Use AC_PROGRAM_EGREP to test for [CRAY && !CRAY2], instead of AC_TEST_PROGRAM. No need to run a program for this. * acgeneral.m4 (AC_PROGRAM_CHECK): Extract the first word of $2 when looking for it in PATH, so it can be a program name with args. Omit default assignment if $4 is empty. Only write verbose msg if $1 was set nonempty. * acspecific.m4 (AC_PROG_YACC): Pass 'bison -y' (quoted like that) in list to AC_PROGRAMS_CHECK. Don't test for bison later to add -y flag. Sat May 8 00:23:58 1993 Jim Meyering (meyering@comco.com) * acgeneral.m4 (AC_REPLACE_FUNCS): Add a trailing newline in code for AC_COMPILE_CHECK. Otherwise it got spurious failures. * acspecific.m4 (TIME_WITH_SYS_TIME): New macro. * Makefile.in (dist): Depend on Makefile. Use gzip instead of compress. Link files individually instead of en masse; if a link fails, use `cp -f' on the losing file. * acspecific.m4 (AC_ALLOCA): Define CRAY_STACKSEG_END (the name of a function used in alloca.c) for CRAY-1, CRAY X-MP, and CRAY Y-MP. Fri May 7 15:56:26 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Check for mach/mach.h, but don't disable nlist checks if found. Fri May 7 04:59:25 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_INSTALL): Don't look for `install' in /usr/ucb. Thu May 6 20:41:35 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_FUNC_CHECK): The test program should choke on #ifdef __stub___$1 as well. (AC_REPLACE_FUNCS): Make the test program choke on stubs. Wed May 5 20:43:13 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoconf.sh ($infile existence check): Fixed test for nonemptiness of $print_version to not always be true. Wed May 5 17:22:42 1993 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_PREFIX, AC_PROGRAM_CHECK), acspecific.m4 (AC_PROG_INSTALL): If IFS wasn't set initially, give it a normal default value. Happens on LynxOS (x86), says Pete Klammer . Wed May 5 13:22:52 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4: Undefine the `shift' builtin. * acspecific.m4 (AC_PROG_YACC): Use AC_PROGRAMS_CHECK to check for both bison and yacc, instead of two AC_PROGRAM_CHECK uses. * autoheader.sh ($# -eq 0): Set var $tmpout to name of temp file, send stdout there instead of config.h.in. (just before exit): If $# -eq 0, then move $tmpout to config.h.in if $status -eq 0, or remove $tmpout otherwise. * acspecific.m4 (AC_STRCOLL): Rewritten to use a test program that verifies that `strcoll' does vaguely reasonable ordering. Tue May 4 19:59:00 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_LONG_DOUBLE): Don't explicitely echo `checking for long double'. Mon May 3 22:04:35 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_GETGROUPS_T): New macro. Sat May 1 22:37:55 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_LONG_DOUBLE): New macro. Wed Apr 28 15:52:42 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_PROGRAM_CHECK): Write msg under --verbose. Thu Apr 22 18:24:40 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_FUNC_CHECK): Remove spurious `#endif' line at end. * acgeneral.m4 (AC_WITH): Fix reversed args to patsubst. Test $with_FOO, not $FOO. Wed Apr 21 18:14:19 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_QUOTE_TOKEN): New macro. (AC_DEFINE_QUOTE): Use it. Tue Apr 20 18:02:46 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_DECLARE_YYTEXT): Guess name of lex output file and do AC_SUBST of `LEX_OUTPUT_ROOT'. Add `dnl' after calling some autoconf macros. Mon Apr 19 15:46:24 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_MINUS_C_MINUS_O): Do each compile a second time after testing for the existence of the output. Some compilers refuse to overwrite an existing .o file with -o, though they will create one. * acspecific.m4 (AC_DECLARE_YYTEXT): Changed lex input to two lines of "%%", not just one. Sat Apr 17 17:26:12 1993 Jim Meyering (meyering@comco.com) * acgeneral.m4 (AC_COMPILE_CHECK): Don't print `checking for ...' message if first argument is empty. Sat Apr 17 01:18:41 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_PID_T): provide self. (AC_VFORK): Require AC_PID_T. Fri Apr 16 11:57:35 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_PROGRAMS_CHECK): Take optional third arg; if given, use it as the default value. Thu Apr 15 16:43:45 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_REPLACE_FUNCS): Print a message under --verbose. * acgeneral.m4 (AC_HAVE_LIBRARY): Use m4's patsubst and translit instead of running sed and tr at runtime. * acconfig.h: Add STACK_DIRECTION. Wed Apr 14 17:08:47 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_ALLOCA): If we chose alloca.c, run a test program to define STACK_DIRECTION. Mon Apr 5 19:02:52 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_LONG_FILE_NAMES): Put test inside a for loop on several directories: . /tmp $prefix $exec_prefix. Define HAVE_LONG_FILE_NAMES iff long names win in all those directories. Sun Apr 4 18:38:23 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * Makefile.in (%.info): Removed pattern rule. (autoconf.info, standards.info): New rules. * autoconf.sh (version_only): New variable, set nonempty for `autoconf --version' with no input file. (output writing): No output if $version_only is set. Wed Mar 31 17:33:57 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_CONST): Uncomment and fix second AIX test. Wed Mar 31 16:58:12 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_CONST): Rewrite first AIX XL C 1.02.0.0 test. Comment out bogosity in second AIX test. Wed Mar 31 12:45:59 1993 Jim Meyering (meyering@comco.com) * acgeneral.m4 (AC_DEFINE): Put single quotes around definition that is echoed with --verbose. AC_DEFINE(MVDIR, "$(libdir)/mvdir") was generating losing code. Mon Mar 29 15:44:24 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acspecific.m4 (AC_STDC_HEADERS): Add a missing pair of [quotes]. Mon Mar 29 14:54:00 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_DECLARE_YYTEXT): Change sed regexp so it won't match other identifiers beginning with `yytext'. Sat Mar 27 00:11:16 1993 Paul Eggert (eggert@twinsun.com) * acspecific.m4 (AC_CONST): Detect broken AIX XL C 1.2.0.0 compiler. Thu Mar 25 19:54:50 1993 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_CONST): Remove single quotes from the C program; they produce shell syntax errors. * acgeneral.m4 (AC_DEFINE): Add a newline after "}" to prevent commands following on the same line of configure.in from generating shell syntax errors. * acgeneral.m4 (AC_COMPILE_CHECK): Use explicit return types to avoid warnings. (AC_TEST_CPP): Add parens to force redirection order. (AC_OUTPUT): Allow hostname to return bogus exit status. From Jean-loup Gailly . Mon Mar 22 16:53:01 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoconf.sh: Use $M4, not m4 explicitly. (M4): If unset in env, initialize to @m4@. * autoheader.sh: Likewise. * Makefile.in (M4): Define new variable. (autoconf.conf, %.conf): Use it. (editsh): New variable: sed command to replace @datadir@; also replace @M4@ with $(M4). (autoconf, autoheader): Use $(editsh) instead of explicit sed command. Mon Mar 22 13:08:10 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_CONST): IBM's /bin/cc under AIX-3.2 on an rs6000 rejects attempts to modify *any* member of a struct that has a member declared like `const int *ap[2]'. Wed Mar 17 18:08:30 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * autoconf.sh, autoheader.sh (MACRODIR): Variable renamed to AC_MACRODIR. Don't initialize it at runtime if it is already set in the environment. (MACROFILES): Don't set until after options are processed. (print_version): New temp variable. * autoconf.sh, autoheader.sh: Rewrote argument parsing. Added `-m', `--macrodir', `-h', `--help', and `--' options. Updated usage string. * autoconf.texi: Documented --macrodir option and its effects for both scripts. Tue Mar 16 09:10:48 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_CONST): Sun's SC1.0 ANSI compiler (acc) won't increment a `const int *' pointer declared through a typedef. Mon Mar 15 16:08:42 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_PARSEARGS): Grok `--verbose' flag; set verbose=yes. (AC_DEFINE): Only echo "defining $1" if $verbose is set. Sun Mar 14 18:19:21 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_INSTALL): Choose `installbsd' if we find it, in preference to `install'. * acspecific.m4 (AC_CONST): Add a check for `const int *foo' not allowing modification of FOO (not *FOO). Fri Mar 12 15:27:53 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT_HEADER): Remove conftest.sh before creating it. Thu Mar 11 12:57:53 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_DEFINE): Surround defn with { and }. * acgeneral.m4 (AC_OUTPUT_HEADER): Split up $SEDDEFS into smaller chunks, since some shells can't handle large here documents. We write several commands in config.status to create conftest.sed in pieces. Mon Mar 8 14:40:53 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_WITH): Don't echo anything. Use the m4 patsubst fn instead of a run-time sed invocation to massage $1. * acspecific.m4 (AC_DIR_HEADER_CHECK): #include before the header we are testing. * acgeneral.m4 (AC_DEFINE): If $2 is empty, echo "defining $1 to be empty", rather than "defining $1 to be ". * acspecific.m4 (AC_DIR_HEADER_CHECK): New; subr of AC_DIR_HEADER. (AC_DIR_HEADER): Use it to test for each possible header file. Tue Mar 2 01:06:25 1993 Noah Friedman (friedman@nutrimat.gnu.ai.mit.edu) * autoheader.sh: Don't use /p1/,/p2/ construct with sed---it's not portable. Handle broken AIX sed that strips \n from hold space when it shouldn't. From Jun Hamano . Tue Mar 02 00:08:39 1993 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_CONST): Fix typo that caused spurious lossage with /bin/cc from Irix-4. From Karl Berry. Fri Feb 26 17:14:58 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_CONST): Add bizarre case that loses on SCO 3.2v4. Mon Feb 22 13:02:27 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_QUOTE_HERE, AC_QUOTE_SED): Change the quote chars to { and } instead of nothing. Then use {} (empty quotes) to separate the patsubst forms from the following dnl. Otherwise the result of patsubst is pasted together with dnl and the result is seen as a single token. * acspecific.m4 (AC_MINUS_C_MINUS_O): Print msg saying what we are doing before we do it. * acgeneral.m4 (AC_PREFIX): Print out the choice made. (AC_DEFINE): Print out the definition being done. * acgeneral.m4 (AC_DEFINE_QUOTE): Add dnl at end of line. * acspecific.m4 (AC_GETLOADAVG): Do changequote around listing of /dev/kmem and sed frobbing which needs to use [ and ]. Sun Feb 21 13:57:55 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh: Use brackets in tr range args. * acspecific.m4 (AC_SETVBUF_REVERSED): Make the test fail if setvbuf returns nonzero. * acspecific.m4 (AC_GETLOADAVG): If we need to install setgid, figure out what group owns /dev/kmem, and set KMEM_GROUP to that. * acspecific.m4 (AC_MINUS_C_MINUS_O): Test plain `cc' after testing $CC. We want to make sure both compilers grok -c -o. Thu Feb 18 18:05:14 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_QUOTE_{DQUOTE,SQUOTE,HERE,SED}): New macros. (AC_DEFINE_{QUOTE,SEDQUOTE}): New macros; subrs of AC_DEFINE. (AC_DEFINE): Use them to quote $2. Wed Feb 17 14:49:14 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_TIMEZONE): Fixed quoting in tzname check. changequote inside quotes lost. Mon Feb 8 14:22:11 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acconfig.h (_ALL_SOURCE): Use #ifndef; AIX compiler way too dumb. Sun Jan 31 16:39:46 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_TIMEZONE): Put newlines before `#include ...' in $defs value. Thu Jan 28 18:06:53 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acconfig.h (_ALL_SOURCE): Use "!defined (_ALL_SOURCE) || _ALL_SOURCE == 0" rather than "!_ALL_SOURCE", which bombs on the AIX compiler. Mon Jan 25 12:09:43 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acconfig.h (HAVE_UNION_WAIT, SYS_SIGLIST_DECLARED): New #undef's. * acconfig.h (_ALL_SOURCE): Surround with #if !_ALL_SOURCE. Fri Jan 22 15:08:33 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): If /usr/local/lib/libgetloadavg.a exists, add -L/usr/local/lib to LDFLAGS. Fri Jan 22 12:49:11 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT_HEADER): Only comment out the #undef NAME part of the line, to avoid causing errors from existing comments. Thu Jan 21 14:50:20 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_HAVE_LIBRARY): Use $libname in "checking for" message, not $1, to avoid "checking for -l-lfoo". * acgeneral.m4 (AC_PREPARE): In compile defn, include $CFLAGS. * acgeneral.m4 (AC_OUTPUT): Broke AC_CONFIG_NAME writing out into: (AC_OUTPUT_HEADER): New macro broken out of AC_OUTPUT. Add to conftest.sed a new sed command to turn #undef's into comments. * acgeneral.m4 (AC_OUTPUT): Use new shell variable, $maxsedlines, for max number of lines to feed to one sed invocation. Lower this limit to 20; UTekV 3.2e can't cope with 40. Tue Jan 19 13:21:02 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * Version 1.3. Fri Jan 15 16:28:18 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_CONFIG_HEADER, AC_HEADER_EGREP, AC_TEST_PROGRAM): Make DEFS always contain -D commands, not C code. Thu Jan 14 17:05:17 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Check for -lkvm; don't assume it. Thu Jan 14 16:46:41 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh (selecting $syms from $TEMPLATES): Use sed to replace lines containing only blanks with empty lines. Thu Jan 14 15:15:31 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acspecific.m4 (AC_MODE_T): New macro. * acgeneral.m4 (AC_OUTPUT): Check for grep -c returning nothing (AIX 3.1) as well as returning 0. Wed Jan 13 16:05:59 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_FUNC_CHECK): Add missing #endif. * acgeneral.m4 (AC_OUTPUT): Use sed, not basename. From Francois Pinard. Wed Jan 13 15:49:18 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Set exec_prefix to ${prefix}, not $(prefix); it now works in both makefiles and shell scripts. Wed Jan 13 15:29:04 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * autoheader.sh: If input is empty, don't print all of acconfig.h. From Francois Pinard. * acgeneral.m4 (AC_OUTPUT): Have config.status check all of its args for validity. Tue Jan 12 11:11:45 1993 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Preserve whitespace around = in prefix and exec_prefix assignments. * acspecific.m4 (AC_GETLOADAVG): Values for getloadavg_missing were reversed. Fri Jan 8 18:45:59 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Make config.status not complain with usage msg when given no args. * acgeneral.m4 (AC_HAVE_LIBRARY): Say "checking for -lfoo", not just "checking for foo". * acgeneral.m4 (AC_HAVE_LIBRARY): Remove excess quoting around $2 and $3. * acspecific.m4 (AC_GETLOADAVG): Check for getloadavg library, both a normally installed one, and one in /usr/local/lib. After figuring out params for getloadavg.c, figure out whether it defined LDAV_PRIVILEGED, and if so, set NEED_SETGID to true, and define GETLOADAVG_PRIVILEGED. * acconfig.h: Added GETLOADAVG_PRIVILEGED. Fri Jan 8 16:16:35 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_DEFINE, AC_OUTPUT): Restore the third sed string. * acgeneral.m4 (AC_FUNC_CHECK): Use __stub_funcname. * autoheader.sh: Use Autoconf version number. * acgeneral.m4 (AC_OUTPUT): Diagnose usage errors for config.status. Use grep -c to count nonempty lines instead of test -s. * acspecific.m4 (AC_GETLOADAVG): Use AC_HAVE_LIBRARY. Wed Jan 6 19:54:47 1993 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * autoheader.sh (coverage check): Use $TEMPLATES in error msg, not hard-wired "config.h". Wed Jan 6 18:24:41 1993 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): If AC_CONFIG_NAME, change @DEFS@ to -DHAVE_CONFIG_H in Makefiles etc. Idea from Roland McGrath. * acgeneral.m4 (AC_FUNC_CHECK): If __STUB_funcname is defined, assume the function isn't present. * acgeneral.m4 (AC_OUTPUT): Make no args to AC_OUTPUT work again. From Ian Lance Taylor. * acspecific.m4 (AC_CONST): Fix quoting problem. * acconfig.h [const]: New addition. Thu Dec 31 17:56:18 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_HAVE_LIBRARY): New macro from Noah Friedman. * acconfig.h: Renamed from config.h. * autoheader.sh: Renamed from autohead.sh. Support a local acconfig.h. Use \\012 instead of \\n for tr for portability. Thu Dec 31 12:30:34 1992 Roland McGrath (roland@geech.gnu.ai.mit.edu) * config.h: Added #undef vfork. Tue Dec 29 14:26:43 1992 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_COMPILE_CHECK): Use cat rather than echo to create conftest.c, to avoid " problems. Fri Dec 25 15:07:06 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acspecific.m4 (AC_CONST): Don't define HAVE_CONST. * acgeneral.m4 (AC_OUTPUT, AC_DEFINE): Combine the two sed commands for #undef lines. * acgeneral.m4 (AC_PROGRAM_EGREP, AC_TEST_PROGRAM, AC_TEST_CPP, AC_OUTPUT), acspecific.m4 (AC_PROG_CC): Put > before << when using both, to avoid HP-UX sh bug. Wed Dec 23 20:47:53 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_PARSEARGS): Use if, not &&, for --with. From Jan Brittenson. Mon Dec 21 17:13:57 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Use sed instead of head and tail. Trap to remove the temp files. * acgeneral.m4 (AC_OUTPUT): Quote DEFS assignment. From Ian Lance Taylor. Mon Dec 21 14:27:44 1992 Jim Meyering (meyering@comco.com) * acspecific.m4 (AC_STDC_HEADERS): Make sure ctype.h macros are ANSI. Nest tests so we don't need shell temporary variable. Sun Dec 20 18:12:33 1992 Roland McGrath (roland@albert.gnu.ai.mit.edu) * Makefile.in (%.h: %.in): New rule using autohead. (all): Do autohead. (install): Install autohead and config.h. (autohead): New rule. (DISTFILES): Added autohead.sh. * autohead: New script. Fri Dec 18 00:21:23 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acgeneral.m4 (AC_HAVE_FUNCS, AC_HAVE_HEADERS): Change method of tr quoting to keep old shells happy. From Ian Lance Taylor. * acgeneral.m4 (AC_DEFINE): Add to SEDDEFS. (AC_OUTPUT): Use sed instead of awk. From Ian Lance Taylor. Mon Dec 14 14:33:29 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acspecific.m4 (AC_STDC_HEADERS): Check for string.h declaring memchr. * acgeneral.m4 (AC_NOTICE): Fix comment. Fri Dec 11 17:59:23 1992 David J. MacKenzie (djm@kropotkin.gnu.ai.mit.edu) * acspecific.m4 (AC_ALLOCA): Don't use libPW; it causes too much trouble. Wed Dec 9 14:04:30 1992 Roland McGrath (roland@geech.gnu.ai.mit.edu) * config.h: Added HAVE_SYS_WAIT, HAVE_WAITPID, SVR4, UMAX, [ugp]id_t, UMAX4_3, DGUX. Thu Dec 3 13:37:17 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_INSTALL): Ignore AFS install. From James Clark, jjc@jclark.com. Tue Nov 24 07:47:45 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_LEX, AC_DECLARE_YYTEXT, AC_VFORK, AC_WAIT3, AC_INT_16_BITS, AC_WORDS_BIGENDIAN, AC_ARG_ARRAY): End with a newline. * acspecific.m4 (AC_DIR_HEADER): If ndir.h exists and the other choices don't, define NDIR. Sat Nov 21 00:14:51 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_RETSIGTYPE): Instead of grepping for the signal declaration, try redeclaring it and see if we get an error. Always define RETSIGTYPE, not just if it's int. From Ian Lance Taylor. Fri Nov 20 17:06:09 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_DEFINE): Only put -D option in quotes if it actually contains blanks. Thu Nov 19 17:18:40 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_PARSEARGS): Set a shell var for --with-*. (AC_WITH): New macro. * acspecific.m4 (AC_CONST): If const works, define HAVE_CONST. * acspecific.m4 (AC_ALLOCA): Don't use libPW on HP-UX. Wed Nov 18 17:36:08 1992 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_DEFINE): When writing a -D with a value, surround it with 's so the value can contain spaces. Thu Nov 12 22:49:35 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_PROG_CC): Don't add -O to CC if GNU C. (-O2, or nothing, might be more appropriate.) Sun Nov 8 23:33:23 1992 david d `zoo' zuhn (zoo at cirdan.cygnus.com) * acspecific.m4 (AC_GETLOADAVG): Check for dwarf.h for general svr4, then elf.h for Solaris 2, which needs additional libraries. Thu Nov 12 22:18:54 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4 (AC_PARSEARGS): --exec_prefix -> --exec-prefix. Tue Nov 10 16:15:10 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4: undef m4 `include' builtin. * acspecific.m4 (AC_STDC_HEADERS): Don't test for limits.h due to Ultrix conflict with float.h. Thu Oct 29 16:16:11 1992 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_PARSEARGS, AC_PREPARE): New macros, broken out parts of AC_INIT. (AC_INIT): Use them. Thu Oct 22 20:48:12 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_INSTALL): Comment out arg to `:'. AIX doesn't like it. Wed Oct 14 12:41:02 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * Version 1.2. * acspecific.m4 (AC_INSTALL): Avoid the AIX install script. * acspecific.m4 (AC_RESTARTABLE_SYSCALLS): Wait for child if sys calls are not restarted, to avoid leaving the child still running. From Ian Lance Taylor. Tue Oct 13 15:43:56 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_CONST): Add more tests for brokenness. From Jim Meyering. * acgeneral.m4: Use % instead of ? to avoid shell variable expansion. Fri Oct 2 06:55:05 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acgeneral.m4: Use ? instead of , to separate parts of sed arg. Mon Sep 14 12:33:41 1992 David J. MacKenzie (djm@apple-gunkies.gnu.ai.mit.edu) * acspecific.m4 (AC_STDC_HEADERS): Also check for float.h. * acspecific.m4 (AC_TIMEZONE): Protect [] from being quotes. Thu Sep 10 17:12:10 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Include the hostname in config.status. * acgeneral.m4 (AC_OUTPUT): Use a separate flag in the awk script instead of checking for non-empty values, so things like defining const as empty work. From Steve Emmerson . Fri Aug 28 18:51:13 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acgeneral.m4 (AC_INIT): If there's no path on $0, use '.'. Thu Aug 27 16:15:14 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * config.h: New file. * acgeneral.m4 (AC_INIT): Look for source files in the directory containing `configure', if not given explicitly. * acspecific.m4 (AC_TIMEZONE): Adjust tzname decl for RS6000. * acspecific.m4 (AC_GETLOADAVG): Don't use double quotes in the test program. Thu Aug 27 15:26:49 1992 Roland McGrath (roland@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Don't check nlist.h if we found one of specific things. Mon Aug 24 16:22:45 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * Version 1.1. * acspecific.m4 (AC_TIMEZONE): Include time.h. Don't declare tzname if it's a macro. From Jim Meyering. Fri Aug 21 14:12:35 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_ALLOCA): Check whether the alloca defined by alloca.h works when given a non-constant argument. * acspecific.m4 (AC_GETLOADAVG): Define NLIST_STRUCT and NLIST_NAME_UNION if appropriate. * acgeneral.m4 (AC_OUTPUT): If no args are given, omit the loop to produce output files. * acgeneral.m4 (AC_TEST_PROGRAM): Add a call to exit to try to suppress core dumped message. From Ian Lance Taylor. * acgeneral.m4 (AC_PREFIX): Only print the message if prefix hasn't been set. From James Clark. * acspecific.m4 (AC_SIZE_T, AC_UID_T, AC_PID_T, AC_RETSIGTYPE): Print a message saying what it's checking for. (AC_SIZE_T): Define size_t to be unsigned, not int, for ANSI-friendliness. * acspecific.m4 (AC_GETLOADAVG): Just check for elf.h, not dwarf.h too. * autoconf.sh: Exit with status 1 if there are unresolved macros. Isolate the pattern to make adding other prefixes easy. Look for aclocal.m4 in . as well as MACRODIR. Tue Aug 18 16:35:46 1992 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_STRCOLL): New macro. Tue Aug 18 15:22:45 1992 Roland McGrath (roland@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): elf.h implies SVR4. Mon Jul 27 14:20:32 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acgeneral.m4 (AC_TEST_PROGRAM): Check for cross-compiling was missing "test -n". From Ian Lance Taylor. Sun Jul 26 16:25:19 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acgeneral.m4 (AC_SUBST): Support multiple substitutions in a line. Mon Jul 20 01:08:01 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * Version 1.0. * acspecific.m4 (AC_ALLOCA): Don't try -lucb -- it's too often broken. Sat Jul 18 13:40:46 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_MAJOR_HEADER): Add missing "$". * acspecific.m4 (AC_ALLOCA): Put -lc before -lucb. Fri Jul 17 00:00:07 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * autoconf.sh: Print the lines of the input file where unresolved macros occur. From Francois Pinard. * acspecific.m4 (AC_PROG_INSTALL), acgeneral.m4 (AC_PROGRAM_CHECK): Use test -f instead of -s. * autoconf.sh: grep for undefined macros in output. Tue Jul 14 01:19:26 1992 David J. MacKenzie (djm@apple-gunkies.gnu.ai.mit.edu) * acgeneral.m4 (AC_PROGRAM_CHECK): Search PATH manually to avoid "command not found" messages on /dev/tty. Remove "args for check" argument. (AC_INIT): Don't define checkfor; no longer needed. * acspecific.m4 (AC_PROG_CC, AC_PROG_RANLIB, AC_PROG_YACC, AC_PROG_LEX): Don't pass "args for check" argument. * acgeneral.m4 (AC_PROGRAMS_CHECK): New macro. * acspecific.m4 (AC_PROG_AWK): Use it. Check for mawk, gawk, nawk, and awk. (AC_PROG_YACC): Check for byacc if bison isn't found. * acspecific.m4 (AC_PROG_CC): Renamed from AC_PROG_GCC. (AC_PROG_YACC): Renamed from AC_PROG_BISON. (AC_PROG_AWK): Renamed from AC_PROG_GAWK. (AC_PROG_LEX): Renamed from AC_PROG_FLEX. * acgeneral.m4 (AC_TEST_PROGRAM): Redirect stderr to /dev/null both inside and outside the subshell to try to prevent core dumped messages. Who knows, it might even help. Thu Jul 9 21:37:45 1992 Roland McGrath (roland@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): Check for DGUX before SVR4. Fri Jul 3 01:01:50 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_REMOTE_TAPE): Renamed from AC_MTIO. Define HAVE_SYS_MTIO_H instead of NO_MTIO. (AC_LONG_FILE_NAMES): Renamed from AC_LONG_FILENAMES. (AC_RSH): Define HAVE_NETDB_H instead of USE_REXEC. Above mostly from Richard Stallman. * acgeneral.m4 (AC_MISSING_FUNCS): Macro removed. * acspecific.m4 (AC_VPRINTF, AC_WAIT3, AC_UTIME_NULL, AC_TIMEZONE, AC_ST_BLOCKS, AC_ST_BLKSIZE): Change from FOO_MISSING to HAVE_FOO. (AC_WAIT3): Renamed from AC_WAIT3_RUSAGE. (AC_TIMEZONE): Require AC_STRUCT_TM. (AC_STRUCT_TM): Provide itself. * acgeneral.m4 (AC_OUTPUT): Add --recheck option to config.status. * acspecific.m4 (AC_ST_RDEV, AC_CONST): New macros. * acgeneral.m4 (AC_DEFINE): Don't consider an empty value arg to be an omitted arg. Thu Jul 2 16:05:05 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_GETLOADAVG): New macro. * autoconf.sh: Only reject an arg that's not a known option if it is an option. Tue Jun 30 16:08:04 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_INIT, AC_OUTPUT): Eliminate vpsub. Thu Jun 25 12:42:10 1992 David J. MacKenzie (djm@apple-gunkies.gnu.ai.mit.edu) * autoconf.sh: Add --version option. * acgeneral.m4: Support it. Wed Jun 24 14:04:13 1992 David J. MacKenzie (djm@goldman.gnu.ai.mit.edu) * acspecific.m4 (AC_TIMEZONE): Do the checks unconditionally, not only if strftime.o is in LIBOBJS. * acspecific.m4 (AC_DIR_HEADER): Don't assume sys/dir.h exists. * acgeneral.m4 (AC_PROGRAM_CHECK): Don't include the program name in the value-if-found. From Rich Murphey. * acspecific.m4 (AC_PROG_{GCC,RANLIB,GAWK,BISON,FLEX}): Change callers. * acgeneral.m4 (AC_OUTPUT): Mention the args given to configure in a comment in config.status. Fri Jun 19 13:18:12 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_FUNC_CHECK): Use the third arg when it's non-null, not when it's null. From Ian Lance Taylor. Thu Jun 18 12:10:27 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_BEFORE): Print message in next-error format. From Franc,ois Pinard. * acgeneral.m4 (AC_PROGRAM_CHECK): If args-for-use is empty, don't put a space after the program name. * acspecific.m4 (AC_DECLARE_YYTEXT): Move AC_REQUIREs from AC_PROG_FLEX to here, where they belong. * acspecific.m4 (AC_MEMORY_H): Look for memchr instead of memcpy. From Karl Berry. Wed Jun 17 09:56:59 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_BEFORE): New macro. * acspecific.m4 (AC_PROG_GCC, AC_DIR_HEADER, AC_AIX, AC_MINIX, AC_ISC_POSIX): Use it. Tue Jun 16 14:46:29 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * autoconf.sh: Remove incomplete output file if interrupted. * acgeneral.m4 (AC_INIT): Avoid running an extra subshell for pwd. From Franc,ois Pinard. Mon Jun 15 21:27:49 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_TEST_PROGRAM): Take another arg for cross-compiling. * acspecific.m4 (AC_CROSS_CHECK): New program. * acgeneral.m4 (AC_REQUIRE, AC_PROVIDE): New macros. (AC_HEADER_EGREP, AC_PROGRAM_EGREP, AC_TEST_CPP): Use them. * acspecific.m4 (AC_PROG_GCC, AC_GCC_TRADITIONAL, AC_PROG_CPP, AC_PROG_FLEX, AC_INLINE): Ditto. Sat Jun 13 17:54:24 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_MEMORY_H): echo what it's doing. Thu Jun 11 14:18:35 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (AC_MINUS_C_MINUS_O, AC_INLINE, AC_SETVBUF_REVERSED): New macros. (AC_ALLOCA): Define HAVE_ALLOCA_H if appropriate. * acgeneral.m4 (AC_INIT): Do pwd in the srcdir, not current dir. Scan through "$@" (implicitly) instead of $*. (AC_HEADER_EGREP, AC_PROGRAM_EGREP, AC_COMPILE_CHECK, AC_TEST_PROGRAM, AC_TEST_CPP): Supply a `:' if `true' argument is empty. * acgeneral.m4, acspecific.m4: Omit `:' in callers. Wed Jun 10 12:03:11 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_HEADER_EGREP, AC_PROGRAM_EGREP, AC_HEADER_CHECK, AC_COMPILE_CHECK, AC_TEST_PROGRAM, AC_TEST_CPP, AC_FUNC_CHECK): Make the last argument (program to run if test fails) optional. (AC_HAVE_FUNCS, AC_HAVE_HEADERS): Don't pass optional last args. * acspecific.m4 (most macros): Likewise. Mon Jun 8 16:27:10 1992 David J. MacKenzie (djm@nutrimat.gnu.ai.mit.edu) * acspecific.m4 (AC_VFORK): Get rid of backquotes. * acgeneral.m4 (AC_OUTPUT): Exit with 0 status when --no-create was given. Only write to the AC_CONFIG_NAME file if it doesn't exist or is different from what we'd write. From Ian Lance Taylor. Thu Jun 4 14:46:22 1992 David J. MacKenzie (djm@geech.gnu.ai.mit.edu) * acspecific.m4 (AC_UNISTD_H, AC_UID_T, AC_SIZE_T, AC_PID_T, AC_ST_BLKSIZE, AC_STRUCT_TM): Quote the whole macro body. * acgeneral.m4 (AC_OUTPUT): Look for config header.in in top_srcdir, not srcdir. From Garrett Wollman. * acgeneral.m4 (AC_OUTPUT): Don't add make .NOEXPORT rule to output files. * acgeneral.m4, acspecific.m4: Rename AC_PROG_CHECK to AC_PROGRAM_CHECK, AC_PROG_EGREP to AC_PROGRAM_EGREP, AC_TEST_PROG to AC_TEST_PROGRAM. Wed Jun 3 14:00:07 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acgeneral.m4 (AC_INIT, AC_OUTPUT): Add --no-create option. * acgeneral.m4 (AC_COMPILE_CHECK): Check the C compiler exit status instead of trying to run the test program. * acspecific.m4 (AC_RESTARTABLE_SYSCALLS): Use AC_TEST_PROG instead of doing it by hand. * acspecific.m4 (AC_PROG_GCC, AC_AIX, AC_XENIX_DIR, AC_SCO_INTL, AC_DYNIX_SEQ): Use AC_PROG_EGREP instead of AC_TEST_PROG. * acgeneral.m4 (AC_TEST_PROG): Renamed from AC_TEST_PROGRAM. * acgeneral.m4 (AC_INIT): Don't relativize `.'. (AC_OUTPUT): Substitute the subdirectory path, not the top path, for srcdir, unless the top path is `.'. * acgeneral.m4 (AC_OUTPUT): Special-case substituting DEFS. From Ian Lance Taylor. * acspecific.m4 (AC_GCC_TRADITIONAL): Use CPP instead of compiling a test program. * acgeneral.m4 (AC_TEST_CPP): Pass DEFS to CPP. (AC_HEADER_EGREP): Don't echo anything. (AC_PROG_EGREP): New macro. Tue Jun 2 14:07:27 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_CONFIG_HEADER): Renamed from AC_HEADER_FILE. Rename AC_HEADER_NAME TO AC_CONFIG_NAME. (AC_SUBST): Add the arg variable to a diversion for config.status. (AC_OUTPUT): Write the code to create output files into config.status, then run that. Always use `awk'; checking for nawk in a subshell doesn't seem to work on 4.3BSD. * acgeneral.m4 (AC_HEADER_EGREP): Pass DEFS to CPP. * acspecific.m4 (AC_SIZE_T): Define size_t as int, not long. From Ian Lance Taylor. * acspecific.m4 (AC_STDC_HEADERS): Also check for stdarg.h. From Garrett Wollman. Wed May 20 00:34:03 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Use nawk if available. * acgeneral.m4 (AC_INIT): Make srcdir=`.' absolute. * acspecific.m4 (AC_DIR_HEADER): Include sys/types.h before dir header in closedir test. * acgeneral.m4, acspecific.m4: AC_LIBTHING_CHECK renamed to AC_COMPILE_CHECK. * acspecific.m4 (AC_AIX, AC_XENIX_DIR, AC_SCO_INTL, AC_DYNIX_SEQ): Use the C preprocessor instead of just looking for files. Mon May 18 20:51:50 1992 David J. MacKenzie (djm@wookumz.gnu.ai.mit.edu) * acspecific.m4 (WORDS_BIGENDIAN): Fix exit expression. (AC_DECLARE_YYTEXT): Eval $CPP. (AC_DIR_HEADER): Compile the test program; don't just preprocess it. Above all from Karl Berry. Fri May 15 00:57:01 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_AIX): Don't define _BSD, to avoid getting union wait. * acgeneral.m4 (AC_HEADER_EGREP): New macro, replacing AC_IDENT*. * acspecific.m4 (AC_MEMORY_H, AC_RETSIGTYPE, AC_{UID,SIZE,PID}_T): Use it. * acgeneral.m4 (AC_TEST_CPP): New macro. (AC_IDENT_{PRESENT,MISSING}): Macros deleted. (AC_HEADER_CHECK): Use AC_TEST_CPP, replaces AC_HEADER_{PRESENT, MISSING}. (AC_LIBTHING_CHECK): Replace AC_LIBTHING_{PRESENT,MISSING}. (AC_FUNC_CHECK): Replace AC_FUNC_PRESENT. (AC_INIT): Don't set INCLUDEDIR. * acspecific.m4 (AC_DIR_HEADER): Use AC_TEST_CPP. * All other macros: Don't refer to INCLUDEDIR; use AC_HEADER_CHECK instead. * acspecific.m4 (AC_PROG_CPP): Don't evaluate $CC until called. Try $CC -E before /lib/cpp. Thu May 14 23:15:02 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Remove each file before creating it. Sat May 9 14:52:57 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acspecific.m4 (AC_WAIT3_RUSAGE): New macro. * acgeneral.m4 (AC_INIT, AC_OUTPUT): Use AC_SUBST instead of a special mechanism to substitute for srcdir. * acgeneral.m4 (AC_OUTPUT): Substitute for exec_prefix if it was given, even if not substituting for prefix. * acgeneral.m4 (AC_INIT, AC_OUTPUT): Remove @VPATH@ substitution; use @srcdir@ instead. Sun May 3 01:21:47 1992 David J. MacKenzie (djm@geech.gnu.ai.mit.edu) * acgeneral.m4 (AC_NOTICE): New macro taken from AC_INIT, to avoid m4 coredump. From Karl Berry. * acgeneral.m4 (AC_OUTPUT): Look for header-file.in in $srcdir, not current dir. * acgeneral.m4 (AC_IDENT_{MISSING,PRESENT}): Make them agree with the documentation -- the third arg is a shell command, not an identifier to define. * acspecific.m4 (AC_DIR_HEADER): Change the caller. Mon Apr 27 09:15:15 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acspecific.m4: Use AC_TEST_PROGRAM wherever $compile was being used directly. * acgeneral.m4 (AC_HAVE_HEADERS, AC_HAVE_FUNCS, AC_FUNC_PRESENT, AC_TEST_PROGRAM): New macros from Ian Lance Taylor. * acspecific.m4 (AC_PROG_INSTALL): Screen out /usr/sbin/install. (AC_CHAR_UNSIGNED): Don't define __CHAR_UNSIGNED__ if it's predefined. Fri Apr 24 10:08:21 1992 David J. MacKenzie (djm@churchy.gnu.ai.mit.edu) * acgeneral.m4 (AC_OUTPUT): Make the arg a list of files instead of directories. * acspecific.m4 (AC_ALLOCA): Check whether the alternate libraries actually contain alloca. From Ian Lance Taylor. * acspecific.m4 (AC_PROG_CPP): New macro. * acgeneral.m4 (AC_OUTPUT): Allow newly defined values to be more than one word for AC_HEADER_FILE. From Karl Berry. * acgeneral.m4 (AC_OUTPUT): Don't substitute DEFS if AC_HEADER_FILE. (AC_LIBTHING{PRESENT,MISSING}): Run conftest in subshell. From Ian Lance Taylor. Local Variables: mode: indented-text left-margin: 8 version-control: never End: autoconf-1.7/autoconf.texi100666 21570 1 304306 5463274640 14242 0ustar djmdaemon\input texinfo @c -*-texinfo-*- @c %**start of header @setfilename autoconf.info @settitle Autoconf @setchapternewpage odd @c %**end of header @c Use on instead of odd in the setchapternewpage for single-sided printing. @set EDITION 1.7 @set VERSION 1.7 @set UPDATED October 1993 @iftex @finalout @end iftex @ifinfo This file documents the GNU Autoconf package for creating scripts to configure source code packages using templates and an @code{m4} macro package. Copyright (C) 1992, 1993 Free Software Foundation, Inc. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. @ignore Permission is granted to process this file through TeX and print the results, provided the printed document carries copying permission notice identical to this one except for the removal of this paragraph (this paragraph not being relevant to the printed manual). @end ignore Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Foundation. @end ifinfo @titlepage @title Autoconf @subtitle Generating Automatic Configuration Scripts @subtitle Edition @value{EDITION}, for Autoconf version @value{VERSION} @subtitle @value{UPDATED} @author by David MacKenzie, Roland McGrath, and Noah Friedman @page @vskip 0pt plus 1filll Copyright @copyright{} 1992, 1993 Free Software Foundation, Inc. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Foundation. @end titlepage @c Define a macro index that @@defmacro doesn't write to. @defcodeindex ma @node Top, Introduction, , (dir) @comment node-name, next, previous, up @ifinfo This file documents the GNU Autoconf package for creating scripts to configure source code packages using templates and an @code{m4} macro package. This is edition @value{EDITION}, for Autoconf version @value{VERSION}. @end ifinfo @c The master menu, created with texinfo-master-menu, goes here. @menu * Introduction:: Autoconf's purpose, strengths, and weaknesses. * Distributing:: Legal restrictions on Autoconf output. * Making configure Scripts:: How to organize and produce Autoconf scripts. * Specific Tests:: Macros that check for particular features. * General Purpose Macros:: Macros that check for kinds of features. * Writing Macros:: How to add your own macros to Autoconf. * Makefiles:: Information Autoconf uses in @file{Makefile}s. * Running configure Scripts:: How to use the Autoconf output. * Example:: Sample Autoconf input files. * Preprocessor Symbol Index:: Index of C preprocessor symbols defined. * Macro Index:: Index of Autoconf macros. --- The Detailed Node Listing --- Making @code{configure} Scripts * Writing configure.in:: What to put in an Autoconf input file. * Invoking autoconf:: How to create configuration scripts. * Invoking autoheader:: How to create configuration header files. Specific Tests * Alternative Programs:: Selecting between alternative programs. * Header Files:: Header files that might be missing. * Typedefs:: @code{typedef}s that might be missing. * Library Functions:: C library functions that might be missing. * Structures:: Structures or members that might be missing. * Compiler Characteristics:: C compiler or machine architecture features. * System Services:: Operating system services. * UNIX Variants:: Special cases for specific UNIX variants. General Purpose Macros * Setup:: Controlling Autoconf operation. * General Tests:: Check for kinds of features. * Setting Variables:: Setting shell and @code{make} variables. * Macro Ordering:: Enforcing ordering constraints. Writing Macros * Macro Format:: Basic format of an Autoconf macro. * Quoting:: Protecting macros from unwanted expansion. * Dependencies Between Macros:: What to do when macros depend on other macros. * Checking for Files:: Finding whether a file exists. * Checking for Symbols:: Finding whether a symbol is defined. * Test Programs:: Writing programs to test for features. * Multiple Cases:: Tests for several possible values. Dependencies Between Macros * Prerequisite Macros:: Ensuring required information. * Suggested Ordering:: Warning about possible ordering problems. Test Programs * Guidelines:: General rules for writing test programs. * Tricks:: Special ways to work around problems. Makefiles * Predefined Variables:: Heavily used @code{make} variables. * Installation Prefixes:: A special variable substitution. * VPATH Substitutions:: Compiling in a different directory. * Automatic Remaking:: Makefile rules for configuring. Running @code{configure} Scripts * Overriding variables:: Workarounds for unusual systems. * Running config.status:: Recreating a configuration. An Example * Sample configure.in:: An example of a @file{configure} template. * Sample Makefile.in:: An example of a @file{Makefile} template. @end menu @node Introduction, Distributing, Top, Top @chapter Introduction Autoconf is a tool for producing shell scripts that automatically configure software source code packages to adapt to many kinds of UNIX-like systems. The configuration scripts produced by Autoconf are independent of Autoconf when they are run, so their users do not need to have Autoconf. The configuration scripts produced by Autoconf normally require no manual user intervention when run; they do not even take an argument specifying the system type. Instead, they test for the presence of each feature that the software package they are for might need individually. (Before each check, they print a one-line message stating what they are checking for, so the user doesn't get too bored while waiting for the script to finish.) As a result, they deal well with systems that are hybrids or customized from the more common UNIX variants. There is no need to maintain files that list the features supported by each release of each variant of UNIX. For each software package that Autoconf is used with, it creates a configuration script from a template file that lists the operating system features that the package can use. After the shell code to recognize and respond to an operating system feature has been written, Autoconf allows it to be shared by many software packages that can use (or need) that feature. If it later turns out that the shell code needs adjustment for some reason, it needs to be changed in only one place; all of the the configuration scripts can be regenerated automatically to take advantage of the updated code. Larry Wall's Metaconfig package is similar in purpose to Autoconf, but is more general. The scripts it produces require manual user intervention, which is quite inconvenient when configuring large source trees. Unlike Metaconfig scripts, Autoconf scripts can support cross-compiling, if some care is taken in writing them. They should avoid executing test programs, since test programs compiled with a cross-compiler can not be executed on the host system. Also, they shouldn't do anything that tests features of the host system instead of the target system. Autoconf imposes some restrictions on the names of macros used with @code{#ifdef} in C programs (@pxref{Preprocessor Symbol Index}). Autoconf requires GNU @code{m4} in order to generate the scripts. It uses features that some UNIX versions of @code{m4} do not have; it is also reported to overflow internal limits of some versions of @code{m4}. Autoconf does not work well with GNU C library releases before 1.06. The GNU C library contains stubs (which always return an error) for functions that are not available instead of omitting them from the library. As a result, Autoconf scripts are fooled into thinking that those functions are available. This problem does not exist with releases 1.06 and later of the GNU C library, which define C preprocessor macros that the Autoconf macros @code{AC_FUNC_CHECK} and @code{AC_REPLACE_FUNCS} test, indicating that certain functions are stubs (@pxref{General Tests}, for more information on checking for functions). @ifinfo Autoconf was written by David MacKenzie, with help from Franc,ois @end ifinfo @tex Autoconf was written by David MacKenzie, with help from Fran\c cois @end tex Pinard, Karl Berry, Richard Pixley, Ian Lance Taylor, Roland McGrath, and Noah Friedman. It was inspired by Brian Fox's automatic configuration system for Bash, by Larry Wall's Metaconfig, and by Richard Stallman, Richard Pixley, and John Gilmore's configuration tools for the GNU compiler and object file utilities. Mail suggestions and bug reports for Autoconf to @code{bug-gnu-utils@@prep.ai.mit.edu}. Please include the Autoconf version number, which you can get by running @samp{autoconf --version}. @node Distributing, Making configure Scripts, Introduction, Top @chapter Distributing Autoconf Output The configuration scripts that Autoconf produces are covered by the GNU General Public License. This is because they consist almost entirely of parts of Autoconf itself, rearranged somewhat, and Autoconf is distributed under the terms of the GPL. As applied to Autoconf, the GPL just means that you need to distribute @file{configure.in} along with @file{configure}. Programs that use Autoconf scripts to configure themselves do not automatically come under the GPL. Distributing an Autoconf configuration script as part of a program is considered to be @emph{mere aggregation} of that work with the Autoconf script. Such programs are not derivative works based on Autoconf; only their configuration scripts are. We still encourage software authors to distribute their work under terms like those of the GPL, but doing so is not required to use Autoconf. @node Making configure Scripts, Specific Tests, Distributing, Top @chapter Making @code{configure} Scripts The configuration scripts that Autoconf produces are by convention called @code{configure} when they are distributed. When run, they create several files: @itemize @bullet @item one or more @file{Makefile} files (one in each subdirectory of the package), from template @file{Makefile.in} files (@pxref{Makefiles}); @item optionally, a C header file, the name of which is configurable, containing @code{#define} statements (@pxref{Setup}); @item a shell script called @file{config.status} that, when run, will recreate the current configuration parameter settings (@pxref{Running config.status}). @end itemize To create a @code{configure} script with Autoconf, you need to write an Autoconf input file and run Autoconf on it to produce the script. And, of course, test the resulting script. Here is a diagram showing how the files that can be used in configuration are produced: @example acgeneral.m4 \ Makefile.in \ acspecific.m4 \ \ autoconf* -> m4* -> configure* -> config.status* -> Makefile \ configure.in / | | \ | | | | make* -> your | | | | /package | | config.status* -> config.h / configure.in \ / autoheader* -> - - - - - - - - - - -> config.h.in / acconfig.h / @end example @noindent Executables are suffixed by @samp{*}, while files appearing twice are linked with lines of @samp{|}. @menu * Writing configure.in:: What to put in an Autoconf input file. * Invoking autoconf:: How to create configuration scripts. * Invoking autoheader:: How to create configuration header files. @end menu @node Writing configure.in, Invoking autoconf, , Making configure Scripts @section Writing @file{configure.in} To produce a @code{configure} script for a software package, create a file called @file{configure.in} that contains invocations of the Autoconf macros that test the system features your package needs or can use. Autoconf macros already exist to check for many features; see @ref{Specific Tests}, for their descriptions. For most other features, you can use Autoconf template macros to produce custom checks; see @ref{General Tests}, for information about them. For especially tricky or specialized features, @file{configure.in} might need to contain some hand-crafted shell commands. @xref{Writing Macros}, for guidelines on writing tests from scratch. Every @file{configure.in} must begin with a call to @code{AC_INIT} and end with a call to @code{AC_OUTPUT} (@pxref{Setup}). Other than that, the order in which @file{configure.in} calls the Autoconf macros is generally not important, except that some macros rely on other macros having been called first, because they check previously set values of some variables to decide what to do. These macros are noted in the individual descriptions (@pxref{Specific Tests}). To encourage consistency, here is a suggested order for calling the Autoconf macros. A few macros need to be called in a different order from the one given here; they are noted in their individual descriptions (@pxref{Specific Tests}). Note that there must not be any space between the macro name and the open parentheses. @display @group @code{AC_INIT(@var{file})} checks for alternative programs checks for UNIX variants that set C preprocessor variables checks for header files checks for typedefs checks for library functions checks for structures checks for compiler characteristics checks for system services other checks for UNIX variants @code{AC_OUTPUT(@r{[}@var{file@dots{}}@r{]})} @end group @end display You can include comments in @file{configure.in} files by starting them with the @code{m4} predefined macro @code{dnl}, which discards text up through the next newline. These comments do not appear in the generated @code{configure} scripts. For example, it is helpful to begin @file{configure.in} files with a line like this: @example dnl Process this file with autoconf to produce a configure script. @end example @xref{Sample configure.in}, for an example of a real @file{configure.in} script. @node Invoking autoconf, Invoking autoheader, Writing configure.in, Making configure Scripts @section Invoking @code{autoconf} To create @code{configure} from @file{configure.in}, run the @code{autoconf} program with no arguments. @code{autoconf} processes @file{configure.in} with the @code{m4} macro processor, using the Autoconf macros. If you give @code{autoconf} an argument, it reads that file instead of @file{configure.in} and writes the configuration script to the standard output instead of to @file{configure}. If you give @code{autoconf} the argument @samp{-}, it reads the standard input instead of @file{configure.in} and writes the configuration script on the standard output. The Autoconf macros are defined in two or more files. Two of the files are distributed with Autoconf: @code{autoconf} first reads @file{acgeneral.m4} (@pxref{General Purpose Macros}), then @file{acspecific.m4} (@pxref{Specific Tests}). After reading them, @code{autoconf} looks for an optional file called @file{aclocal.m4}, first in the directory that contains other installed Autoconf macro files, and then in the current directory. If both files exist, it uses both of them. Those files can contain your site's own locally written Autoconf macro definitions (@pxref{Writing Macros}, for more information). If a macro is defined in more than one of the files that @code{autoconf} reads, the last definition it reads overrides the earlier ones. You can override the directory where @code{autoconf} looks for the installed macro files by setting the @code{AC_MACRODIR} environment variable to a different directory. You can also give @code{autoconf} the @samp{--macrodir} option, which overrides @code{AC_MACRODIR}. @code{autoconf} also accepts the options @code{--version}, which prints the Autoconf version number and exits, and @code{--help}, which prints a summary of the command-line options and exits. @node Invoking autoheader, , Invoking autoconf, Making configure Scripts @section Invoking @code{autoheader} You can use the program @code{autoheader} to create a template file of C @samp{#define} statements for @code{configure} to use. By default, the file that @code{autoheader} creates is called @file{config.h.in}; if @file{configure.in} invokes @code{AC_CONFIG_HEADER(@var{file})}, @code{autoheader} creates @file{@var{file}.in}. @code{autoheader} scans @file{configure.in} and figures out which C preprocessor symbols it might define. It copies comments and @code{#define} and @code{#undef} statements from a file called @file{acconfig.h}, which comes with and is installed with Autoconf. It also uses a file called @file{acconfig.h} in the current directory, if present; you must create that file to contain entries for symbols that you @code{AC_DEFINE}. For symbols that @code{AC_HAVE_HEADERS} or @code{AC_HAVE_FUNCS} define, @code{autoheader} generates comments and @code{#undef} statements itself rather than copying them from a file, since the possible symbols are effectively limitless. The file that @code{autoheader} creates contains only @code{#define} and @code{#undef} statements and their accompanying comments; there is no provision in @code{autoheader} for adding other code or comments to the file. If you give @code{autoheader} an argument, it uses that file instead of @file{configure.in} and writes the header file to the standard output instead of to @file{config.h.in}. If you give @code{autoheader} an argument of @samp{-}, it reads the standard input instead of @file{configure.in} and writes the header file to the standard output. You can override the directory where @code{autoheader} looks for the installed macro and @file{acconfig.h} files by setting the @code{AC_MACRODIR} environment variable to a different directory. You can also give @code{autoheader} the @samp{--macrodir} option, which overrides @code{AC_MACRODIR}. @code{autoheader} also accepts the options @code{--version}, which prints the Autoconf version number and exits, and @code{--help}, which prints a summary of the command-line options and exits. @node Specific Tests, General Purpose Macros, Making configure Scripts, Top @chapter Specific Tests These macros test for particular operating system features that packages might need or want to use. If you need to test for a feature that none of these macros check for, you can probably do it by calling one of the general purpose test macros with appropriate arguments (@pxref{General Tests}). All of these macros that set @code{make} variables call @code{AC_SUBST} on those variables (@pxref{Setting Variables}, for details about @code{AC_SUBST}). The phrase ``define @var{name}'' is used below as a shorthand to mean either add @samp{-D@var{name}=1} to the @code{make} variable @code{DEFS}, or put @samp{#define @var{name} 1} in the configuration header file, depending on whether @code{AC_CONFIG_HEADER} has been called. @xref{Setting Variables}, for more information. Within each section below, the macros are listed in alphabetical order. The macros are generally named for the @code{make} variables or C preprocessor macros that they define; those names are based largely on what existing GNU programs use. These macros are defined in the file @file{acspecific.m4}. @menu * Alternative Programs:: Selecting between alternative programs. * Header Files:: Header files that might be missing. * Typedefs:: @code{typedef}s that might be missing. * Library Functions:: C library functions that might be missing. * Structures:: Structures or members that might be missing. * Compiler Characteristics:: C compiler or machine architecture features. * System Services:: Operating system services. * UNIX Variants:: Special cases for specific UNIX variants. @end menu @node Alternative Programs, Header Files, , Specific Tests @section Alternative Programs The following macros check for the presence or behavior of particular programs: @defmac AC_DECLARE_YYTEXT @maindex DECLARE_YYTEXT @vindex DECLARE_YYTEXT Define @code{DECLARE_YYTEXT} to declare @code{yytext} appropriately, depending on whether @code{lex} or @code{flex} is being used. This macro calls @code{AC_PROG_CPP} and @code{AC_PROG_LEX} if they haven't been called already. @end defmac @defmac AC_LN_S @maindex LN_S If @samp{ln -s} works on the current filesystem (the O.S. and filesystem support symbolic links), set shell and @code{make} variable @code{LN_S} to @samp{ln -s}, otherwise set it to @samp{ln}. @end defmac @defmac AC_MINUS_C_MINUS_O @maindex MINUS_C_MINUS_O @vindex NO_MINUS_C_MINUS_O If the C compiler does not accept the @samp{-c} and @samp{-o} options simultaneously, define @code{NO_MINUS_C_MINUS_O}. @end defmac @defmac AC_PROG_YACC @maindex PROG_YACC If @code{bison} is found, set @code{make} variable @code{YACC} to @samp{bison -y}. Otherwise, if @code{byacc} is found, set @code{YACC} to @samp{byacc}. Otherwise set @code{YACC} to @samp{yacc}. @end defmac @defmac AC_PROG_CPP @maindex PROG_CPP Set shell and @code{make} variable @code{CPP} to a command that runs the C preprocessor. If @samp{$CC -E} doesn't work, it uses @file{/lib/cpp}. Many of the specific test macros use the value of @code{CPP} indirectly by calling @code{AC_TEST_CPP}, @code{AC_HEADER_CHECK}, @code{AC_HEADER_EGREP}, or @code{AC_PROGRAM_EGREP}. Those macros call this macro first if it hasn't been called already. It calls @code{AC_PROG_CC} if it hasn't been called already. @end defmac @defmac AC_PROG_LEX @maindex PROG_LEX If @code{flex} is found, set @code{make} variable @code{LEX} to @samp{flex} and @code{LEXLIB} to @samp{-lfl}, if that library is in a standard place. Otherwise set @code{LEX} to @samp{lex} and @code{LEXLIB} to @samp{-ll}. @end defmac @defmac AC_PROG_AWK @maindex PROG_AWK Check for @code{mawk}, @code{gawk}, @code{nawk}, and @code{awk}, in that order, and set @code{make} variable @code{AWK} to the first one that it finds. @end defmac @defmac AC_PROG_CC @maindex PROG_CC If @code{gcc} is found, set @code{make} variable @code{CC} to @samp{gcc}, and set shell variable @code{GCC} to 1 for use by macros such as @code{AC_GCC_TRADITIONAL}. @end defmac @defmac AC_GCC_TRADITIONAL @maindex GCC_TRADITIONAL Add @samp{-traditional} to @code{make} variable @code{CC} if using the GNU C compiler and @code{ioctl} does not work properly without @samp{-traditional}. This macro calls @code{AC_PROG_CC} and @code{AC_PROG_CPP} if they haven't been called already. @end defmac @defmac AC_PROG_INSTALL @maindex PROG_INSTALL Set @code{make} variable @code{INSTALL} to @samp{install -c} if @code{install} is found, otherwise set it to @samp{cp}. Screens out the false matches @file{/etc/install}, @file{/usr/sbin/install}, and other instances of @code{install} known not to work. The variable @code{INSTALL_PROGRAM} is set to @samp{$(INSTALL)} and @code{INSTALL_DATA} is set to @samp{$(INSTALL) -m 644}. @end defmac @defmac AC_PROG_RANLIB @maindex PROG_RANLIB Set @code{make} variable @code{RANLIB} to @samp{ranlib} if @code{ranlib} is found, otherwise to @samp{:} (do nothing). @end defmac @defmac AC_RSH @maindex RSH @vindex NO_REMOTE @vindex HAVE_NETDB_H If a remote shell is available, put @samp{rtapelib.o} in @code{make} variable @code{RTAPELIB}. Otherwise, also do so if @file{netdb.h} exists (implying the @code{rexec} function), and in addition define @code{HAVE_NETDB_H}. If neither a remote shell nor @code{rexec} is available, define @code{NO_REMOTE}. @end defmac @node Header Files, Typedefs, Alternative Programs, Specific Tests @section Header Files The following macros check for the presence of certain C header files: @defmac AC_DIR_HEADER @maindex DIR_HEADER @vindex DIRENT @vindex SYSDIR @vindex SYSNDIR @vindex NDIR @vindex VOID_CLOSEDIR If the system has @file{dirent.h}, define @code{DIRENT}; otherwise, if it has @file{sys/ndir.h}, define @code{SYSNDIR}; otherwise, if it has @file{sys/dir.h}, define @code{SYSDIR}; otherwise, if it has @file{ndir.h}, define @code{NDIR}. Also, if the directory library header file contains a declaration of the @code{closedir} function with a @code{void} return type, define @code{VOID_CLOSEDIR}. The directory library declarations in the source code should look something like the following, which assumes that you have also called @samp{AC_HAVE_HEADERS(unistd.h)}: @example @group #ifdef HAVE_UNISTD_H #include #include #endif /* unistd.h defines _POSIX_VERSION on POSIX.1 systems. */ #if defined(DIRENT) || defined(_POSIX_VERSION) #include #define NLENGTH(dirent) (strlen((dirent)->d_name)) #else /* not (DIRENT or _POSIX_VERSION) */ #define dirent direct #define NLENGTH(dirent) ((dirent)->d_namlen) #ifdef SYSNDIR #include #endif /* SYSNDIR */ #ifdef SYSDIR #include #endif /* SYSDIR */ #ifdef NDIR #include #endif /* NDIR */ #endif /* not (DIRENT or _POSIX_VERSION) */ @end group @end example Using the above declarations, the program would declare variables to be type @code{struct dirent}, not @code{struct direct}, and would access the length of a directory entry name by passing a pointer to a @code{struct dirent} to the @code{NLENGTH} macro. @end defmac @defmac AC_MAJOR_HEADER @maindex MAJOR_HEADER @vindex MAJOR_IN_MKDEV @vindex MAJOR_IN_SYSMACROS If @file{sys/types.h} does not define @code{major}, @code{minor}, and @code{makedev}, but @file{sys/mkdev.h} does, define @code{MAJOR_IN_MKDEV}; otherwise, if @file{sys/sysmacros.h} does, define @code{MAJOR_IN_SYSMACROS}. @end defmac @defmac AC_MEMORY_H @maindex MEMORY_H @vindex NEED_MEMORY_H Define @code{NEED_MEMORY_H} if @code{memcpy}, @code{memcmp}, etc. are not declared in @file{string.h} and @file{memory.h} exists. This macro is obsolete; instead, use @code{AC_HAVE_HEADERS(memory.h)}. See the example for @code{AC_STDC_HEADERS}. @end defmac @defmac AC_STDC_HEADERS @maindex STDC_HEADERS @vindex STDC_HEADERS Define @code{STDC_HEADERS} if the system has ANSI C header files. Specifically, this macro checks for @file{stdlib.h}, @file{stdarg.h}, @file{string.h}, and @file{float.h}; if the system has those, it probably has the rest of the ANSI C header files. This macro also checks whether @file{string.h} declares @code{memchr} (and thus presumably the other @code{mem} functions) and whether the @file{ctype.h} macros work on characters with the high bit set, as ANSI C requires. Use @code{STDC_HEADERS} instead of @code{__STDC__} to determine whether the system has ANSI-compliant header files (and probably C library functions) because many systems that have GCC do not have ANSI C header files. To check whether to use the System V/ANSI C string functions and header file, you can put the following in @file{configure.in}: @example AC_STDC_HEADERS AC_HAVE_HEADERS(string.h memory.h) @end example @noindent Then, in the code, use a test like this: @example @group #if STDC_HEADERS || HAVE_STRING_H #include /* An ANSI string.h and pre-ANSI memory.h might conflict. */ #if !STDC_HEADERS && HAVE_MEMORY_H #include #endif /* not STDC_HEADERS and HAVE_MEMORY_H */ #define index strchr #define rindex strrchr #define bcopy(s, d, n) memcpy ((d), (s), (n)) #define bcmp(s1, s2, n) memcmp ((s1), (s2), (n)) #define bzero(s, n) memset ((s), 0, (n)) #else /* not STDC_HEADERS and not HAVE_STRING_H */ #include /* memory.h and strings.h conflict on some systems. */ #endif /* not STDC_HEADERS and not HAVE_STRING_H */ @end group @end example @noindent This example asssumes that your code uses the BSD style functions. If you use the System V/ANSI C style functions, you will need to replace the macro definitions with ones that go in the other direction. @end defmac @defmac AC_UNISTD_H @maindex UNISTD_H @vindex HAVE_UNISTD_H Define @code{HAVE_UNISTD_H} if the system has @file{unistd.h}. This macro is obsolete; instead, use @samp{AC_HAVE_HEADERS(unistd.h)}. The way to check if the system supports POSIX.1 is: @example @group #if HAVE_UNISTD_H #include #include #endif #ifdef _POSIX_VERSION /* Code for POSIX.1 systems. */ #endif @end group @end example @vindex _POSIX_VERSION @code{_POSIX_VERSION} is defined when @file{unistd.h} is included on POSIX.1 systems. If there is no @file{unistd.h}, it is definitely not a POSIX.1 system. However, some non-POSIX.1 systems do have @file{unistd.h}. @end defmac @defmac AC_USG @maindex USG @vindex USG Define @code{USG} if the system does not have @file{strings.h}, @code{rindex}, @code{bzero}, etc. This implies that it has @file{string.h}, @code{strrchr}, @code{memset}, etc. The symbol @code{USG} is obsolete. Instead of this macro, use @code{AC_HAVE_HEADERS(string.h)} and use @code{HAVE_STRING_H} in your code. See the example for @code{AC_STDC_HEADERS}. @end defmac @node Typedefs, Library Functions, Header Files, Specific Tests @section Typedefs The following macros check for predefined C types: @defmac AC_GETGROUPS_T @maindex GETGROUPS_T @vindex GETGROUPS_T Define @code{GETGROUPS_T} to be whichever of @code{gid_t} or @code{int} is the base type of the array argument to @code{getgroups}. @end defmac @defmac AC_MODE_T @maindex MODE_T @vindex mode_t If @code{mode_t} is not defined in @file{sys/types.h}, define @code{mode_t} to be @code{int}. @end defmac @defmac AC_OFF_T @maindex OFF_T @vindex off_t If @code{off_t} is not defined in @file{sys/types.h}, define @code{off_t} to be @code{long}. @end defmac @defmac AC_PID_T @maindex PID_T @vindex pid_t If @code{pid_t} is not defined in @file{sys/types.h}, define @code{pid_t} to be @code{int}. @end defmac @defmac AC_RETSIGTYPE @maindex RETSIGTYPE @vindex RETSIGTYPE If @file{signal.h} declares @code{signal} as returning a pointer to a function returning @code{void}, define @code{RETSIGTYPE} to be @code{void}; otherwise, define it to be @code{int}. Define signal handlers as returning type @code{RETSIGTYPE}: @example @group RETSIGTYPE hup_handler () @{ @dots{} @} @end group @end example @end defmac @defmac AC_SIZE_T @maindex SIZE_T @vindex size_t If @code{size_t} is not defined in @file{sys/types.h}, define @code{size_t} to be @code{unsigned}. @end defmac @defmac AC_UID_T @maindex UID_T @vindex uid_t @vindex gid_t If @code{uid_t} is not defined in @file{sys/types.h}, define @code{uid_t} to be @code{int} and @code{gid_t} to be @code{int}. @end defmac @node Library Functions, Structures, Typedefs, Specific Tests @section Library Functions The following macros check for particular C library functions: @defmac AC_ALLOCA @maindex ALLOCA @vindex C_ALLOCA @vindex HAVE_ALLOCA_H Check how to get @code{alloca}. Tries to get a builtin version by checking for @file{alloca.h} or the predefined C preprocessor macros @code{__GNUC__} and @code{_AIX}. If that fails, it looks for a function in the standard C library. If that fails, it sets the @code{make} variable @code{ALLOCA} to @samp{alloca.o} and defines @code{C_ALLOCA} (so programs can periodically call @samp{alloca(0)} to garbage collect). This variable is separate from @code{LIBOBJS} so multiple programs can share the value of @code{ALLOCA} without needing to create an actual library, in case only some of them use the code in @code{LIBOBJS}. If this macro finds @file{alloca.h}, it defines @code{HAVE_ALLOCA_H}. This macro does not try to get @code{alloca} from the SVR3 @file{libPW} or the SVR4 @file{libucb} because those libraries contain some incompatible functions that cause trouble. Some versions do not even contain @code{alloca} or contain a buggy version. If you still want to use their @code{alloca}, use @code{ar} to extract @file{alloca.o} from them instead of compiling @file{alloca.c}. Source files that use @code{alloca} should start with a piece of code like the following, to declare it properly. Note that in some versions of AIX, the declaration of @code{alloca} must precede everything else except for comments and preprocessor directives. The @code{#pragma} directive is indented so that pre-ANSI C compilers will ignore it, rather than choke on it. @example @group /* AIX requires this to be the first thing in the file. */ #ifdef __GNUC__ #define alloca __builtin_alloca #else /* not __GNUC__ */ #if HAVE_ALLOCA_H #include #else /* not HAVE_ALLOCA_H */ #ifdef _AIX #pragma alloca #else /* not _AIX */ char *alloca (); #endif /* not _AIX */ #endif /* not HAVE_ALLOCA_H */ #endif /* not __GNUC__ */ @end group @end example @end defmac @defmac AC_GETLOADAVG @maindex GETLOADAVG @vindex SVR4 @vindex DGUX @vindex UMAX @vindex UMAX4_3 @vindex NLIST_STRUCT @vindex NLIST_NAME_UNION @vindex GETLODAVG_PRIVILEGED @vindex NEED_SETGID Check how to get the system load averages. If the system has the @code{getloadavg} function, this macro defines @code{HAVE_GETLOADAVG}, and adds to @code{LIBS} any libraries needed to get that function. Otherwise, it adds @samp{getloadavg.o} to the @code{make} variable @code{LIBOBJS}, and possibly defines several other C preprocessor macros and @code{make} variables: @enumerate @item It defines @code{SVR4}, @code{DGUX}, @code{UMAX}, or @code{UMAX4_3} if on those systems. @item If it finds @file{nlist.h}, it defines @code{NLIST_STRUCT}. @item If @samp{struct nlist} has an @samp{n_un} member, it defines @code{NLIST_NAME_UNION}. @item If compiling @file{getloadavg.c} defines @code{LDAV_PRIVILEGED}, programs need to be installed specially on this system for @code{getloadavg} to work, and this macro defines @code{GETLOADAVG_PRIVILEGED}. @item This macro always defines @code{NEED_SETGID}, for @code{make}. The value is @samp{true} if special installation is required, @samp{false} if not. If @code{NEED_SETGID} is @samp{true}, it sets @code{KMEM_GROUP} to the name of the group that should own the installed program. @end enumerate @end defmac @defmac AC_SETVBUF_REVERSED @maindex SETVBUF_REVERSED @vindex SETVBUF_REVERSED If @code{setvbuf} takes the buffering type as its second argument and the buffer pointer as the third, instead of the other way around, define @code{SETVBUF_REVERSED}. This is the case on System V before release 3. @end defmac @defmac AC_STRCOLL @maindex STRCOLL @vindex HAVE_STRCOLL If the @code{strcoll} function exists and works correctly, define @code{HAVE_STRCOLL}. This does a bit more than @samp{AC_HAVE_FUNCS(strcoll)}, because some systems have incorrect definitions of @code{strcoll}, which should not be used. @end defmac @defmac AC_UTIME_NULL @maindex UTIME_NULL @vindex HAVE_UTIME_NULL If @samp{utime(@var{file}, NULL)} sets @var{file}'s timestamp to the present, define @code{HAVE_UTIME_NULL}. @end defmac @defmac AC_VFORK @maindex VFORK @vindex HAVE_VFORK_H @vindex vfork If @file{vfork.h} is found, define @code{HAVE_VFORK_H}. If a working @code{vfork} is not found, define @code{vfork} to be @code{fork}. This macro checks for several known errors in implementations of @code{vfork} and considers the system to not have a working @code{vfork} if it detects any of them. @end defmac @defmac AC_VPRINTF @maindex VPRINTF @vindex HAVE_VPRINTF @vindex HAVE_DOPRNT If @code{vprintf} is found, define @code{HAVE_VPRINTF}. Otherwise, if @code{_doprnt} is found, define @code{HAVE_DOPRNT}. @end defmac @defmac AC_WAIT3 @maindex WAIT3 @vindex HAVE_WAIT3 If @code{wait3} is found and fills in the contents of its third argument (a @samp{struct rusage *}), which HP-UX does not do, define @code{HAVE_WAIT3}. @end defmac @node Structures, Compiler Characteristics, Library Functions, Specific Tests @section Structures The following macros check for certain structures or structure members: @defmac AC_STAT_MACROS_BROKEN @maindex STAT_MACROS_BROKEN @maindex STAT_MACROS_BROKEN If the macros @code{S_ISDIR}, @code{S_ISREG} et al. defined in @file{sys/stat.h} do not work properly (returning false positives), define @code{STAT_MACROS_BROKEN}. This is the case on Tektronix UTekV, Amdahl UTS and Motorola System V/88. @end defmac @defmac AC_ST_BLKSIZE @maindex ST_BLKSIZE @vindex HAVE_ST_BLKSIZE If @code{struct stat} contains an @code{st_blksize} member, define @code{HAVE_ST_BLKSIZE}. @end defmac @defmac AC_ST_BLOCKS @maindex ST_BLOCKS @vindex HAVE_ST_BLOCKS If @code{struct stat} contains an @code{st_blocks} member, define @code{HAVE_ST_BLOCKS}. Otherwise, add @samp{fileblocks.o} to the @code{make} variable @code{LIBOBJS}. @end defmac @defmac AC_ST_RDEV @maindex ST_RDEV @vindex HAVE_ST_RDEV If @code{struct stat} contains an @code{st_rdev} member, define @code{HAVE_ST_RDEV}. @end defmac @defmac AC_TIME_WITH_SYS_TIME @maindex TIME_WITH_SYS_TIME @vindex TIME_WITH_SYS_TIME If a program may include both @file{time.h} and @file{sys/time.h}, define @code{TIME_WITH_SYS_TIME}. On some older systems, @file{sys/time.h} includes @file{time.h}, but @file{time.h} is not protected against multiple inclusion, so programs should not explicitly include both files. This macro is useful in programs that use, for example, @code{struct timeval} or @code{struct timezone} as well as @code{struct tm}. It is best used in conjunction with @code{HAVE_SYS_TIME_H}. @example #ifdef TIME_WITH_SYS_TIME #include #include #else #ifdef HAVE_SYS_TIME_H #include #else #include #endif #endif @end example @end defmac @defmac AC_STRUCT_TM @maindex STRUCT_TM @vindex TM_IN_SYS_TIME If @file{time.h} does not define @code{struct tm}, define @code{TM_IN_SYS_TIME}, which means that including @file{sys/time.h} defines @code{struct tm}. @end defmac @defmac AC_TIMEZONE @maindex TIMEZONE @vindex HAVE_TM_ZONE @vindex HAVE_TZNAME Figure out how to get the current timezone. If @code{struct tm} has a @code{tm_zone} member, define @code{HAVE_TM_ZONE}. Otherwise, if the external array @code{tzname} is found, define @code{HAVE_TZNAME}. This macro calls @code{AC_STRUCT_TM} if it hasn't been called already. @end defmac @node Compiler Characteristics, System Services, Structures, Specific Tests @section Compiler Characteristics The following macros check for C compiler or machine architecture features: @defmac AC_ARG_ARRAY @maindex ARG_ARRAY @vindex NO_ARG_ARRAY If the address of an argument to a C function can not be used like the start of an array, define @code{NO_ARG_ARRAY}. This ability allows a sequence of arguments with the same type to be accessed as if they were an array of values. @end defmac @defmac AC_CROSS_CHECK @maindex CROSS_CHECK If the C compiler being used does not produce executables that can run on the system where @code{configure} is being run, set the shell variable @code{cross_compiling} to 1. This information can be used by @code{AC_TEST_PROGRAM} to determine whether to take a default action instead of trying to run a test program (@pxref{General Tests}). @end defmac @defmac AC_CHAR_UNSIGNED @maindex CHAR_UNSIGNED @vindex __CHAR_UNSIGNED__ If the C type @code{char} is unsigned, define @code{__CHAR_UNSIGNED__}, unless the C compiler predefines it. @end defmac @defmac AC_CONST @maindex CONST @vindex const If the C compiler does not fully support the keyword @code{const}, define @code{const} to be empty. Some C compilers that do not define @code{__STDC__} do support @code{const}; some compilers that define @code{__STDC__} do not completely support @code{const}. Programs can simply use @code{const} as if every C compiler supported it; for those that don't, the @file{Makefile} or configuration header file will define it as empty. (If using a configuration header file, the program should include it before any other header files, to prevent inconsistencies in declarations.) @end defmac @defmac AC_INLINE @maindex INLINE @vindex inline If the C compiler is a version of GCC that supports the keyword @code{__inline} but not @code{inline} (such as some NeXT versions), define @code{inline} to be @code{__inline}. This macro calls @code{AC_PROG_CC} if it hasn't been called already. @end defmac @defmac AC_INT_16_BITS @maindex INT_16_BITS @vindex INT_16_BITS If the C type @code{int} is 16 bits wide, define @code{INT_16_BITS}. @end defmac @defmac AC_LONG_64_BITS @maindex LONG_64_BITS @vindex LONG_64_BITS If the C type @code{long int} is 64 bits wide, define @code{LONG_64_BITS}. @end defmac @defmac AC_LONG_DOUBLE @maindex HAVE_LONG_DOUBLE @vindex HAVE_LONG_DOUBLE If the C compiler supports the @code{long double} type, define @code{HAVE_LONG_DOUBLE}. Some C compilers that do not define @code{__STDC__} do support the @code{long double} type; some compilers that define @code{__STDC__} do not support @code{long double}. @end defmac @defmac AC_WORDS_BIGENDIAN @maindex WORDS_BIGENDIAN @vindex WORDS_BIGENDIAN If words are stored with the most significant byte first (like Motorola and SPARC, but not Intel and VAX, CPUs), define @code{WORDS_BIGENDIAN}. @end defmac @node System Services, UNIX Variants, Compiler Characteristics, Specific Tests @section System Services The following macros check for operating system services: @defmac AC_FIND_X @maindex FIND_X Try to locate the X Window System include files and libraries. Try first by running @code{xmkmf} on a trivial @file{Imakefile} and examining the @file{Makefile} that it produces. If that fails (such as if @code{xmkmf} is not present), look for them in several directories where they often reside. If either method is successful, set the shell variables @code{x_includes} and @code{x_libraries} to their locations, unless they are in directories the compiler searches by default. If both methods fail, set the shell variable @code{no_x} to @samp{true}; otherwise set it to the empty string. @end defmac @defmac AC_HAVE_POUNDBANG (@var{action-if-supported} @r{[}, @var{action-if-unsupported}@r{]}) @maindex HAVE_POUNDBANG Print @samp{checking if `#!' works in shell scripts} to the standard output. Then check whether the system supports starting shell scripts with a line of the form @samp{#!/bin/csh} to select the shell to run them with. @var{action-if-supported} is a list of shell commands to run if @samp{#!} works; @var{action-if-not-supported} is a list of shell commands to run otherwise. @end defmac @defmac AC_LONG_FILE_NAMES @maindex LONG_FILE_NAMES @vindex HAVE_LONG_FILE_NAMES If the system supports file names longer than 14 characters, define @code{HAVE_LONG_FILE_NAMES}. @end defmac @defmac AC_REMOTE_TAPE @maindex REMOTE_TAPE @vindex HAVE_SYS_MTIO_H If BSD tape drive ioctls are available, define @code{HAVE_SYS_MTIO_H}, and if sockets are available add @code{rmt} to @code{make} variable @code{PROGS}. @end defmac @defmac AC_RESTARTABLE_SYSCALLS @maindex RESTARTABLE_SYSCALLS @vindex HAVE_RESTARTABLE_SYSCALLS If the system automatically restarts a system call that is interrupted by a signal, define @code{HAVE_RESTARTABLE_SYSCALLS}. @end defmac @node UNIX Variants, , System Services, Specific Tests @section UNIX Variants The following macros check for certain operating systems that need special treatment for some programs, due to exceptional oddities in their header files or libraries: @defmac AC_AIX @maindex AIX @vindex _ALL_SOURCE If on AIX, define @code{_ALL_SOURCE}. Allows the use of some BSD functions. Should be called before any macros that run the C compiler. @end defmac @defmac AC_DYNIX_SEQ @maindex DYNIX_SEQ If on DYNIX/ptx (Sequent UNIX), add @samp{-lseq} to @code{make} variable @code{LIBS}. Allows use of some BSD system calls and @code{getmntent}. @end defmac @defmac AC_IRIX_SUN @maindex IRIX_SUN If on IRIX (Silicon Graphics UNIX), add @samp{-lsun} to @code{make} variable @code{LIBS}. Needed to get @code{getmntent}. At sites using Yellow Pages/NIS, it is also needed to get properly working @code{gethostby*}, @code{getpw*}, @code{getgr*}, @code{getnetby*}, and so on. @c (According to Garrett.Wollman@uvm.edu.) @end defmac @defmac AC_ISC_POSIX @maindex ISC_POSIX @vindex _POSIX_SOURCE If on a POSIXized ISC UNIX, define @code{_POSIX_SOURCE} and add @samp{-posix} (for the GNU C compiler) or @samp{-Xp} (for other C compilers) to @code{make} variable @code{CC}. This allows the use of POSIX facilities. Must be called after @code{AC_PROG_CC} and before any other macros that run the C compiler. @end defmac @defmac AC_MINIX @maindex MINIX @vindex _MINIX @vindex _POSIX_SOURCE @vindex _POSIX_1_SOURCE If on Minix, define @code{_MINIX} and @code{_POSIX_SOURCE} and define @code{_POSIX_1_SOURCE} to be 2. This allows the use of POSIX facilities. Should be called before any macros that run the C compiler. @end defmac @defmac AC_SCO_INTL @maindex SCO_INTL If on SCO UNIX, add @samp{-lintl} to @code{make} variable @code{LIBS}. Used to get @code{strftime}. It must be called before checking for @code{strftime}. @end defmac @defmac AC_XENIX_DIR @maindex XENIX_DIR @vindex VOID_CLOSEDIR If on Xenix, define @code{VOID_CLOSEDIR} and add @samp{-lx} to @code{make} variable @code{LIBS}. Also, if @file{sys/ndir.h} is not being used, add @samp{-ldir} to @code{LIBS}. Needed when using the directory reading functions. This macro must be called after @code{AC_DIR_HEADER}. @end defmac @node General Purpose Macros, Writing Macros, Specific Tests, Top @chapter General Purpose Macros These macros provide ways for other macros to control the kind of output that Autoconf produces or to check whether various kinds of features are available. They all take arguments. When calling these macros, there must not be any blank space between the macro name and the open parentheses. Arguments to these macros can be more than one line long if they are enclosed within the @code{m4} quote characters @samp{[} and @samp{]}. Within each section below, the macros are listed in alphabetical order. These macros are defined in the file @file{acgeneral.m4}. @menu * Setup:: Controlling Autoconf operation. * General Tests:: Check for kinds of features. * Setting Variables:: Setting shell and @code{make} variables. * Macro Ordering:: Enforcing ordering constraints. @end menu @node Setup, General Tests, , General Purpose Macros @section Controlling Autoconf Setup The following macros control the kind of output that Autoconf produces. @defmac AC_CONFIG_HEADER (@var{header-to-create} @dots{}) @maindex CONFIG_HEADER @vindex HAVE_CONFIG_H Make @code{AC_OUTPUT} create the file(s) in the whitespace-separated list @var{header-to-create} containing C preprocessor @code{#define} statements and replace @samp{@@DEFS@@} in generated files with @samp{-DHAVE_CONFIG_H} instead of the value of @code{DEFS}. This macro should be called right after @code{AC_INIT}. The usual name for @var{header-to-create} is @file{config.h}. If @var{header-to-create} already exists and its contents are identical to what @code{AC_OUTPUT} would put in it, it is left alone. Doing this allows some changes in configuration without needlessly causing object files that depend on the header file to be recompiled. Your distribution should contain a file @file{@var{header-to-create}.in} that looks as you want the final header file to look, including comments, with default values in the @code{#define} statements. A default value can be to @code{#undef} the variable instead of to define it to a value, if your code tests for configuration options using @code{#ifdef} instead of @code{#if}. You can use the program @code{autoheader} to create @file{@var{header-to-create}.in} (@pxref{Invoking autoheader}). @end defmac @defmac AC_INIT (@var{unique-file-in-source-dir}) @maindex INIT Process the command-line arguments and find the source code directory. @var{unique-file-in-source-dir} is some file that is in the package's source directory; @code{configure} checks for this file's existence to make sure that the directory that it is told contains the source code in fact does (@pxref{Running configure Scripts}, for more information). @end defmac @defmac AC_PREPARE (@var{unique-file-in-source-dir}) @maindex PREPARE Find the source code directory and set up shell variables necessary for other Autoconf macros to work. @var{unique-file-in-source-dir} is some file that is in the package's source directory; @code{configure} checks for this file's existence to make sure that the directory that it is told contains the source code in fact does (@pxref{Running configure Scripts}, for more information). @code{AC_PREPARE} is the last thing done by @code{AC_INIT}. Use @code{AC_PREPARE} instead of @code{AC_INIT} if you want to do argument parsing yourself; never use both. @end defmac @defmac AC_OUTPUT (@r{[}@var{file}@dots{}@r{] [,}@var{extra-cmds}@r{]}) @maindex OUTPUT Create output files (typically one or more @file{Makefile}s) and @file{config.status}. If @code{AC_CONFIG_HEADER} has been called, also create the header file that was named as its argument. The argument is a whitespace-separated list of files to create; if it is omitted, no files are created. @code{AC_OUTPUT} creates each file @file{@var{file}} in the list by copying @file{@var{file}.in}, substituting the variable values that have been selected by calling @code{AC_SUBST}. It creates the directory that each file is in if it doesn't exist (but not the parents of that directory). A plausible value for the argument to @code{AC_OUTPUT} is @samp{Makefile src/Makefile man/Makefile X/Imakefile}. If you pass @var{extra-cmds}, those commands will be inserted into @file{config.status} to be run after all its other processing. @end defmac @defmac AC_REVISION (@var{revision-info}) @maindex REVISION Copy revision stamp @var{revision-info} into the @file{configure} script, with any dollar signs or double-quotes removed. This macro lets you put a revision stamp from @file{configure.in} into @file{configure} without RCS or CVS changing it when you check in @file{configure}. That way, you can determine easily which revision of @file{configure.in} a particular @file{configure} corresponds to. It is a good idea to call this macro before @code{AC_INIT} so that the revision number is near the top of both @file{configure.in} and @file{configure}. To support doing that, the @code{AC_REVISION} output begins with @samp{#!/bin/sh}, like the normal start of a @file{configure} script does. For example, this line in @file{configure.in}: @example AC_REVISION($Revision: 1.30 $)dnl @end example @noindent produces this in @file{configure}: @example #!/bin/sh # From configure.in Revision: 1.30 @end example @end defmac @node General Tests, Setting Variables, Setup, General Purpose Macros @section Checking for Kinds of Features These macros are templates that, when called with actual parameters, check for various kinds of features. Many of these macros handle two cases: what to do if the given condition is met, and what to do if the condition is not met. In some places you you might want to do something if a condition is true but do nothing if it's false, or vice versa. To omit the true case, pass an empty value for the @var{action-if-found} argument to the macro. To omit the false case, omit the @var{action-if-not-found} argument to the macro, including the comma before it. One shell programming construction that you should not use in the action arguments to these macros is @samp{@var{var}=$@{@var{var}:-@var{value}@}}. Old BSD shells, including the Ultrix @code{sh}, don't understand the colon, and complain and die. If you omit the colon, it works fine: @samp{@var{var}=$@{@var{var}-@var{value}@}}. @xref{Writing Macros}, for more information on how best to use these macros. @defmac AC_COMPILE_CHECK (@var{echo-text}, @var{includes}, @var{function-body}, @var{action-if-found} @r{[}, @var{action-if-not-found}@r{]}) @maindex COMPILE_CHECK Print @samp{checking for @var{echo-text}} to the standard output. Then create a test C program to see whether a function whose body consists of @var{function-body} can be compiled and linked; @var{includes} is any @code{#include} statements needed by the code in @var{function-body}. If the file compiles and links successfully, run shell commands @var{action-if-found}, otherwise run @var{action-if-not-found}. @end defmac @defmac AC_FUNC_CHECK (@var{function}, @var{action-if-found} @r{[}, @var{action-if-not-found}@r{]}) @maindex FUNC_CHECK If @var{function} is available, run shell commands @var{action-if-found}, otherwise @var{action-if-not-found}. If the functions might be in libraries other than the default C library, first call @code{AC_HAVE_LIBRARY} for those libraries. If you just want to define a symbol if the function is available, consider using @code{AC_HAVE_FUNCS} instead. @end defmac @defmac AC_HAVE_FUNCS (@var{function}@dots{}) @maindex HAVE_FUNCS @vindex HAVE_@var{function} For each given @var{function} in the whitespace-separated argument list that is available, define @code{HAVE_@var{function}} (in all caps). @xref{Specific Tests}, for a precise definition of ``define'' as it is used here. If the functions might be in libraries other than the default C library, first call @code{AC_HAVE_LIBRARY} for those libraries. @end defmac @defmac AC_HAVE_HEADERS (@var{header-file}@dots{}) @maindex HAVE_HEADERS @vindex HAVE_@var{header} For each given @var{header-file} in the whitespace-separated argument list that exists, define @code{HAVE_@var{header-file}} (in all caps). @xref{Specific Tests}, for a precise definition of ``define'' as it is used here. @end defmac @defmac AC_HAVE_LIBRARY (@var{library} @r{[}, @var{action-if-found} @r{[}, @var{action-if-not-found}@r{]]}) @maindex HAVE_LIBRARY Print @samp{checking for @var{library}} to the standard output. Then create a test C program to see whether that program can be linked with the specified library. @var{action-if-found} is a list of shell commands to run if the link succeeds (which means that the library is present); @var{action-if-not-found} is a list of shell commands to run if the link fails. If @var{action-if-found} and @var{action-if-not-found} are not specified, the default action is to add @samp{-lfoo} to @code{LIBS} and define @samp{HAVE_LIBfoo} for library @samp{foo}. @var{library} can be written as any of @samp{foo}, @samp{-lfoo}, or @samp{libfoo.a}. In all of those cases, the compiler is passed @samp{-lfoo}. @end defmac @defmac AC_HEADER_CHECK (@var{header-file}, @var{action-if-found} @r{[}, @var{action-if-not-found}@r{]}) @maindex HEADER_CHECK If @var{header-file} exists, execute shell commands @var{action-if-found}, otherwise execute @var{action-if-not-found}. If you just want to define a symbol if the header file is available, consider using @code{AC_HAVE_HEADERS} instead. @end defmac @defmac AC_HEADER_EGREP (@var{pattern}, @var{header-file}, @var{action-if-found} @r{[}, @var{action-if-not-found}@r{]}) @maindex HEADER_EGREP If the output of running the C preprocessor on @var{header-file} contains the @code{egrep} regular expression @var{pattern}, execute shell commands @var{action-if-found}, otherwise execute @var{action-if-not-found}. You can not check whether preprocessor symbols are defined this way, because they get expanded before @code{egrep} sees them. But you can almost always detect them by simply using @code{#ifdef} directives in your C programs. @end defmac @defmac AC_PREFIX (@var{program}) @maindex PREFIX If the user did not specify an installation prefix on the command line, guess a value for it by looking for @var{program} in @code{PATH}, the way the shell does. If @var{program} is found, set the prefix to the parent of the directory containing @var{program}; otherwise leave the prefix specified in @file{Makefile.in} unchanged. For example, if @var{program} is @code{gcc} and the @code{PATH} contains @file{/usr/local/gnu/bin/gcc}, set the prefix to @file{/usr/local/gnu}. @end defmac @defmac AC_PROGRAM_CHECK (@var{variable}, @var{prog-to-check-for}, @var{value-if-found}, @var{value-if-not-found}) @maindex PROGRAM_CHECK Check whether program @var{prog-to-check-for} exists in @code{PATH}. If it is found, set @var{variable} to @var{value-if-found}, otherwise to @var{value-if-not-found}. Calls @code{AC_SUBST} for @var{variable}. @end defmac @defmac AC_PROGRAM_EGREP (@var{pattern}, @var{program}, @var{action-if-found} @r{[}, @var{action-if-not-found}@r{]}) @maindex PROGRAM_EGREP @var{program} is the text of a C program, on which shell variable and backquote substitutions are performed. If the output of running the C preprocessor on @var{program} contains the @code{egrep} regular expression @var{pattern}, execute shell commands @var{action-if-found}, otherwise execute @var{action-if-not-found}. (It is an unfortunate oversight that we use the word @code{PROGRAM} in Autoconf macro names to sometimes mean C source code and sometimes mean a UNIX command.) @end defmac @defmac AC_PROGRAMS_CHECK (@var{variable}, @var{progs-to-check-for} @r{[}, @var{value-if-not-found}@r{]}) @maindex PROGRAMS_CHECK Check for each program in the whitespace-separated list @var{progs-to-check-for} exists in @code{PATH}. If it is found, set @var{variable} to the name of that program. Otherwise, continue checking the next program in the list. If none of the programs in the list are found, set @var{variable} to @var{value-if-not-found}; if @var{value-if-not-found} is not specified, the value of @var{variable} will not be changed. Calls @code{AC_SUBST} for @var{variable}. @end defmac @defmac AC_REPLACE_FUNCS (@var{function-name}@dots{}) @maindex REPLACE_FUNCS For each given @var{function-name} in the whitespace-separated argument list that is not in the C library, add @samp{@var{function-name}.o} to the value of the @code{make} variable @code{LIBOBJS}. If the functions might be in libraries other than the default C library, first call @code{AC_HAVE_LIBRARY} for those libraries. @end defmac @defmac AC_TEST_PROGRAM (@var{program}, @var{action-if-true} @r{[}, @var{action-if-false} @r{[}, @var{action-if-cross-compiling}@r{]]}) @maindex TEST_PROGRAM @var{program} is the text of a C program, on which shell variable and backquote substitutions are performed. If it compiles and links successfully and returns an exit status of 0 when executed, run shell commands @var{action-if-true}. Otherwise run shell commands @var{action-if-false}. If the optional argument @var{action-if-cross-compiling} is given and the C compiler being used does not produce executables that run on the system where @code{configure} is being run, then the test program is not run. Instead, the shell commands @var{action-if-cross-compiling} are run. If that argument is given, this macro calls @code{AC_CROSS_CHECK} if it has not already been called (@pxref{Compiler Characteristics}). @end defmac @defmac AC_TEST_CPP (@var{includes}, @var{action-if-true} @r{[}, @var{action-if-false}@r{]}) @maindex TEST_CPP @var{includes} is C @code{#include} statements and declarations, on which shell variable and backquote substitutions are performed. (Actually, it can be any C program, but other statements are probably not useful.) If the C preprocessor produces no error messages while processing it, run shell commands @var{action-if-true}. Otherwise run shell commands @var{action-if-false}. This macro calls @code{AC_PROG_CPP} if it hasn't been called already. @end defmac @defmac AC_WITH (@var{feature}, @var{action-if-true} @r{[}, @var{action-if-false}@r{]}) @maindex WITH If the user gave @code{configure} the option @samp{--with-@var{feature}}, run shell commands @var{action-if-true}. Otherwise run shell commands @var{action-if-false}. The name @var{feature} should consist only of alphanumeric characters and dashes; typical feature names are @samp{gnu-libc} and @samp{x}. The user can give a value by following the feature name with @samp{=} and the value; for example, @example ./configure --with-targets=sun4,hp300bsd @end example The value given is available to the shell commands @var{action-if-true} in the shell variable @code{withval}. If no value was given, @code{withval} is 1. For example, @example AC_WITH(fubar, echo "got --with-fubar=$withval", echo no --with-fubar) @end example @end defmac @node Setting Variables, Macro Ordering, General Tests, General Purpose Macros @section Setting Variables These macros help other macros to define shell and @code{make} variables. @defmac AC_DEFINE (@var{variable} @r{[}, @var{value}@r{]}) @maindex DEFINE Define C preprocessor variable @var{variable}. If @var{value} is given, set @var{variable} to that value, otherwise set it to 1. To use a shell variable as the value, use @code{AC_DEFINE_UNQUOTED} instead. Precede double quotes in the value with backslashes. This macro adds to the shell variable @code{DEFS}. @code{AC_OUTPUT} later substitutes the values in @code{DEFS} into the file(s) that it generates (typically @file{Makefile}). Alternately, if @code{AC_CONFIG_HEADER} has been called, @code{AC_OUTPUT} creates a header file by substituting the correct values into @code{#define} statements in a template file. For example, suppose your @file{configure.in} calls @code{AC_CONFIG_HEADER(conf.h)} and @code{AC_UNISTD_H}. You could have code like this in @file{conf.h.in}: @example @group /* Define as 1 if you have unistd.h. */ #define HAVE_UNISTD_H 0 @end group @end example On systems that have @file{unistd.h}, @code{configure} will change the 0 to a 1. On other systems, it will leave the line unchanged. Alternately, if you prefer to use @code{#ifdef}, your @file{conf.h.in} could have code like this: @example @group /* Define if you have unistd.h. */ #undef HAVE_UNISTD_H @end group @end example On systems that have @file{unistd.h}, @code{configure} will change the second line to read @samp{#define HAVE_UNISTD_H 1}. On other systems, it will leave the line unchanged. Due to the syntactical bizarreness of the Bourne shell, do not use semicolons to separate @code{AC_DEFINE} calls from other macro calls or shell code; that can cause syntax errors in the resulting @code{configure} script. Use either spaces or newlines. That is, do this: @example AC_HEADER_CHECK(elf.h, AC_DEFINE(SVR4) LIBS="$LIBS -lelf") @end example @noindent or this: @example AC_HEADER_CHECK(elf.h, AC_DEFINE(SVR4) LIBS="$LIBS -lelf") @end example @noindent instead of this: @example AC_HEADER_CHECK(elf.h, AC_DEFINE(SVR4); LIBS="$LIBS -lelf") @end example @end defmac @defmac AC_DEFINE_UNQUOTED (@var{variable} @r{[}, @var{value}@r{]}) @maindex DEFINE_UNQUOTED Like @code{AC_DEFINE}, but it does nothing to quote @var{value} from various shell and @code{sed} expansions it will undergo. @var{value} will be used in many different contexts requiring different quoting, and it is up to you to make sure it works right. Use this macro instead of @code{AC_DEFINE} when @var{value} contains a shell variable. For example: @example AC_DEFINE_UNQUOTED(config_machfile, $@{machfile@}) @end example @end defmac @defmac AC_SUBST (@var{variable}) @maindex SUBST Substitute the variable @var{variable} when creating the output files (typically one or more @file{Makefile}s). This means replace instances of @samp{@@@var{variable}@@}, e.g. in @file{Makefile.in}, with the current value of the shell variable @var{variable}. If this macro were not called, the value of @var{variable} would not be set in the output files, even though @code{configure} had figured out a value for it. You can set or add to the value of @var{variable} in the usual shell way. For example, to add @samp{-ltermcap} to the value of the variable @code{LIBS}: @example LIBS="$LIBS -ltermcap" @end example @end defmac @node Macro Ordering, , Setting Variables, General Purpose Macros @section Macro Ordering These macros provide ways for other macros to make sure that they are called in the correct order. @defmac AC_BEFORE (@var{this-macro-name}, @var{called-macro-name}) @maindex BEFORE Make @code{m4} print a warning message on the standard error output if @var{called-macro-name} has already been called. @var{this-macro-name} should be the name of the macro that is calling @code{AC_BEFORE}. The macro @var{called-macro-name} must contain a call to @code{AC_PROVIDE} to indicate that it has been called. This macro should be used when one macro makes changes that might affect another macro, so that the other macro should probably not be called first. For example, @code{AC_PROG_CPP} checks whether the C compiler can run the C preprocessor when given the @samp{-E} option. It should therefore be called after any macros that change which C compiler is being used, such as @code{AC_PROG_CC}. So @code{AC_PROG_CC} contains: @example AC_BEFORE([$0], [AC_PROG_CPP]) @end example @noindent This warns the user if a call to @code{AC_PROG_CPP} has already occurred when @code{AC_PROG_CC} is called. @end defmac @defmac AC_OBSOLETE (@var{this-macro-name} @r{[}, @var{suggestion}@r{]}) @maindex OBSOLETE Make @code{m4} print a message on the standard error output warning that @var{this-macro-name} is obsolete, and giving the file and line number where it was called. @var{this-macro-name} should be the name of the macro that is calling @code{AC_BEFORE}. If @var{suggestion} is given, it is printed at the end of the warning message; for example, it can be a suggestion for what to use instead of @var{this-macro-name}. A sample call is: @example AC_OBSOLETE([$0], [; use AC_HAVE_HEADERS(unistd.h) instead]) @end example @end defmac @defmac AC_PROVIDE (@var{macro-name}) @maindex PROVIDE Set a flag recording that @var{macro-name} has been called. The argument should be the name of the macro that is calling @code{AC_PROVIDE}. An easy way to get it is from the @code{m4} builtin variable @code{$0}, like this: @example AC_PROVIDE([$0]) @end example @end defmac @defmac AC_REQUIRE (@var{macro-name}) @maindex REQUIRE If the @code{m4} macro @var{macro-name} has not already been called, call it (without any arguments). Make sure to quote @var{macro-name} with square brackets. The body of @var{macro-name} must contain a call to @code{AC_PROVIDE} to indicate that it has been called. Macros that need some other macro to be called before they are called can use @code{AC_REQUIRE} to ensure that it has been, in case the person who made @file{configure.in} forgot or didn't know to do it. @code{AC_REQUIRE} and @code{AC_PROVIDE} together can ensure that a macro is only called if it is needed, and only called once. @xref{Dependencies Between Macros}, for more information. @end defmac @node Writing Macros, Makefiles, General Purpose Macros, Top @chapter Writing Macros If your package needs to test for some feature that none of the macros supplied with Autoconf handles, you'll need to write one or more new Autoconf macros. Here are some suggestions and some of the rationale behind why the existing macros are written the way they are. You can also learn a lot about how to write Autoconf macros by looking at the existing ones. If something goes wrong in one or more of the Autoconf tests, this information can help you understand why they work the way they do and the assumptions behind them, which might help you figure out how to best solve the problem. If you add macros that you think would be useful to other people, or find problems with the distributed macros, please send electronic mail to @file{bug-gnu-utils@@prep.ai.mit.edu}, so we can consider them for future releases of Autoconf. Please include the Autoconf version number, which you can get by running @samp{autoconf --version}. @menu * Macro Format:: Basic format of an Autoconf macro. * Quoting:: Protecting macros from unwanted expansion. * Dependencies Between Macros:: What to do when macros depend on other macros. * Checking for Files:: Finding whether a file exists. * Checking for Symbols:: Finding whether a symbol is defined. * Test Programs:: Writing programs to test for features. * Multiple Cases:: Tests for several possible values. @end menu @node Macro Format, Quoting, , Writing Macros @section Macro Format Autoconf macros are defined as arguments to the @code{m4} builtin command @code{define}. Their overall structure looks like this: @example define(@var{macro-name}, [@var{macro-body}])dnl @end example @noindent The square brackets here do not indicate optional text: they should literally be present in the macro definition. All of the Autoconf macros have names starting with @samp{AC_} to prevent them from accidentally conflicting with other text. You should prefix your own macro names with some other sequence, such as your initials or an abbreviation for the name of your organization or software package, to ensure that their names don't conflict with the names of present or future Autoconf macros. The @code{m4} builtin @code{dnl} prevents a newline from being inserted in the output where the macro is defined; without it, the generated @code{configure} script would begin with dozens of blank lines. @code{dnl} is also used to introduce comments in @code{m4}; it causes @code{m4} to discard the rest of the input line. You should quote the entire macro body with square brackets to avoid macro expansion problems (@pxref{Quoting}). You can refer to any arguments passed to the macro as @samp{$1}, @samp{$2}, etc. @xref{Definitions, , How to define new macros, m4.info, GNU m4}, for more complete information on writing @code{m4} macros. @node Quoting, Dependencies Between Macros, Macro Format, Writing Macros @section Quoting Macros that are called by other macros are evaluated by @code{m4} several times; each evaluation might require another layer of quotes to prevent unwanted expansions of macros or @code{m4} builtins, such as @samp{define} and @samp{$1}. Quotes are also required around macro arguments that contain commas, since commas separate the arguments from each other. Autoconf (in @file{acgeneral.m4}) changes the @code{m4} quote characters from the default @samp{`} and @samp{'} to @samp{[} and @samp{]}, because many of the macros use @samp{`} and @samp{'}, mismatched. However, in a few places the macros need to use brackets. In those places, they use the @code{m4} builtin command @code{changequote} to temporarily disable quoting before the code that uses brackets, like this: @example changequote(,)dnl @end example @noindent Then they turn quoting back on again with another call to @code{changequote}: @example changequote([,])dnl @end example When you create a @code{configure} script using newly written macros, examine it carefully to check whether you need to add more quotes in your macros. If one or more words have disappeared in the @code{m4} output, you need more quotes. When in doubt, quote. However, it's also possible to put on too many layers of quotes. If this happens, the resulting @code{configure} script will contain unexpanded macros. The @code{autoconf} program checks for this problem by doing @samp{grep AC_ configure}. @node Dependencies Between Macros, Checking for Files, Quoting, Writing Macros @section Dependencies Between Macros Some Autoconf macros depend on other macros having been called first in order to work correctly. Autoconf provides a way to ensure that certain macros are called if needed and a way to warn the user if macros are called in an order that might cause incorrect operation. @menu * Prerequisite Macros:: Ensuring required information. * Suggested Ordering:: Warning about possible ordering problems. @end menu @node Prerequisite Macros, Suggested Ordering, , Dependencies Between Macros @subsection Prerequisite Macros A macro that you write might need to use values that have previously been computed by other macros. For example, if you write a new macro that uses the C preprocessor, it depends on @code{AC_PROG_CPP} having been called first to set the shell variable @code{CPP} (@pxref{Alternative Programs}). Rather than forcing the user of the macros to keep track of all of the dependencies between them, you can use the macros @code{AC_PROVIDE} and @code{AC_REQUIRE} to do it automatically. @xref{Macro Ordering}, for more information on their syntax. The new macro that runs the C preprocessor should contain, somewhere before @code{CPP} is used, the statement @example AC_REQUIRE([AC_PROG_CPP]) @end example @noindent and the macro @code{AC_PROG_CPP} should contain the statement (anywhere in its body) @example AC_PROVIDE([$0]) @end example @noindent Then, when the new macro is run, it will invoke @code{AC_PROG_CPP} if and only if @code{AC_PROG_CPP} has not already been run. @node Suggested Ordering, , Prerequisite Macros, Dependencies Between Macros @subsection Suggested Ordering Some macros should be run before another macro if both are called, but neither requires the other to be called. For example, a macro like @code{AC_AIX} that changes the behavior of the C compiler (@pxref{UNIX Variants}) should be called before any macros that run the C compiler. Many of these dependencies are noted in the documentation. Autoconf provides a way to warn users when macros with this kind of dependency appear out of order in a @file{configure.in} file. The warning occurs when creating @file{configure} from @file{configure.in}, not when running @file{configure}. It is not a fatal error; @file{configure} is created as usual. The @code{AC_BEFORE} macro causes @code{m4} to print a warning message on the standard error output when a macro is used before another macro which might change its behavior. The macro which should come first should contain a call to @code{AC_BEFORE} and the macro which should come later should contain a call to @code{AC_PROVIDE}. For example, @code{AC_AIX} contains @example AC_BEFORE([$0], [AC_COMPILE_CHECK]) @end example @noindent and @code{AC_COMPILE_CHECK} contains @example AC_PROVIDE([$0]) @end example @noindent As a result, if @code{AC_AIX} is called after @code{AC_COMPILE_CHECK}, it will note that @code{AC_COMPILE_CHECK} has already been called and print a warning message. @node Checking for Files, Checking for Symbols, Dependencies Between Macros, Writing Macros @section Checking for Files If you need to check whether a file other than a C header file exists, use @samp{test -f @var{filename}}. If you need to make multiple checks using @code{test}, combine them with the shell operators @samp{&&} and @samp{||} instead of using the @code{test} operators @samp{-a} and @samp{-o}. On System V, the precedence of @samp{-a} and @samp{-o} is wrong relative to the unary operators; consequently, POSIX does not specify them, so using them is nonportable. If you combine @samp{&&} and @samp{||} in the same statement, keep in mind that they have equal precedence. Do not use @samp{test -x}, because 4.3BSD does not have it. Use @samp{test -f} or @samp{test -r} instead. @node Checking for Symbols, Test Programs, Checking for Files, Writing Macros @section Checking for Symbols If you need to check whether a symbol is defined in a C header file, you can use @code{AC_HEADER_EGREP} if the symbol is not a C preprocessor macro (@pxref{General Tests}), or compile a small test program that includes the file and references the symbol (@pxref{Test Programs}). Don't directly @code{grep} for the symbol in the file, because on some systems it might be defined in another header file that the file you are checking @samp{#include}s. However, if you need to check for a particular UNIX variant which is distinguished by having certain text in a certain file, then use @code{grep} (or @code{egrep}). But don't use @samp{grep -s} to suppress output, because @samp{grep -s} on System V does not suppress output, only error messages. Instead, redirect the standard output and standard error (in case the file doesn't exist) of @code{grep} to @file{/dev/null}. Check the exit status of @code{grep} to determine whether it found a match. To check whether the Autoconf macros have already defined a certain C preprocessor symbol, you can use a @code{case} statement like this: @example case "$DEFS" in *HAVE_FOO*) ;; *) LIBOBJS="$LIBOBJS foo.o" ;; esac @end example @noindent Make sure to enclose the variable name you are checking (usually @code{DEFS}) in double quotes, because otherwise some old versions of @code{bash} misinterpret the statement. @node Test Programs, Multiple Cases, Checking for Symbols, Writing Macros @section Test Programs Autoconf checks for many features by compiling small test programs. To find out whether a library function is available, Autoconf tries to compile a small program that uses it. This is unlike Larry Wall's Metaconfig, which uses @code{nm} or @code{ar} on the C library to try to figure out which functions are available. Trying to link with the function is usually a more reliable and flexible approach because it avoids dealing with the variations in the options and output formats of @code{nm} and @code{ar} and in the location of the standard libraries. It also allows @code{configure} to check aspects of the function's runtime behavior if needed. On the other hand, it is sometimes slower than scanning the libraries. If you need to check for a condition other than whether some symbol exists on the system or has a certain value, then you can't use @code{AC_COMPILE_CHECK} (@pxref{General Tests}). You have to write a test program by hand. You can compile and run it using @code{AC_TEST_PROGRAM} (@pxref{General Tests}). Try to avoid writing test programs if possible, because using them prevents people from configuring your package for cross-compiling. If it's really best that you test for a run-time behavior, try to provide a default ``worst case'' value to use when cross-compiling makes run-time tests impossible. You do this by passing the optional last argument to @code{AC_TEST_PROGRAM}. @menu * Guidelines:: General rules for writing test programs. * Tricks:: Special ways to work around problems. @end menu @node Guidelines, Tricks, , Test Programs @subsection Guidelines for Test Programs Test programs should return 0 if the test succeeds, nonzero otherwise, so that success can be distinguished easily from a core dump or other failure; segmentation violations and other failures produce a nonzero exit status. Test programs should @code{exit}, not @code{return}, from @code{main}, because on some systems the argument to @code{return} in @code{main} is ignored. They should not write anything to the standard output. Test programs can use @code{#if} or @code{#ifdef} to check the values of preprocessor macros defined by tests that have already run. For example, if you call @code{AC_STDC_HEADERS}, then later on in @file{configure.in} you can have a test program that includes an ANSI C header file conditionally: @example @group #if STDC_HEADERS #include #endif @end group @end example If a test program needs to use or create a data file, give it a name that starts with @file{conftest}, such as @file{conftestdata}. The @code{configure} script cleans up by running @samp{rm -rf conftest*} after running test programs and if the script is interrupted. @node Tricks, , Guidelines, Test Programs @subsection Tricks for Test Programs If a test program calls a function with invalid parameters (just to see whether it exists), organize the program to ensure that it never invokes that function. You can do this by calling it in another function that is never invoked. You can't do it by putting it after a call to @code{exit}, because GCC version 2 knows that @code{exit} never returns and optimizes out any code that follows it in the same block. If you include any header files, make sure to call the functions relevant to them with the correct number of arguments, even if they are just 0, to avoid compilation errors due to prototypes. GCC version 2 has internal prototypes for several functions that it automatically inlines; for example, @code{memcpy}. To avoid errors when checking for them, either pass them the correct number of arguments or redeclare them with a different return type (such as @code{char}). @node Multiple Cases, , Test Programs, Writing Macros @section Multiple Cases Some operations are accomplished in several possible ways, depending on the UNIX variant. Checking for them essentially requires a ``case statement''. Autoconf does not directly provide one; however, it is easy to simulate by using a shell variable to keep track of whether a way to perform the operation has been found yet. Here is an example excerpted from the @file{configure.in} for GNU @code{find}. It uses the shell variable @code{fstype} to keep track of whether the remaining cases need to be checked. There are several more cases which are not shown here but follow the same pattern. @example @group echo checking how to get filesystem type # SVR4. AC_TEST_CPP([#include #include ], AC_DEFINE(FSTYPE_STATVFS) fstype=1) if test -z "$fstype"; then # SVR3. AC_TEST_CPP([#include #include ], AC_DEFINE(FSTYPE_USG_STATFS) fstype=1) fi if test -z "$fstype"; then # AIX. AC_TEST_CPP([#include #include ], AC_DEFINE(FSTYPE_AIX_STATFS) fstype=1) fi @end group @end example @node Makefiles, Running configure Scripts, Writing Macros, Top @chapter Makefiles Each subdirectory in a distribution should come with a file @file{Makefile.in}, from which @code{configure} will produce a @file{Makefile} in that directory. Most of the substitutions that @code{configure} does are simple: for each configuration variable that the package uses, it just replaces occurrences of @samp{@@@var{variable}@@} with the value that @code{configure} has determined for that variable. Any occurrences of @samp{@@@var{variable}@@} for variables that @code{configure} does not know about are passed through unchanged. There is no point in checking for the correct value to give a variable that is never used. Every variable that the @code{configure} script might set a value for should appear in a @samp{@@@var{variable}@@} reference in at least one @file{Makefile.in}. If @code{AC_CONFIG_HEADER} is called, @code{configure} replaces @samp{@@DEFS@@} with @samp{-DHAVE_CONFIG_H}, since the contents of @code{DEFS} would be redundant. @xref{Makefiles, , Makefile Conventions, standards.info, The GNU Coding Standards}, for more information on what to put in Makefiles. @xref{Sample Makefile.in}, for an example of a real @file{Makefile.in}. @menu * Predefined Variables:: Heavily used @code{make} variables. * Installation Prefixes:: A special variable substitution. * VPATH Substitutions:: Compiling in a different directory. * Automatic Remaking:: Makefile rules for configuring. @end menu @node Predefined Variables, Installation Prefixes, , Makefiles @section Predefined Variables Some @code{make} variables are predefined by the Autoconf macros. @code{AC_SUBST} is called for them automatically (@pxref{Setting Variables}), so in your @file{Makefile.in} files you can get their values by enclosing their names in @samp{@@} characters. @ifinfo @xref{Makefiles}, for more information on @samp{@@} substitutions. @end ifinfo The variables that are defined by the general purpose Autoconf macros are: @defvar srcdir The directory that contains the source code for that @file{Makefile}. @end defvar @defvar DEFS @samp{-D} options to pass to the C compiler. If @code{AC_CONFIG_HEADER} is called, @code{configure} replaces @samp{@@DEFS@@} with @samp{-DHAVE_CONFIG_H}, since the contents of @code{DEFS} would be redundant. @end defvar @defvar LIBS @samp{-l} and @samp{-L} options to pass to the linker. @end defvar @defvar LIBOBJS Names of object files (ending in @file{.o}). Set by @code{AC_REPLACE_FUNCS} (@pxref{General Tests}). @end defvar @node Installation Prefixes, VPATH Substitutions, Predefined Variables, Makefiles @section Installation Prefixes If @code{configure} has figured out a value for the installation prefix, either by the user supplying one on the command line (@pxref{Running configure Scripts}) or with @code{AC_PREFIX}, then it substitutes that value in @file{Makefile}s that it creates. Wherever a @file{Makefile.in} contains lines like @example prefix = /usr/local exec_prefix = $@{prefix@} @end example @noindent @code{configure} substitutes the value it figured out. The word @samp{prefix} or @samp{exec_prefix} must not be preceded by any other characters on the line. If @code{configure} has not figured out a value for the prefix, the value in the @file{Makefile.in} is left alone. There can be separate installation prefixes for architecture-specific files (@code{exec_prefix}) and architecture-independent files (@code{prefix}). @xref{Running configure Scripts}, for more information on setting them. Autoconf @code{configure} scripts replace these two variables without requiring them to be enclosed in @samp{@@} characters, and only if they have been set, because the Cygnus @code{configure} does so. In retrospect, being compatible in this way was a bad decision, because it created an inconsistency in Autoconf without giving significant benefits. @node VPATH Substitutions, Automatic Remaking, Installation Prefixes, Makefiles @section @code{VPATH} Substitutions You might want to compile a software package in a different directory from the one that contains the source code. Doing this allows you to compile the package for several architectures simultaneously from the same copy of the source code and keep multiple sets of object files on disk. To support doing this, @code{make} uses the @code{VPATH} variable to find the files that are in the source directory. GNU @code{make} and most other recent @code{make} programs can do this. Older @code{make} programs do not support @code{VPATH}; when using them, the source code must be in the same directory as the object files. To support @code{VPATH}, each @file{Makefile.in} should contain two lines that look like: @example srcdir = @@srcdir@@ VPATH = @@srcdir@@ @end example Do not set @code{VPATH} to the value of another variable, for example @samp{VPATH = $(srcdir)}, because some versions of @code{make} do not do variable substitutions on the value of @code{VPATH}. @code{configure} substitutes in the correct value for @code{srcdir} when it produces @file{Makefile.in}. Do not use the @code{make} variable @code{$<}, which expands to the pathname of the file in the source directory (found with @code{VPATH}), except in implicit rules. (An implicit rule is one such as @samp{.c.o}, which tells how to create a @file{.o} file from a @file{.c} file.) Some versions of @code{make} do not set @code{$<} in explicit rules; they expand it to an empty value. Instead, @file{Makefile} command lines should always refer to source files by prefixing them with @samp{$(srcdir)/}. For example: @example time.info: time.texinfo $(MAKEINFO) $(srcdir)/time.texinfo @end example @node Automatic Remaking, , VPATH Substitutions, Makefiles @section Automatic Remaking You can put rules like the following in the top-level @file{Makefile.in} for a package to automatically update the configuration information when you change the configuration files. @example # The next rule also takes care of making config.h from config.h.in. # If remaking config.h does not change it, its timestamp is untouched. Makefile: Makefile.in config.status $(SHELL) config.status config.status: configure $(SHELL) config.status --recheck configure: configure.in cd $(srcdir); autoconf config.h.in: configure.in cd $(srcdir); autoheader @end example @node Running configure Scripts, Example, Makefiles, Top @chapter Running @code{configure} Scripts A software package that uses a @code{configure} script should be distributed with a file @file{Makefile.in}, but no @file{Makefile}; that way, the user has to properly configure the package for the local system before compiling it. Here is how to configure a package that uses a @code{configure} script. Normally, you just @code{cd} to the directory containing the package's source code and type @samp{./configure}. If you're using @code{csh} on an old version of System V, you might need to type @samp{sh configure} instead to prevent @code{csh} from trying to execute @code{configure} itself. Running @code{configure} takes a minute or two. While it is running, it prints some messages that tell what it is doing. If you don't want to see the messages, run @code{configure} with its standard output redirected to @file{/dev/null}; for example, @samp{./configure >/dev/null}. To compile the package in a different directory from the one containing the source code, you must use a version of @code{make} that supports the @code{VPATH} variable, such as GNU @code{make}. @code{cd} to the directory where you want the object files and executables to go and run the @code{configure} script. @code{configure} automatically checks for the source code in the directory that @code{configure} is in and in @file{..}. If for some reason @code{configure} is not in the source code directory that you are configuring, then it will report that it can't find the source code. In that case, run @code{configure} with the option @samp{--srcdir=@var{dir}}, where @var{dir} is the directory that contains the source code. By default, @samp{make install} will install the package's files in @file{/usr/local/bin}, @file{/usr/local/man}, etc. You can specify an installation prefix other than @file{/usr/local} by giving @code{configure} the option @samp{--prefix=@var{path}}. Alternately, you can do so by consistently giving a value for the @samp{prefix} variable when you run @code{make}, e.g., @example make prefix=/usr/gnu make prefix=/usr/gnu install @end example You can specify separate installation prefixes for architecture-specific files and architecture-independent files. If you give @code{configure} the option @samp{--exec-prefix=@var{path}} or set the @code{make} variable @samp{exec_prefix} to @var{path}, the package will use @var{path} as the prefix for installing programs and libraries. Data files and documentation will still use the regular prefix. Normally, all files are installed using the same prefix. Some packages pay attention to @samp{--with-@var{package}} options to @code{configure}, where @var{package} is something like @samp{gnu-as} or @samp{x} (for the X Window System). The README should mention any @samp{--with-} options that the package recognizes. @code{configure} ignores any other arguments that you give it. @menu * Overriding variables:: Workarounds for unusual systems. * Running config.status:: Recreating a configuration. @end menu @node Overriding variables, Running config.status, , Running configure Scripts @section Overriding variables On systems that require unusual options for compilation or linking that the package's @code{configure} script does not know about, you can give @code{configure} initial values for variables by setting them in the environment. In Bourne-compatible shells, you can do that on the command line like this: @example CC='gcc -traditional' LIBS=-lposix ./configure @end example Here are the @code{make} variables that you might want to override with environment variables when running @code{configure}. For these variables, any value given in the environment overrides the value that @code{configure} would choose: @defvar CC C compiler program. The default is @code{cc}. @end defvar @defvar INSTALL Program to use to install files. The default is @code{install} if you have it, @code{cp} otherwise. @end defvar For these variables, any value given in the environment is added to the value that @code{configure} chooses: @defvar DEFS Configuration options, in the form @samp{-Dfoo -Dbar@dots{}}. Do not use this variable in packages that create a configuration header file. @end defvar @defvar LIBS Libraries to link with, in the form @samp{-lfoo -lbar@dots{}}. @end defvar In the long term, most problems requiring manual intervention should be fixed by updating either the Autoconf macros or the @file{configure.in} file for that package. @xref{Making configure Scripts}, for a discussion of that subject. @node Running config.status, , Overriding variables, Running configure Scripts @section Recreating a Configuration The @code{configure} script creates a file named @file{config.status} which describes which configuration options were specified when the package was last configured. This file is a shell script which, if run, will recreate the same configuration. You can give @file{config.status} the @samp{--recheck} option, which makes it re-run @code{configure} with the same arguments you used before. This option is useful if you change @code{configure}, so that the results of some tests might be different from the previous run. @file{config.status} checks several optional environment variables that can alter its behavior: @defvar CONFIG_SHELL The shell with which to run @code{configure} for the @samp{--recheck} option. The default is @file{/bin/sh}. @end defvar The following two variables provide one way for separately distributed packages to share the values computed by @code{configure}. Doing so can be useful if some of the packages need a superset of the features that one of them, perhaps a common library, does. These variables allow a @file{config.status} file to create files other than the ones that its @file{configure.in} specifies, so it can be used for a different package. @defvar CONFIG_FILES The files in which to perform @samp{@@@var{variable}@@} substitutions. The default is the arguments given to @code{AC_OUTPUT} in @file{configure.in}. @end defvar @defvar CONFIG_HEADERS The files in which to substitute C @code{#define} statements. The default is the arguments given to @code{AC_CONFIG_HEADER}; if that macro was not called, @file{config.status} ignores this variable. @end defvar @node Example, Preprocessor Symbol Index, Running configure Scripts, Top @chapter An Example Here are sample @file{configure.in} and @file{Makefile.in} files, to give a real illustration of using Autoconf. They are from the GNU @code{cpio} package, which also includes the @code{mt} and @code{rmt} programs. This package does not use a configuration header file; it passes @samp{-D} options to the C compiler on the command line. @menu * Sample configure.in:: An example of a @file{configure} template. * Sample Makefile.in:: An example of a @file{Makefile} template. @end menu @node Sample configure.in, Sample Makefile.in, , Example @section Sample @file{configure.in} @noindent Here is @file{configure.in} from GNU @code{cpio}. The @code{dnl} macro after @code{AC_SUBST} is suppresses an extra (though harmless) newline in the generated @code{configure} script (because the @code{AC_SUBST} macro does not produce any output where it is called). @example dnl Process this file with autoconf to produce a configure script. AC_INIT(cpio.h) PROGS="cpio" AC_SUBST(PROGS)dnl AC_PROG_CC AC_PROG_CPP AC_GCC_TRADITIONAL AC_PROG_INSTALL AC_AIX AC_MINIX AC_ISC_POSIX AC_RETSIGTYPE AC_MAJOR_HEADER AC_REMOTE_TAPE test -n "$have_mtio" && PROGS="$PROGS mt" AC_RSH AC_CONST AC_UID_T AC_STDC_HEADERS AC_UNISTD_H AC_HAVE_HEADERS(string.h fcntl.h utime.h sys/io/trioctl.h) AC_REPLACE_FUNCS(fnmatch bcopy mkdir strdup) AC_HAVE_FUNCS(strerror lchown) AC_VPRINTF AC_ALLOCA AC_XENIX_DIR AC_HAVE_LIBRARY(socket, [LIBS="$LIBS -lsocket"]) AC_HAVE_LIBRARY(nsl, [LIBS="$LIBS -lnsl"]) AC_OUTPUT(Makefile) @end example @node Sample Makefile.in, , Sample configure.in, Example @section Sample @file{Makefile.in} @noindent Here is @file{Makefile.in} from GNU @code{cpio}, with some irrelevant lines omitted, for brevity. @example srcdir = @@srcdir@@ VPATH = @@srcdir@@ CC = @@CC@@ INSTALL = @@INSTALL@@ INSTALL_PROGRAM = @@INSTALL_PROGRAM@@ INSTALL_DATA = @@INSTALL_DATA@@ DEFS = @@DEFS@@ LIBS = @@LIBS@@ RTAPELIB = @@RTAPELIB@@ CFLAGS = -g LDFLAGS = -g prefix = /usr/local exec_prefix = $(prefix) binprefix = manprefix = bindir = $(exec_prefix)/bin libdir = $(exec_prefix)/lib mandir = $(prefix)/man/man1 manext = 1 SHELL = /bin/sh SRCS = copyin.c copyout.c copypass.c defer.c dstring.c global.c \ main.c tar.c util.c error.c getopt.c getopt1.c filemode.c version.c \ rtapelib.c dirname.c idcache.c makepath.c xmalloc.c stripslash.c \ userspec.c xstrdup.c bcopy.c fnmatch.c mkdir.c strdup.c OBJS = copyin.o copyout.o copypass.o defer.o dstring.o global.o \ main.o tar.o util.o error.o getopt.o getopt1.o filemode.o version.o \ $(RTAPELIB) dirname.o idcache.o makepath.o xmalloc.o stripslash.o \ userspec.o xstrdup.o @@LIBOBJS@@ @@ALLOCA@@ # mt source files not shared with cpio. MT_SRCS = mt.c argmatch.c MT_OBJS = mt.o argmatch.o error.o getopt.o getopt1.o \ xmalloc.o version.o $(RTAPELIB) @@ALLOCA@@ HDRS = cpio.h cpiohdr.h tar.h tarhdr.h defer.h dstring.h extern.h filetypes.h \ system.h fnmatch.h getopt.h rmt.h DISTFILES = $(SRCS) $(HDRS) COPYING COPYING.LIB ChangeLog Makefile.in \ README NEWS INSTALL cpio.1 mt.1 makefile.pc makefile.os2 cpio.def \ configure configure.in mkinstalldirs $(MT_SRCS) rmt.c tcexparg.c alloca.c all: @@PROGS@@ .c.o: $(CC) -c $(CPPFLAGS) $(DEFS) -I$(srcdir) $(CFLAGS) $< install: installdirs all $(srcdir)/cpio.1 $(srcdir)/mt.1 $(INSTALL_PROGRAM) cpio $(bindir)/$(binprefix)cpio test ! -f mt || $(INSTALL_PROGRAM) mt $(bindir)/$(binprefix)mt -test ! -f rmt || $(INSTALL_PROGRAM) rmt $(libdir)/rmt $(INSTALL_DATA) $(srcdir)/cpio.1 $(mandir)/$(manprefix)cpio.$(manext) test ! -f mt || \ $(INSTALL_DATA) $(srcdir)/mt.1 $(mandir)/$(manprefix)mt.$(manext) installdirs: $(srcdir)/mkinstalldirs $(bindir) $(libdir) $(mandir) uninstall: cd $(bindir); rm -f $(binprefix)cpio $(binprefix)mt -rm -f $(libdir)/rmt cd $(mandir); rm -f $(manprefix)cpio.$(manext) $(manprefix)mt.$(manext) check: @@echo No tests are supplied. cpio: $(OBJS) $(CC) $(LDFLAGS) -o $@@ $(OBJS) $(LIBS) rmt: rmt.o $(CC) $(LDFLAGS) -o $@@ rmt.o $(LIBS) mt: $(MT_OBJS) $(CC) $(LDFLAGS) -o $@@ $(MT_OBJS) $(LIBS) Makefile: Makefile.in config.status $(SHELL) config.status config.status: configure $(SHELL) config.status --recheck configure: configure.in cd $(srcdir); autoconf TAGS: $(SRCS) etags $(SRCS) clean: rm -f cpio rmt mt *.o core mostlyclean: clean distclean: clean rm -f Makefile config.status realclean: distclean rm -f TAGS dist: $(DISTFILES) echo cpio-`sed -e '/version_string/!d' \ -e 's/[^0-9.]*\([0-9.]*\).*/\1/' -e q version.c` > .fname rm -rf `cat .fname` mkdir `cat .fname` -ln $(DISTFILES) `cat .fname` for file in $(DISTFILES); do \ test -r `cat .fname`/$$file || cp -p $$file `cat .fname`; \ done tar chzf `cat .fname`.tar.gz `cat .fname` rm -rf `cat .fname` .fname @end example @node Preprocessor Symbol Index, Macro Index, Example, Top @unnumbered Preprocessor Symbol Index This is an alphabetical list of the C preprocessor symbols that the Autoconf macros define. To work with Autoconf, C source code needs to use these names in @code{#if} directives. @printindex vr @node Macro Index, , Preprocessor Symbol Index, Top @unnumbered Macro Index This is an alphabetical list of the Autoconf macros. To make the list easier to use, the macros are listed without their preceding @samp{AC_}. @printindex ma @contents @bye @c Local variables: @c compile-command: "makeinfo --no-split autoconf.texi" @c End: autoconf-1.7/autoconf.info100666 21570 1 315672 5463274725 14240 0ustar djmdaemonThis is Info file autoconf.info, produced by Makeinfo-1.54 from the input file ./autoconf.texi. This file documents the GNU Autoconf package for creating scripts to configure source code packages using templates and an `m4' macro package. Copyright (C) 1992, 1993 Free Software Foundation, Inc. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Foundation.  File: autoconf.info, Node: Top, Next: Introduction, Up: (dir) This file documents the GNU Autoconf package for creating scripts to configure source code packages using templates and an `m4' macro package. This is edition 1.7, for Autoconf version 1.7. * Menu: * Introduction:: Autoconf's purpose, strengths, and weaknesses. * Distributing:: Legal restrictions on Autoconf output. * Making configure Scripts:: How to organize and produce Autoconf scripts. * Specific Tests:: Macros that check for particular features. * General Purpose Macros:: Macros that check for kinds of features. * Writing Macros:: How to add your own macros to Autoconf. * Makefiles:: Information Autoconf uses in `Makefile's. * Running configure Scripts:: How to use the Autoconf output. * Example:: Sample Autoconf input files. * Preprocessor Symbol Index:: Index of C preprocessor symbols defined. * Macro Index:: Index of Autoconf macros. -- The Detailed Node Listing -- Making `configure' Scripts * Writing configure.in:: What to put in an Autoconf input file. * Invoking autoconf:: How to create configuration scripts. * Invoking autoheader:: How to create configuration header files. Specific Tests * Alternative Programs:: Selecting between alternative programs. * Header Files:: Header files that might be missing. * Typedefs:: `typedef's that might be missing. * Library Functions:: C library functions that might be missing. * Structures:: Structures or members that might be missing. * Compiler Characteristics:: C compiler or machine architecture features. * System Services:: Operating system services. * UNIX Variants:: Special cases for specific UNIX variants. General Purpose Macros * Setup:: Controlling Autoconf operation. * General Tests:: Check for kinds of features. * Setting Variables:: Setting shell and `make' variables. * Macro Ordering:: Enforcing ordering constraints. Writing Macros * Macro Format:: Basic format of an Autoconf macro. * Quoting:: Protecting macros from unwanted expansion. * Dependencies Between Macros:: What to do when macros depend on other macros. * Checking for Files:: Finding whether a file exists. * Checking for Symbols:: Finding whether a symbol is defined. * Test Programs:: Writing programs to test for features. * Multiple Cases:: Tests for several possible values. Dependencies Between Macros * Prerequisite Macros:: Ensuring required information. * Suggested Ordering:: Warning about possible ordering problems. Test Programs * Guidelines:: General rules for writing test programs. * Tricks:: Special ways to work around problems. Makefiles * Predefined Variables:: Heavily used `make' variables. * Installation Prefixes:: A special variable substitution. * VPATH Substitutions:: Compiling in a different directory. * Automatic Remaking:: Makefile rules for configuring. Running `configure' Scripts * Overriding variables:: Workarounds for unusual systems. * Running config.status:: Recreating a configuration. An Example * Sample configure.in:: An example of a `configure' template. * Sample Makefile.in:: An example of a `Makefile' template.  File: autoconf.info, Node: Introduction, Next: Distributing, Prev: Top, Up: Top Introduction ************ Autoconf is a tool for producing shell scripts that automatically configure software source code packages to adapt to many kinds of UNIX-like systems. The configuration scripts produced by Autoconf are independent of Autoconf when they are run, so their users do not need to have Autoconf. The configuration scripts produced by Autoconf normally require no manual user intervention when run; they do not even take an argument specifying the system type. Instead, they test for the presence of each feature that the software package they are for might need individually. (Before each check, they print a one-line message stating what they are checking for, so the user doesn't get too bored while waiting for the script to finish.) As a result, they deal well with systems that are hybrids or customized from the more common UNIX variants. There is no need to maintain files that list the features supported by each release of each variant of UNIX. For each software package that Autoconf is used with, it creates a configuration script from a template file that lists the operating system features that the package can use. After the shell code to recognize and respond to an operating system feature has been written, Autoconf allows it to be shared by many software packages that can use (or need) that feature. If it later turns out that the shell code needs adjustment for some reason, it needs to be changed in only one place; all of the the configuration scripts can be regenerated automatically to take advantage of the updated code. Larry Wall's Metaconfig package is similar in purpose to Autoconf, but is more general. The scripts it produces require manual user intervention, which is quite inconvenient when configuring large source trees. Unlike Metaconfig scripts, Autoconf scripts can support cross-compiling, if some care is taken in writing them. They should avoid executing test programs, since test programs compiled with a cross-compiler can not be executed on the host system. Also, they shouldn't do anything that tests features of the host system instead of the target system. Autoconf imposes some restrictions on the names of macros used with `#ifdef' in C programs (*note Preprocessor Symbol Index::.). Autoconf requires GNU `m4' in order to generate the scripts. It uses features that some UNIX versions of `m4' do not have; it is also reported to overflow internal limits of some versions of `m4'. Autoconf does not work well with GNU C library releases before 1.06. The GNU C library contains stubs (which always return an error) for functions that are not available instead of omitting them from the library. As a result, Autoconf scripts are fooled into thinking that those functions are available. This problem does not exist with releases 1.06 and later of the GNU C library, which define C preprocessor macros that the Autoconf macros `AC_FUNC_CHECK' and `AC_REPLACE_FUNCS' test, indicating that certain functions are stubs (*note General Tests::., for more information on checking for functions). Autoconf was written by David MacKenzie, with help from Franc,ois Pinard, Karl Berry, Richard Pixley, Ian Lance Taylor, Roland McGrath, and Noah Friedman. It was inspired by Brian Fox's automatic configuration system for Bash, by Larry Wall's Metaconfig, and by Richard Stallman, Richard Pixley, and John Gilmore's configuration tools for the GNU compiler and object file utilities. Mail suggestions and bug reports for Autoconf to `bug-gnu-utils@prep.ai.mit.edu'. Please include the Autoconf version number, which you can get by running `autoconf --version'.  File: autoconf.info, Node: Distributing, Next: Making configure Scripts, Prev: Introduction, Up: Top Distributing Autoconf Output **************************** The configuration scripts that Autoconf produces are covered by the GNU General Public License. This is because they consist almost entirely of parts of Autoconf itself, rearranged somewhat, and Autoconf is distributed under the terms of the GPL. As applied to Autoconf, the GPL just means that you need to distribute `configure.in' along with `configure'. Programs that use Autoconf scripts to configure themselves do not automatically come under the GPL. Distributing an Autoconf configuration script as part of a program is considered to be *mere aggregation* of that work with the Autoconf script. Such programs are not derivative works based on Autoconf; only their configuration scripts are. We still encourage software authors to distribute their work under terms like those of the GPL, but doing so is not required to use Autoconf.  File: autoconf.info, Node: Making configure Scripts, Next: Specific Tests, Prev: Distributing, Up: Top Making `configure' Scripts ************************** The configuration scripts that Autoconf produces are by convention called `configure' when they are distributed. When run, they create several files: * one or more `Makefile' files (one in each subdirectory of the package), from template `Makefile.in' files (*note Makefiles::.); * optionally, a C header file, the name of which is configurable, containing `#define' statements (*note Setup::.); * a shell script called `config.status' that, when run, will recreate the current configuration parameter settings (*note Running config.status::.). To create a `configure' script with Autoconf, you need to write an Autoconf input file and run Autoconf on it to produce the script. And, of course, test the resulting script. Here is a diagram showing how the files that can be used in configuration are produced: acgeneral.m4 \ Makefile.in \ acspecific.m4 \ \ autoconf* -> m4* -> configure* -> config.status* -> Makefile \ configure.in / | | \ | | | | make* -> your | | | | /package | | config.status* -> config.h / configure.in \ / autoheader* -> - - - - - - - - - - -> config.h.in / acconfig.h / Executables are suffixed by `*', while files appearing twice are linked with lines of `|'. * Menu: * Writing configure.in:: What to put in an Autoconf input file. * Invoking autoconf:: How to create configuration scripts. * Invoking autoheader:: How to create configuration header files.  File: autoconf.info, Node: Writing configure.in, Next: Invoking autoconf, Up: Making configure Scripts Writing `configure.in' ====================== To produce a `configure' script for a software package, create a file called `configure.in' that contains invocations of the Autoconf macros that test the system features your package needs or can use. Autoconf macros already exist to check for many features; see *Note Specific Tests::, for their descriptions. For most other features, you can use Autoconf template macros to produce custom checks; see *Note General Tests::, for information about them. For especially tricky or specialized features, `configure.in' might need to contain some hand-crafted shell commands. *Note Writing Macros::, for guidelines on writing tests from scratch. Every `configure.in' must begin with a call to `AC_INIT' and end with a call to `AC_OUTPUT' (*note Setup::.). Other than that, the order in which `configure.in' calls the Autoconf macros is generally not important, except that some macros rely on other macros having been called first, because they check previously set values of some variables to decide what to do. These macros are noted in the individual descriptions (*note Specific Tests::.). To encourage consistency, here is a suggested order for calling the Autoconf macros. A few macros need to be called in a different order from the one given here; they are noted in their individual descriptions (*note Specific Tests::.). Note that there must not be any space between the macro name and the open parentheses. `AC_INIT(FILE)' checks for alternative programs checks for UNIX variants that set C preprocessor variables checks for header files checks for typedefs checks for library functions checks for structures checks for compiler characteristics checks for system services other checks for UNIX variants `AC_OUTPUT([FILE...])' You can include comments in `configure.in' files by starting them with the `m4' predefined macro `dnl', which discards text up through the next newline. These comments do not appear in the generated `configure' scripts. For example, it is helpful to begin `configure.in' files with a line like this: dnl Process this file with autoconf to produce a configure script. *Note Sample configure.in::, for an example of a real `configure.in' script.  File: autoconf.info, Node: Invoking autoconf, Next: Invoking autoheader, Prev: Writing configure.in, Up: Making configure Scripts Invoking `autoconf' =================== To create `configure' from `configure.in', run the `autoconf' program with no arguments. `autoconf' processes `configure.in' with the `m4' macro processor, using the Autoconf macros. If you give `autoconf' an argument, it reads that file instead of `configure.in' and writes the configuration script to the standard output instead of to `configure'. If you give `autoconf' the argument `-', it reads the standard input instead of `configure.in' and writes the configuration script on the standard output. The Autoconf macros are defined in two or more files. Two of the files are distributed with Autoconf: `autoconf' first reads `acgeneral.m4' (*note General Purpose Macros::.), then `acspecific.m4' (*note Specific Tests::.). After reading them, `autoconf' looks for an optional file called `aclocal.m4', first in the directory that contains other installed Autoconf macro files, and then in the current directory. If both files exist, it uses both of them. Those files can contain your site's own locally written Autoconf macro definitions (*note Writing Macros::., for more information). If a macro is defined in more than one of the files that `autoconf' reads, the last definition it reads overrides the earlier ones. You can override the directory where `autoconf' looks for the installed macro files by setting the `AC_MACRODIR' environment variable to a different directory. You can also give `autoconf' the `--macrodir' option, which overrides `AC_MACRODIR'. `autoconf' also accepts the options `--version', which prints the Autoconf version number and exits, and `--help', which prints a summary of the command-line options and exits.  File: autoconf.info, Node: Invoking autoheader, Prev: Invoking autoconf, Up: Making configure Scripts Invoking `autoheader' ===================== You can use the program `autoheader' to create a template file of C `#define' statements for `configure' to use. By default, the file that `autoheader' creates is called `config.h.in'; if `configure.in' invokes `AC_CONFIG_HEADER(FILE)', `autoheader' creates `FILE.in'. `autoheader' scans `configure.in' and figures out which C preprocessor symbols it might define. It copies comments and `#define' and `#undef' statements from a file called `acconfig.h', which comes with and is installed with Autoconf. It also uses a file called `acconfig.h' in the current directory, if present; you must create that file to contain entries for symbols that you `AC_DEFINE'. For symbols that `AC_HAVE_HEADERS' or `AC_HAVE_FUNCS' define, `autoheader' generates comments and `#undef' statements itself rather than copying them from a file, since the possible symbols are effectively limitless. The file that `autoheader' creates contains only `#define' and `#undef' statements and their accompanying comments; there is no provision in `autoheader' for adding other code or comments to the file. If you give `autoheader' an argument, it uses that file instead of `configure.in' and writes the header file to the standard output instead of to `config.h.in'. If you give `autoheader' an argument of `-', it reads the standard input instead of `configure.in' and writes the header file to the standard output. You can override the directory where `autoheader' looks for the installed macro and `acconfig.h' files by setting the `AC_MACRODIR' environment variable to a different directory. You can also give `autoheader' the `--macrodir' option, which overrides `AC_MACRODIR'. `autoheader' also accepts the options `--version', which prints the Autoconf version number and exits, and `--help', which prints a summary of the command-line options and exits.  File: autoconf.info, Node: Specific Tests, Next: General Purpose Macros, Prev: Making configure Scripts, Up: Top Specific Tests ************** These macros test for particular operating system features that packages might need or want to use. If you need to test for a feature that none of these macros check for, you can probably do it by calling one of the general purpose test macros with appropriate arguments (*note General Tests::.). All of these macros that set `make' variables call `AC_SUBST' on those variables (*note Setting Variables::., for details about `AC_SUBST'). The phrase "define NAME" is used below as a shorthand to mean either add `-DNAME=1' to the `make' variable `DEFS', or put `#define NAME 1' in the configuration header file, depending on whether `AC_CONFIG_HEADER' has been called. *Note Setting Variables::, for more information. Within each section below, the macros are listed in alphabetical order. The macros are generally named for the `make' variables or C preprocessor macros that they define; those names are based largely on what existing GNU programs use. These macros are defined in the file `acspecific.m4'. * Menu: * Alternative Programs:: Selecting between alternative programs. * Header Files:: Header files that might be missing. * Typedefs:: `typedef's that might be missing. * Library Functions:: C library functions that might be missing. * Structures:: Structures or members that might be missing. * Compiler Characteristics:: C compiler or machine architecture features. * System Services:: Operating system services. * UNIX Variants:: Special cases for specific UNIX variants.  File: autoconf.info, Node: Alternative Programs, Next: Header Files, Up: Specific Tests Alternative Programs ==================== The following macros check for the presence or behavior of particular programs: - Macro: AC_DECLARE_YYTEXT Define `DECLARE_YYTEXT' to declare `yytext' appropriately, depending on whether `lex' or `flex' is being used. This macro calls `AC_PROG_CPP' and `AC_PROG_LEX' if they haven't been called already. - Macro: AC_LN_S If `ln -s' works on the current filesystem (the O.S. and filesystem support symbolic links), set shell and `make' variable `LN_S' to `ln -s', otherwise set it to `ln'. - Macro: AC_MINUS_C_MINUS_O If the C compiler does not accept the `-c' and `-o' options simultaneously, define `NO_MINUS_C_MINUS_O'. - Macro: AC_PROG_YACC If `bison' is found, set `make' variable `YACC' to `bison -y'. Otherwise, if `byacc' is found, set `YACC' to `byacc'. Otherwise set `YACC' to `yacc'. - Macro: AC_PROG_CPP Set shell and `make' variable `CPP' to a command that runs the C preprocessor. If `$CC -E' doesn't work, it uses `/lib/cpp'. Many of the specific test macros use the value of `CPP' indirectly by calling `AC_TEST_CPP', `AC_HEADER_CHECK', `AC_HEADER_EGREP', or `AC_PROGRAM_EGREP'. Those macros call this macro first if it hasn't been called already. It calls `AC_PROG_CC' if it hasn't been called already. - Macro: AC_PROG_LEX If `flex' is found, set `make' variable `LEX' to `flex' and `LEXLIB' to `-lfl', if that library is in a standard place. Otherwise set `LEX' to `lex' and `LEXLIB' to `-ll'. - Macro: AC_PROG_AWK Check for `mawk', `gawk', `nawk', and `awk', in that order, and set `make' variable `AWK' to the first one that it finds. - Macro: AC_PROG_CC If `gcc' is found, set `make' variable `CC' to `gcc', and set shell variable `GCC' to 1 for use by macros such as `AC_GCC_TRADITIONAL'. - Macro: AC_GCC_TRADITIONAL Add `-traditional' to `make' variable `CC' if using the GNU C compiler and `ioctl' does not work properly without `-traditional'. This macro calls `AC_PROG_CC' and `AC_PROG_CPP' if they haven't been called already. - Macro: AC_PROG_INSTALL Set `make' variable `INSTALL' to `install -c' if `install' is found, otherwise set it to `cp'. Screens out the false matches `/etc/install', `/usr/sbin/install', and other instances of `install' known not to work. The variable `INSTALL_PROGRAM' is set to `$(INSTALL)' and `INSTALL_DATA' is set to `$(INSTALL) -m 644'. - Macro: AC_PROG_RANLIB Set `make' variable `RANLIB' to `ranlib' if `ranlib' is found, otherwise to `:' (do nothing). - Macro: AC_RSH If a remote shell is available, put `rtapelib.o' in `make' variable `RTAPELIB'. Otherwise, also do so if `netdb.h' exists (implying the `rexec' function), and in addition define `HAVE_NETDB_H'. If neither a remote shell nor `rexec' is available, define `NO_REMOTE'.  File: autoconf.info, Node: Header Files, Next: Typedefs, Prev: Alternative Programs, Up: Specific Tests Header Files ============ The following macros check for the presence of certain C header files: - Macro: AC_DIR_HEADER If the system has `dirent.h', define `DIRENT'; otherwise, if it has `sys/ndir.h', define `SYSNDIR'; otherwise, if it has `sys/dir.h', define `SYSDIR'; otherwise, if it has `ndir.h', define `NDIR'. Also, if the directory library header file contains a declaration of the `closedir' function with a `void' return type, define `VOID_CLOSEDIR'. The directory library declarations in the source code should look something like the following, which assumes that you have also called `AC_HAVE_HEADERS(unistd.h)': #ifdef HAVE_UNISTD_H #include #include #endif /* unistd.h defines _POSIX_VERSION on POSIX.1 systems. */ #if defined(DIRENT) || defined(_POSIX_VERSION) #include #define NLENGTH(dirent) (strlen((dirent)->d_name)) #else /* not (DIRENT or _POSIX_VERSION) */ #define dirent direct #define NLENGTH(dirent) ((dirent)->d_namlen) #ifdef SYSNDIR #include #endif /* SYSNDIR */ #ifdef SYSDIR #include #endif /* SYSDIR */ #ifdef NDIR #include #endif /* NDIR */ #endif /* not (DIRENT or _POSIX_VERSION) */ Using the above declarations, the program would declare variables to be type `struct dirent', not `struct direct', and would access the length of a directory entry name by passing a pointer to a `struct dirent' to the `NLENGTH' macro. - Macro: AC_MAJOR_HEADER If `sys/types.h' does not define `major', `minor', and `makedev', but `sys/mkdev.h' does, define `MAJOR_IN_MKDEV'; otherwise, if `sys/sysmacros.h' does, define `MAJOR_IN_SYSMACROS'. - Macro: AC_MEMORY_H Define `NEED_MEMORY_H' if `memcpy', `memcmp', etc. are not declared in `string.h' and `memory.h' exists. This macro is obsolete; instead, use `AC_HAVE_HEADERS(memory.h)'. See the example for `AC_STDC_HEADERS'. - Macro: AC_STDC_HEADERS Define `STDC_HEADERS' if the system has ANSI C header files. Specifically, this macro checks for `stdlib.h', `stdarg.h', `string.h', and `float.h'; if the system has those, it probably has the rest of the ANSI C header files. This macro also checks whether `string.h' declares `memchr' (and thus presumably the other `mem' functions) and whether the `ctype.h' macros work on characters with the high bit set, as ANSI C requires. Use `STDC_HEADERS' instead of `__STDC__' to determine whether the system has ANSI-compliant header files (and probably C library functions) because many systems that have GCC do not have ANSI C header files. To check whether to use the System V/ANSI C string functions and header file, you can put the following in `configure.in': AC_STDC_HEADERS AC_HAVE_HEADERS(string.h memory.h) Then, in the code, use a test like this: #if STDC_HEADERS || HAVE_STRING_H #include /* An ANSI string.h and pre-ANSI memory.h might conflict. */ #if !STDC_HEADERS && HAVE_MEMORY_H #include #endif /* not STDC_HEADERS and HAVE_MEMORY_H */ #define index strchr #define rindex strrchr #define bcopy(s, d, n) memcpy ((d), (s), (n)) #define bcmp(s1, s2, n) memcmp ((s1), (s2), (n)) #define bzero(s, n) memset ((s), 0, (n)) #else /* not STDC_HEADERS and not HAVE_STRING_H */ #include /* memory.h and strings.h conflict on some systems. */ #endif /* not STDC_HEADERS and not HAVE_STRING_H */ This example asssumes that your code uses the BSD style functions. If you use the System V/ANSI C style functions, you will need to replace the macro definitions with ones that go in the other direction. - Macro: AC_UNISTD_H Define `HAVE_UNISTD_H' if the system has `unistd.h'. This macro is obsolete; instead, use `AC_HAVE_HEADERS(unistd.h)'. The way to check if the system supports POSIX.1 is: #if HAVE_UNISTD_H #include #include #endif #ifdef _POSIX_VERSION /* Code for POSIX.1 systems. */ #endif `_POSIX_VERSION' is defined when `unistd.h' is included on POSIX.1 systems. If there is no `unistd.h', it is definitely not a POSIX.1 system. However, some non-POSIX.1 systems do have `unistd.h'. - Macro: AC_USG Define `USG' if the system does not have `strings.h', `rindex', `bzero', etc. This implies that it has `string.h', `strrchr', `memset', etc. The symbol `USG' is obsolete. Instead of this macro, use `AC_HAVE_HEADERS(string.h)' and use `HAVE_STRING_H' in your code. See the example for `AC_STDC_HEADERS'.  File: autoconf.info, Node: Typedefs, Next: Library Functions, Prev: Header Files, Up: Specific Tests Typedefs ======== The following macros check for predefined C types: - Macro: AC_GETGROUPS_T Define `GETGROUPS_T' to be whichever of `gid_t' or `int' is the base type of the array argument to `getgroups'. - Macro: AC_MODE_T If `mode_t' is not defined in `sys/types.h', define `mode_t' to be `int'. - Macro: AC_OFF_T If `off_t' is not defined in `sys/types.h', define `off_t' to be `long'. - Macro: AC_PID_T If `pid_t' is not defined in `sys/types.h', define `pid_t' to be `int'. - Macro: AC_RETSIGTYPE If `signal.h' declares `signal' as returning a pointer to a function returning `void', define `RETSIGTYPE' to be `void'; otherwise, define it to be `int'. Define signal handlers as returning type `RETSIGTYPE': RETSIGTYPE hup_handler () { ... } - Macro: AC_SIZE_T If `size_t' is not defined in `sys/types.h', define `size_t' to be `unsigned'. - Macro: AC_UID_T If `uid_t' is not defined in `sys/types.h', define `uid_t' to be `int' and `gid_t' to be `int'.  File: autoconf.info, Node: Library Functions, Next: Structures, Prev: Typedefs, Up: Specific Tests Library Functions ================= The following macros check for particular C library functions: - Macro: AC_ALLOCA Check how to get `alloca'. Tries to get a builtin version by checking for `alloca.h' or the predefined C preprocessor macros `__GNUC__' and `_AIX'. If that fails, it looks for a function in the standard C library. If that fails, it sets the `make' variable `ALLOCA' to `alloca.o' and defines `C_ALLOCA' (so programs can periodically call `alloca(0)' to garbage collect). This variable is separate from `LIBOBJS' so multiple programs can share the value of `ALLOCA' without needing to create an actual library, in case only some of them use the code in `LIBOBJS'. If this macro finds `alloca.h', it defines `HAVE_ALLOCA_H'. This macro does not try to get `alloca' from the SVR3 `libPW' or the SVR4 `libucb' because those libraries contain some incompatible functions that cause trouble. Some versions do not even contain `alloca' or contain a buggy version. If you still want to use their `alloca', use `ar' to extract `alloca.o' from them instead of compiling `alloca.c'. Source files that use `alloca' should start with a piece of code like the following, to declare it properly. Note that in some versions of AIX, the declaration of `alloca' must precede everything else except for comments and preprocessor directives. The `#pragma' directive is indented so that pre-ANSI C compilers will ignore it, rather than choke on it. /* AIX requires this to be the first thing in the file. */ #ifdef __GNUC__ #define alloca __builtin_alloca #else /* not __GNUC__ */ #if HAVE_ALLOCA_H #include #else /* not HAVE_ALLOCA_H */ #ifdef _AIX #pragma alloca #else /* not _AIX */ char *alloca (); #endif /* not _AIX */ #endif /* not HAVE_ALLOCA_H */ #endif /* not __GNUC__ */ - Macro: AC_GETLOADAVG Check how to get the system load averages. If the system has the `getloadavg' function, this macro defines `HAVE_GETLOADAVG', and adds to `LIBS' any libraries needed to get that function. Otherwise, it adds `getloadavg.o' to the `make' variable `LIBOBJS', and possibly defines several other C preprocessor macros and `make' variables: 1. It defines `SVR4', `DGUX', `UMAX', or `UMAX4_3' if on those systems. 2. If it finds `nlist.h', it defines `NLIST_STRUCT'. 3. If `struct nlist' has an `n_un' member, it defines `NLIST_NAME_UNION'. 4. If compiling `getloadavg.c' defines `LDAV_PRIVILEGED', programs need to be installed specially on this system for `getloadavg' to work, and this macro defines `GETLOADAVG_PRIVILEGED'. 5. This macro always defines `NEED_SETGID', for `make'. The value is `true' if special installation is required, `false' if not. If `NEED_SETGID' is `true', it sets `KMEM_GROUP' to the name of the group that should own the installed program. - Macro: AC_SETVBUF_REVERSED If `setvbuf' takes the buffering type as its second argument and the buffer pointer as the third, instead of the other way around, define `SETVBUF_REVERSED'. This is the case on System V before release 3. - Macro: AC_STRCOLL If the `strcoll' function exists and works correctly, define `HAVE_STRCOLL'. This does a bit more than `AC_HAVE_FUNCS(strcoll)', because some systems have incorrect definitions of `strcoll', which should not be used. - Macro: AC_UTIME_NULL If `utime(FILE, NULL)' sets FILE's timestamp to the present, define `HAVE_UTIME_NULL'. - Macro: AC_VFORK If `vfork.h' is found, define `HAVE_VFORK_H'. If a working `vfork' is not found, define `vfork' to be `fork'. This macro checks for several known errors in implementations of `vfork' and considers the system to not have a working `vfork' if it detects any of them. - Macro: AC_VPRINTF If `vprintf' is found, define `HAVE_VPRINTF'. Otherwise, if `_doprnt' is found, define `HAVE_DOPRNT'. - Macro: AC_WAIT3 If `wait3' is found and fills in the contents of its third argument (a `struct rusage *'), which HP-UX does not do, define `HAVE_WAIT3'.  File: autoconf.info, Node: Structures, Next: Compiler Characteristics, Prev: Library Functions, Up: Specific Tests Structures ========== The following macros check for certain structures or structure members: - Macro: AC_STAT_MACROS_BROKEN If the macros `S_ISDIR', `S_ISREG' et al. defined in `sys/stat.h' do not work properly (returning false positives), define `STAT_MACROS_BROKEN'. This is the case on Tektronix UTekV, Amdahl UTS and Motorola System V/88. - Macro: AC_ST_BLKSIZE If `struct stat' contains an `st_blksize' member, define `HAVE_ST_BLKSIZE'. - Macro: AC_ST_BLOCKS If `struct stat' contains an `st_blocks' member, define `HAVE_ST_BLOCKS'. Otherwise, add `fileblocks.o' to the `make' variable `LIBOBJS'. - Macro: AC_ST_RDEV If `struct stat' contains an `st_rdev' member, define `HAVE_ST_RDEV'. - Macro: AC_TIME_WITH_SYS_TIME If a program may include both `time.h' and `sys/time.h', define `TIME_WITH_SYS_TIME'. On some older systems, `sys/time.h' includes `time.h', but `time.h' is not protected against multiple inclusion, so programs should not explicitly include both files. This macro is useful in programs that use, for example, `struct timeval' or `struct timezone' as well as `struct tm'. It is best used in conjunction with `HAVE_SYS_TIME_H'. #ifdef TIME_WITH_SYS_TIME #include #include #else #ifdef HAVE_SYS_TIME_H #include #else #include #endif #endif - Macro: AC_STRUCT_TM If `time.h' does not define `struct tm', define `TM_IN_SYS_TIME', which means that including `sys/time.h' defines `struct tm'. - Macro: AC_TIMEZONE Figure out how to get the current timezone. If `struct tm' has a `tm_zone' member, define `HAVE_TM_ZONE'. Otherwise, if the external array `tzname' is found, define `HAVE_TZNAME'. This macro calls `AC_STRUCT_TM' if it hasn't been called already.  File: autoconf.info, Node: Compiler Characteristics, Next: System Services, Prev: Structures, Up: Specific Tests Compiler Characteristics ======================== The following macros check for C compiler or machine architecture features: - Macro: AC_ARG_ARRAY If the address of an argument to a C function can not be used like the start of an array, define `NO_ARG_ARRAY'. This ability allows a sequence of arguments with the same type to be accessed as if they were an array of values. - Macro: AC_CROSS_CHECK If the C compiler being used does not produce executables that can run on the system where `configure' is being run, set the shell variable `cross_compiling' to 1. This information can be used by `AC_TEST_PROGRAM' to determine whether to take a default action instead of trying to run a test program (*note General Tests::.). - Macro: AC_CHAR_UNSIGNED If the C type `char' is unsigned, define `__CHAR_UNSIGNED__', unless the C compiler predefines it. - Macro: AC_CONST If the C compiler does not fully support the keyword `const', define `const' to be empty. Some C compilers that do not define `__STDC__' do support `const'; some compilers that define `__STDC__' do not completely support `const'. Programs can simply use `const' as if every C compiler supported it; for those that don't, the `Makefile' or configuration header file will define it as empty. (If using a configuration header file, the program should include it before any other header files, to prevent inconsistencies in declarations.) - Macro: AC_INLINE If the C compiler is a version of GCC that supports the keyword `__inline' but not `inline' (such as some NeXT versions), define `inline' to be `__inline'. This macro calls `AC_PROG_CC' if it hasn't been called already. - Macro: AC_INT_16_BITS If the C type `int' is 16 bits wide, define `INT_16_BITS'. - Macro: AC_LONG_64_BITS If the C type `long int' is 64 bits wide, define `LONG_64_BITS'. - Macro: AC_LONG_DOUBLE If the C compiler supports the `long double' type, define `HAVE_LONG_DOUBLE'. Some C compilers that do not define `__STDC__' do support the `long double' type; some compilers that define `__STDC__' do not support `long double'. - Macro: AC_WORDS_BIGENDIAN If words are stored with the most significant byte first (like Motorola and SPARC, but not Intel and VAX, CPUs), define `WORDS_BIGENDIAN'.  File: autoconf.info, Node: System Services, Next: UNIX Variants, Prev: Compiler Characteristics, Up: Specific Tests System Services =============== The following macros check for operating system services: - Macro: AC_FIND_X Try to locate the X Window System include files and libraries. Try first by running `xmkmf' on a trivial `Imakefile' and examining the `Makefile' that it produces. If that fails (such as if `xmkmf' is not present), look for them in several directories where they often reside. If either method is successful, set the shell variables `x_includes' and `x_libraries' to their locations, unless they are in directories the compiler searches by default. If both methods fail, set the shell variable `no_x' to `true'; otherwise set it to the empty string. - Macro: AC_HAVE_POUNDBANG (ACTION-IF-SUPPORTED [, ACTION-IF-UNSUPPORTED]) Print `checking if `#!' works in shell scripts' to the standard output. Then check whether the system supports starting shell scripts with a line of the form `#!/bin/csh' to select the shell to run them with. ACTION-IF-SUPPORTED is a list of shell commands to run if `#!' works; ACTION-IF-NOT-SUPPORTED is a list of shell commands to run otherwise. - Macro: AC_LONG_FILE_NAMES If the system supports file names longer than 14 characters, define `HAVE_LONG_FILE_NAMES'. - Macro: AC_REMOTE_TAPE If BSD tape drive ioctls are available, define `HAVE_SYS_MTIO_H', and if sockets are available add `rmt' to `make' variable `PROGS'. - Macro: AC_RESTARTABLE_SYSCALLS If the system automatically restarts a system call that is interrupted by a signal, define `HAVE_RESTARTABLE_SYSCALLS'.  File: autoconf.info, Node: UNIX Variants, Prev: System Services, Up: Specific Tests UNIX Variants ============= The following macros check for certain operating systems that need special treatment for some programs, due to exceptional oddities in their header files or libraries: - Macro: AC_AIX If on AIX, define `_ALL_SOURCE'. Allows the use of some BSD functions. Should be called before any macros that run the C compiler. - Macro: AC_DYNIX_SEQ If on DYNIX/ptx (Sequent UNIX), add `-lseq' to `make' variable `LIBS'. Allows use of some BSD system calls and `getmntent'. - Macro: AC_IRIX_SUN If on IRIX (Silicon Graphics UNIX), add `-lsun' to `make' variable `LIBS'. Needed to get `getmntent'. At sites using Yellow Pages/NIS, it is also needed to get properly working `gethostby*', `getpw*', `getgr*', `getnetby*', and so on. - Macro: AC_ISC_POSIX If on a POSIXized ISC UNIX, define `_POSIX_SOURCE' and add `-posix' (for the GNU C compiler) or `-Xp' (for other C compilers) to `make' variable `CC'. This allows the use of POSIX facilities. Must be called after `AC_PROG_CC' and before any other macros that run the C compiler. - Macro: AC_MINIX If on Minix, define `_MINIX' and `_POSIX_SOURCE' and define `_POSIX_1_SOURCE' to be 2. This allows the use of POSIX facilities. Should be called before any macros that run the C compiler. - Macro: AC_SCO_INTL If on SCO UNIX, add `-lintl' to `make' variable `LIBS'. Used to get `strftime'. It must be called before checking for `strftime'. - Macro: AC_XENIX_DIR If on Xenix, define `VOID_CLOSEDIR' and add `-lx' to `make' variable `LIBS'. Also, if `sys/ndir.h' is not being used, add `-ldir' to `LIBS'. Needed when using the directory reading functions. This macro must be called after `AC_DIR_HEADER'.  File: autoconf.info, Node: General Purpose Macros, Next: Writing Macros, Prev: Specific Tests, Up: Top General Purpose Macros ********************** These macros provide ways for other macros to control the kind of output that Autoconf produces or to check whether various kinds of features are available. They all take arguments. When calling these macros, there must not be any blank space between the macro name and the open parentheses. Arguments to these macros can be more than one line long if they are enclosed within the `m4' quote characters `[' and `]'. Within each section below, the macros are listed in alphabetical order. These macros are defined in the file `acgeneral.m4'. * Menu: * Setup:: Controlling Autoconf operation. * General Tests:: Check for kinds of features. * Setting Variables:: Setting shell and `make' variables. * Macro Ordering:: Enforcing ordering constraints.  File: autoconf.info, Node: Setup, Next: General Tests, Up: General Purpose Macros Controlling Autoconf Setup ========================== The following macros control the kind of output that Autoconf produces. - Macro: AC_CONFIG_HEADER (HEADER-TO-CREATE ...) Make `AC_OUTPUT' create the file(s) in the whitespace-separated list HEADER-TO-CREATE containing C preprocessor `#define' statements and replace `@DEFS@' in generated files with `-DHAVE_CONFIG_H' instead of the value of `DEFS'. This macro should be called right after `AC_INIT'. The usual name for HEADER-TO-CREATE is `config.h'. If HEADER-TO-CREATE already exists and its contents are identical to what `AC_OUTPUT' would put in it, it is left alone. Doing this allows some changes in configuration without needlessly causing object files that depend on the header file to be recompiled. Your distribution should contain a file `HEADER-TO-CREATE.in' that looks as you want the final header file to look, including comments, with default values in the `#define' statements. A default value can be to `#undef' the variable instead of to define it to a value, if your code tests for configuration options using `#ifdef' instead of `#if'. You can use the program `autoheader' to create `HEADER-TO-CREATE.in' (*note Invoking autoheader::.). - Macro: AC_INIT (UNIQUE-FILE-IN-SOURCE-DIR) Process the command-line arguments and find the source code directory. UNIQUE-FILE-IN-SOURCE-DIR is some file that is in the package's source directory; `configure' checks for this file's existence to make sure that the directory that it is told contains the source code in fact does (*note Running configure Scripts::., for more information). - Macro: AC_PREPARE (UNIQUE-FILE-IN-SOURCE-DIR) Find the source code directory and set up shell variables necessary for other Autoconf macros to work. uNIQUE-FILE-IN-SOURCE-DIR is some file that is in the package's source directory; `configure' checks for this file's existence to make sure that the directory that it is told contains the source code in fact does (*note Running configure Scripts::., for more information). `AC_PREPARE' is the last thing done by `AC_INIT'. Use `AC_PREPARE' instead of `AC_INIT' if you want to do argument parsing yourself; never use both. - Macro: AC_OUTPUT ([FILE...] [,EXTRA-CMDS]) Create output files (typically one or more `Makefile's) and `config.status'. If `AC_CONFIG_HEADER' has been called, also create the header file that was named as its argument. The argument is a whitespace-separated list of files to create; if it is omitted, no files are created. `AC_OUTPUT' creates each file `FILE' in the list by copying `FILE.in', substituting the variable values that have been selected by calling `AC_SUBST'. It creates the directory that each file is in if it doesn't exist (but not the parents of that directory). A plausible value for the argument to `AC_OUTPUT' is `Makefile src/Makefile man/Makefile X/Imakefile'. If you pass EXTRA-CMDS, those commands will be inserted into `config.status' to be run after all its other processing. - Macro: AC_REVISION (REVISION-INFO) Copy revision stamp REVISION-INFO into the `configure' script, with any dollar signs or double-quotes removed. This macro lets you put a revision stamp from `configure.in' into `configure' without RCS or CVS changing it when you check in `configure'. That way, you can determine easily which revision of `configure.in' a particular `configure' corresponds to. It is a good idea to call this macro before `AC_INIT' so that the revision number is near the top of both `configure.in' and `configure'. To support doing that, the `AC_REVISION' output begins with `#!/bin/sh', like the normal start of a `configure' script does. For example, this line in `configure.in': AC_REVISION($Revision: 1.30 $)dnl produces this in `configure': #!/bin/sh # From configure.in Revision: 1.30  File: autoconf.info, Node: General Tests, Next: Setting Variables, Prev: Setup, Up: General Purpose Macros Checking for Kinds of Features ============================== These macros are templates that, when called with actual parameters, check for various kinds of features. Many of these macros handle two cases: what to do if the given condition is met, and what to do if the condition is not met. In some places you you might want to do something if a condition is true but do nothing if it's false, or vice versa. To omit the true case, pass an empty value for the ACTION-IF-FOUND argument to the macro. To omit the false case, omit the ACTION-IF-NOT-FOUND argument to the macro, including the comma before it. One shell programming construction that you should not use in the action arguments to these macros is `VAR=${VAR:-VALUE}'. Old BSD shells, including the Ultrix `sh', don't understand the colon, and complain and die. If you omit the colon, it works fine: `VAR=${VAR-VALUE}'. *Note Writing Macros::, for more information on how best to use these macros. - Macro: AC_COMPILE_CHECK (ECHO-TEXT, INCLUDES, FUNCTION-BODY, ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND]) Print `checking for ECHO-TEXT' to the standard output. Then create a test C program to see whether a function whose body consists of FUNCTION-BODY can be compiled and linked; INCLUDES is any `#include' statements needed by the code in FUNCTION-BODY. If the file compiles and links successfully, run shell commands ACTION-IF-FOUND, otherwise run ACTION-IF-NOT-FOUND. - Macro: AC_FUNC_CHECK (FUNCTION, ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND]) If FUNCTION is available, run shell commands ACTION-IF-FOUND, otherwise ACTION-IF-NOT-FOUND. If the functions might be in libraries other than the default C library, first call `AC_HAVE_LIBRARY' for those libraries. If you just want to define a symbol if the function is available, consider using `AC_HAVE_FUNCS' instead. - Macro: AC_HAVE_FUNCS (FUNCTION...) For each given FUNCTION in the whitespace-separated argument list that is available, define `HAVE_FUNCTION' (in all caps). *Note Specific Tests::, for a precise definition of "define" as it is used here. If the functions might be in libraries other than the default C library, first call `AC_HAVE_LIBRARY' for those libraries. - Macro: AC_HAVE_HEADERS (HEADER-FILE...) For each given HEADER-FILE in the whitespace-separated argument list that exists, define `HAVE_HEADER-FILE' (in all caps). *Note Specific Tests::, for a precise definition of "define" as it is used here. - Macro: AC_HAVE_LIBRARY (LIBRARY [, ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND]]) Print `checking for LIBRARY' to the standard output. Then create a test C program to see whether that program can be linked with the specified library. ACTION-IF-FOUND is a list of shell commands to run if the link succeeds (which means that the library is present); ACTION-IF-NOT-FOUND is a list of shell commands to run if the link fails. If ACTION-IF-FOUND and ACTION-IF-NOT-FOUND are not specified, the default action is to add `-lfoo' to `LIBS' and define `HAVE_LIBfoo' for library `foo'. LIBRARY can be written as any of `foo', `-lfoo', or `libfoo.a'. In all of those cases, the compiler is passed `-lfoo'. - Macro: AC_HEADER_CHECK (HEADER-FILE, ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND]) If HEADER-FILE exists, execute shell commands ACTION-IF-FOUND, otherwise execute ACTION-IF-NOT-FOUND. If you just want to define a symbol if the header file is available, consider using `AC_HAVE_HEADERS' instead. - Macro: AC_HEADER_EGREP (PATTERN, HEADER-FILE, ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND]) If the output of running the C preprocessor on HEADER-FILE contains the `egrep' regular expression PATTERN, execute shell commands ACTION-IF-FOUND, otherwise execute ACTION-IF-NOT-FOUND. You can not check whether preprocessor symbols are defined this way, because they get expanded before `egrep' sees them. But you can almost always detect them by simply using `#ifdef' directives in your C programs. - Macro: AC_PREFIX (PROGRAM) If the user did not specify an installation prefix on the command line, guess a value for it by looking for PROGRAM in `PATH', the way the shell does. If PROGRAM is found, set the prefix to the parent of the directory containing PROGRAM; otherwise leave the prefix specified in `Makefile.in' unchanged. For example, if PROGRAM is `gcc' and the `PATH' contains `/usr/local/gnu/bin/gcc', set the prefix to `/usr/local/gnu'. - Macro: AC_PROGRAM_CHECK (VARIABLE, PROG-TO-CHECK-FOR, VALUE-IF-FOUND, VALUE-IF-NOT-FOUND) Check whether program PROG-TO-CHECK-FOR exists in `PATH'. If it is found, set VARIABLE to VALUE-IF-FOUND, otherwise to VALUE-IF-NOT-FOUND. Calls `AC_SUBST' for VARIABLE. - Macro: AC_PROGRAM_EGREP (PATTERN, PROGRAM, ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND]) PROGRAM is the text of a C program, on which shell variable and backquote substitutions are performed. If the output of running the C preprocessor on PROGRAM contains the `egrep' regular expression PATTERN, execute shell commands ACTION-IF-FOUND, otherwise execute ACTION-IF-NOT-FOUND. (It is an unfortunate oversight that we use the word `PROGRAM' in Autoconf macro names to sometimes mean C source code and sometimes mean a UNIX command.) - Macro: AC_PROGRAMS_CHECK (VARIABLE, PROGS-TO-CHECK-FOR [, VALUE-IF-NOT-FOUND]) Check for each program in the whitespace-separated list PROGS-TO-CHECK-FOR exists in `PATH'. If it is found, set VARIABLE to the name of that program. Otherwise, continue checking the next program in the list. If none of the programs in the list are found, set VARIABLE to VALUE-IF-NOT-FOUND; if VALUE-IF-NOT-FOUND is not specified, the value of VARIABLE will not be changed. Calls `AC_SUBST' for VARIABLE. - Macro: AC_REPLACE_FUNCS (FUNCTION-NAME...) For each given FUNCTION-NAME in the whitespace-separated argument list that is not in the C library, add `FUNCTION-NAME.o' to the value of the `make' variable `LIBOBJS'. If the functions might be in libraries other than the default C library, first call `AC_HAVE_LIBRARY' for those libraries. - Macro: AC_TEST_PROGRAM (PROGRAM, ACTION-IF-TRUE [, ACTION-IF-FALSE [, ACTION-IF-CROSS-COMPILING]]) PROGRAM is the text of a C program, on which shell variable and backquote substitutions are performed. If it compiles and links successfully and returns an exit status of 0 when executed, run shell commands ACTION-IF-TRUE. Otherwise run shell commands ACTION-IF-FALSE. If the optional argument ACTION-IF-CROSS-COMPILING is given and the C compiler being used does not produce executables that run on the system where `configure' is being run, then the test program is not run. Instead, the shell commands ACTION-IF-CROSS-COMPILING are run. If that argument is given, this macro calls `AC_CROSS_CHECK' if it has not already been called (*note Compiler Characteristics::.). - Macro: AC_TEST_CPP (INCLUDES, ACTION-IF-TRUE [, ACTION-IF-FALSE]) INCLUDES is C `#include' statements and declarations, on which shell variable and backquote substitutions are performed. (Actually, it can be any C program, but other statements are probably not useful.) If the C preprocessor produces no error messages while processing it, run shell commands ACTION-IF-TRUE. Otherwise run shell commands ACTION-IF-FALSE. This macro calls `AC_PROG_CPP' if it hasn't been called already. - Macro: AC_WITH (FEATURE, ACTION-IF-TRUE [, ACTION-IF-FALSE]) If the user gave `configure' the option `--with-FEATURE', run shell commands ACTION-IF-TRUE. Otherwise run shell commands ACTION-IF-FALSE. The name FEATURE should consist only of alphanumeric characters and dashes; typical feature names are `gnu-libc' and `x'. The user can give a value by following the feature name with `=' and the value; for example, ./configure --with-targets=sun4,hp300bsd The value given is available to the shell commands ACTION-IF-TRUE in the shell variable `withval'. If no value was given, `withval' is 1. For example, AC_WITH(fubar, echo "got --with-fubar=$withval", echo no --with-fubar)  File: autoconf.info, Node: Setting Variables, Next: Macro Ordering, Prev: General Tests, Up: General Purpose Macros Setting Variables ================= These macros help other macros to define shell and `make' variables. - Macro: AC_DEFINE (VARIABLE [, VALUE]) Define C preprocessor variable VARIABLE. If VALUE is given, set VARIABLE to that value, otherwise set it to 1. To use a shell variable as the value, use `AC_DEFINE_UNQUOTED' instead. Precede double quotes in the value with backslashes. This macro adds to the shell variable `DEFS'. `AC_OUTPUT' later substitutes the values in `DEFS' into the file(s) that it generates (typically `Makefile'). Alternately, if `AC_CONFIG_HEADER' has been called, `AC_OUTPUT' creates a header file by substituting the correct values into `#define' statements in a template file. For example, suppose your `configure.in' calls `AC_CONFIG_HEADER(conf.h)' and `AC_UNISTD_H'. You could have code like this in `conf.h.in': /* Define as 1 if you have unistd.h. */ #define HAVE_UNISTD_H 0 On systems that have `unistd.h', `configure' will change the 0 to a 1. On other systems, it will leave the line unchanged. Alternately, if you prefer to use `#ifdef', your `conf.h.in' could have code like this: /* Define if you have unistd.h. */ #undef HAVE_UNISTD_H On systems that have `unistd.h', `configure' will change the second line to read `#define HAVE_UNISTD_H 1'. On other systems, it will leave the line unchanged. Due to the syntactical bizarreness of the Bourne shell, do not use semicolons to separate `AC_DEFINE' calls from other macro calls or shell code; that can cause syntax errors in the resulting `configure' script. Use either spaces or newlines. That is, do this: AC_HEADER_CHECK(elf.h, AC_DEFINE(SVR4) LIBS="$LIBS -lelf") or this: AC_HEADER_CHECK(elf.h, AC_DEFINE(SVR4) LIBS="$LIBS -lelf") instead of this: AC_HEADER_CHECK(elf.h, AC_DEFINE(SVR4); LIBS="$LIBS -lelf") - Macro: AC_DEFINE_UNQUOTED (VARIABLE [, VALUE]) Like `AC_DEFINE', but it does nothing to quote VALUE from various shell and `sed' expansions it will undergo. VALUE will be used in many different contexts requiring different quoting, and it is up to you to make sure it works right. Use this macro instead of `AC_DEFINE' when VALUE contains a shell variable. For example: AC_DEFINE_UNQUOTED(config_machfile, ${machfile}) - Macro: AC_SUBST (VARIABLE) Substitute the variable VARIABLE when creating the output files (typically one or more `Makefile's). This means replace instances of `@VARIABLE@', e.g. in `Makefile.in', with the current value of the shell variable VARIABLE. If this macro were not called, the value of VARIABLE would not be set in the output files, even though `configure' had figured out a value for it. You can set or add to the value of VARIABLE in the usual shell way. For example, to add `-ltermcap' to the value of the variable `LIBS': LIBS="$LIBS -ltermcap"  File: autoconf.info, Node: Macro Ordering, Prev: Setting Variables, Up: General Purpose Macros Macro Ordering ============== These macros provide ways for other macros to make sure that they are called in the correct order. - Macro: AC_BEFORE (THIS-MACRO-NAME, CALLED-MACRO-NAME) Make `m4' print a warning message on the standard error output if CALLED-MACRO-NAME has already been called. THIS-MACRO-NAME should be the name of the macro that is calling `AC_BEFORE'. The macro CALLED-MACRO-NAME must contain a call to `AC_PROVIDE' to indicate that it has been called. This macro should be used when one macro makes changes that might affect another macro, so that the other macro should probably not be called first. For example, `AC_PROG_CPP' checks whether the C compiler can run the C preprocessor when given the `-E' option. It should therefore be called after any macros that change which C compiler is being used, such as `AC_PROG_CC'. So `AC_PROG_CC' contains: AC_BEFORE([$0], [AC_PROG_CPP]) This warns the user if a call to `AC_PROG_CPP' has already occurred when `AC_PROG_CC' is called. - Macro: AC_OBSOLETE (THIS-MACRO-NAME [, SUGGESTION]) Make `m4' print a message on the standard error output warning that THIS-MACRO-NAME is obsolete, and giving the file and line number where it was called. THIS-MACRO-NAME should be the name of the macro that is calling `AC_BEFORE'. If SUGGESTION is given, it is printed at the end of the warning message; for example, it can be a suggestion for what to use instead of THIS-MACRO-NAME. A sample call is: AC_OBSOLETE([$0], [; use AC_HAVE_HEADERS(unistd.h) instead]) - Macro: AC_PROVIDE (MACRO-NAME) Set a flag recording that MACRO-NAME has been called. The argument should be the name of the macro that is calling `AC_PROVIDE'. An easy way to get it is from the `m4' builtin variable `$0', like this: AC_PROVIDE([$0]) - Macro: AC_REQUIRE (MACRO-NAME) If the `m4' macro MACRO-NAME has not already been called, call it (without any arguments). Make sure to quote MACRO-NAME with square brackets. The body of MACRO-NAME must contain a call to `AC_PROVIDE' to indicate that it has been called. Macros that need some other macro to be called before they are called can use `AC_REQUIRE' to ensure that it has been, in case the person who made `configure.in' forgot or didn't know to do it. `AC_REQUIRE' and `AC_PROVIDE' together can ensure that a macro is only called if it is needed, and only called once. *Note Dependencies Between Macros::, for more information.  File: autoconf.info, Node: Writing Macros, Next: Makefiles, Prev: General Purpose Macros, Up: Top Writing Macros ************** If your package needs to test for some feature that none of the macros supplied with Autoconf handles, you'll need to write one or more new Autoconf macros. Here are some suggestions and some of the rationale behind why the existing macros are written the way they are. You can also learn a lot about how to write Autoconf macros by looking at the existing ones. If something goes wrong in one or more of the Autoconf tests, this information can help you understand why they work the way they do and the assumptions behind them, which might help you figure out how to best solve the problem. If you add macros that you think would be useful to other people, or find problems with the distributed macros, please send electronic mail to `bug-gnu-utils@prep.ai.mit.edu', so we can consider them for future releases of Autoconf. Please include the Autoconf version number, which you can get by running `autoconf --version'. * Menu: * Macro Format:: Basic format of an Autoconf macro. * Quoting:: Protecting macros from unwanted expansion. * Dependencies Between Macros:: What to do when macros depend on other macros. * Checking for Files:: Finding whether a file exists. * Checking for Symbols:: Finding whether a symbol is defined. * Test Programs:: Writing programs to test for features. * Multiple Cases:: Tests for several possible values.  File: autoconf.info, Node: Macro Format, Next: Quoting, Up: Writing Macros Macro Format ============ Autoconf macros are defined as arguments to the `m4' builtin command `define'. Their overall structure looks like this: define(MACRO-NAME, [MACRO-BODY])dnl The square brackets here do not indicate optional text: they should literally be present in the macro definition. All of the Autoconf macros have names starting with `AC_' to prevent them from accidentally conflicting with other text. You should prefix your own macro names with some other sequence, such as your initials or an abbreviation for the name of your organization or software package, to ensure that their names don't conflict with the names of present or future Autoconf macros. The `m4' builtin `dnl' prevents a newline from being inserted in the output where the macro is defined; without it, the generated `configure' script would begin with dozens of blank lines. `dnl' is also used to introduce comments in `m4'; it causes `m4' to discard the rest of the input line. You should quote the entire macro body with square brackets to avoid macro expansion problems (*note Quoting::.). You can refer to any arguments passed to the macro as `$1', `$2', etc. *Note How to define new macros: (m4.info)Definitions, for more complete information on writing `m4' macros.  File: autoconf.info, Node: Quoting, Next: Dependencies Between Macros, Prev: Macro Format, Up: Writing Macros Quoting ======= Macros that are called by other macros are evaluated by `m4' several times; each evaluation might require another layer of quotes to prevent unwanted expansions of macros or `m4' builtins, such as `define' and `$1'. Quotes are also required around macro arguments that contain commas, since commas separate the arguments from each other. Autoconf (in `acgeneral.m4') changes the `m4' quote characters from the default ``' and `'' to `[' and `]', because many of the macros use ``' and `'', mismatched. However, in a few places the macros need to use brackets. In those places, they use the `m4' builtin command `changequote' to temporarily disable quoting before the code that uses brackets, like this: changequote(,)dnl Then they turn quoting back on again with another call to `changequote': changequote([,])dnl When you create a `configure' script using newly written macros, examine it carefully to check whether you need to add more quotes in your macros. If one or more words have disappeared in the `m4' output, you need more quotes. When in doubt, quote. However, it's also possible to put on too many layers of quotes. If this happens, the resulting `configure' script will contain unexpanded macros. The `autoconf' program checks for this problem by doing `grep AC_ configure'.  File: autoconf.info, Node: Dependencies Between Macros, Next: Checking for Files, Prev: Quoting, Up: Writing Macros Dependencies Between Macros =========================== Some Autoconf macros depend on other macros having been called first in order to work correctly. Autoconf provides a way to ensure that certain macros are called if needed and a way to warn the user if macros are called in an order that might cause incorrect operation. * Menu: * Prerequisite Macros:: Ensuring required information. * Suggested Ordering:: Warning about possible ordering problems.  File: autoconf.info, Node: Prerequisite Macros, Next: Suggested Ordering, Up: Dependencies Between Macros Prerequisite Macros ------------------- A macro that you write might need to use values that have previously been computed by other macros. For example, if you write a new macro that uses the C preprocessor, it depends on `AC_PROG_CPP' having been called first to set the shell variable `CPP' (*note Alternative Programs::.). Rather than forcing the user of the macros to keep track of all of the dependencies between them, you can use the macros `AC_PROVIDE' and `AC_REQUIRE' to do it automatically. *Note Macro Ordering::, for more information on their syntax. The new macro that runs the C preprocessor should contain, somewhere before `CPP' is used, the statement AC_REQUIRE([AC_PROG_CPP]) and the macro `AC_PROG_CPP' should contain the statement (anywhere in its body) AC_PROVIDE([$0]) Then, when the new macro is run, it will invoke `AC_PROG_CPP' if and only if `AC_PROG_CPP' has not already been run.  File: autoconf.info, Node: Suggested Ordering, Prev: Prerequisite Macros, Up: Dependencies Between Macros Suggested Ordering ------------------ Some macros should be run before another macro if both are called, but neither requires the other to be called. For example, a macro like `AC_AIX' that changes the behavior of the C compiler (*note UNIX Variants::.) should be called before any macros that run the C compiler. Many of these dependencies are noted in the documentation. Autoconf provides a way to warn users when macros with this kind of dependency appear out of order in a `configure.in' file. The warning occurs when creating `configure' from `configure.in', not when running `configure'. It is not a fatal error; `configure' is created as usual. The `AC_BEFORE' macro causes `m4' to print a warning message on the standard error output when a macro is used before another macro which might change its behavior. The macro which should come first should contain a call to `AC_BEFORE' and the macro which should come later should contain a call to `AC_PROVIDE'. For example, `AC_AIX' contains AC_BEFORE([$0], [AC_COMPILE_CHECK]) and `AC_COMPILE_CHECK' contains AC_PROVIDE([$0]) As a result, if `AC_AIX' is called after `AC_COMPILE_CHECK', it will note that `AC_COMPILE_CHECK' has already been called and print a warning message.  File: autoconf.info, Node: Checking for Files, Next: Checking for Symbols, Prev: Dependencies Between Macros, Up: Writing Macros Checking for Files ================== If you need to check whether a file other than a C header file exists, use `test -f FILENAME'. If you need to make multiple checks using `test', combine them with the shell operators `&&' and `||' instead of using the `test' operators `-a' and `-o'. On System V, the precedence of `-a' and `-o' is wrong relative to the unary operators; consequently, POSIX does not specify them, so using them is nonportable. If you combine `&&' and `||' in the same statement, keep in mind that they have equal precedence. Do not use `test -x', because 4.3BSD does not have it. Use `test -f' or `test -r' instead.  File: autoconf.info, Node: Checking for Symbols, Next: Test Programs, Prev: Checking for Files, Up: Writing Macros Checking for Symbols ==================== If you need to check whether a symbol is defined in a C header file, you can use `AC_HEADER_EGREP' if the symbol is not a C preprocessor macro (*note General Tests::.), or compile a small test program that includes the file and references the symbol (*note Test Programs::.). Don't directly `grep' for the symbol in the file, because on some systems it might be defined in another header file that the file you are checking `#include's. However, if you need to check for a particular UNIX variant which is distinguished by having certain text in a certain file, then use `grep' (or `egrep'). But don't use `grep -s' to suppress output, because `grep -s' on System V does not suppress output, only error messages. Instead, redirect the standard output and standard error (in case the file doesn't exist) of `grep' to `/dev/null'. Check the exit status of `grep' to determine whether it found a match. To check whether the Autoconf macros have already defined a certain C preprocessor symbol, you can use a `case' statement like this: case "$DEFS" in *HAVE_FOO*) ;; *) LIBOBJS="$LIBOBJS foo.o" ;; esac Make sure to enclose the variable name you are checking (usually `DEFS') in double quotes, because otherwise some old versions of `bash' misinterpret the statement.  File: autoconf.info, Node: Test Programs, Next: Multiple Cases, Prev: Checking for Symbols, Up: Writing Macros Test Programs ============= Autoconf checks for many features by compiling small test programs. To find out whether a library function is available, Autoconf tries to compile a small program that uses it. This is unlike Larry Wall's Metaconfig, which uses `nm' or `ar' on the C library to try to figure out which functions are available. Trying to link with the function is usually a more reliable and flexible approach because it avoids dealing with the variations in the options and output formats of `nm' and `ar' and in the location of the standard libraries. It also allows `configure' to check aspects of the function's runtime behavior if needed. On the other hand, it is sometimes slower than scanning the libraries. If you need to check for a condition other than whether some symbol exists on the system or has a certain value, then you can't use `AC_COMPILE_CHECK' (*note General Tests::.). You have to write a test program by hand. You can compile and run it using `AC_TEST_PROGRAM' (*note General Tests::.). Try to avoid writing test programs if possible, because using them prevents people from configuring your package for cross-compiling. If it's really best that you test for a run-time behavior, try to provide a default "worst case" value to use when cross-compiling makes run-time tests impossible. You do this by passing the optional last argument to `AC_TEST_PROGRAM'. * Menu: * Guidelines:: General rules for writing test programs. * Tricks:: Special ways to work around problems.  File: autoconf.info, Node: Guidelines, Next: Tricks, Up: Test Programs Guidelines for Test Programs ---------------------------- Test programs should return 0 if the test succeeds, nonzero otherwise, so that success can be distinguished easily from a core dump or other failure; segmentation violations and other failures produce a nonzero exit status. Test programs should `exit', not `return', from `main', because on some systems the argument to `return' in `main' is ignored. They should not write anything to the standard output. Test programs can use `#if' or `#ifdef' to check the values of preprocessor macros defined by tests that have already run. For example, if you call `AC_STDC_HEADERS', then later on in `configure.in' you can have a test program that includes an ANSI C header file conditionally: #if STDC_HEADERS #include #endif If a test program needs to use or create a data file, give it a name that starts with `conftest', such as `conftestdata'. The `configure' script cleans up by running `rm -rf conftest*' after running test programs and if the script is interrupted.  File: autoconf.info, Node: Tricks, Prev: Guidelines, Up: Test Programs Tricks for Test Programs ------------------------ If a test program calls a function with invalid parameters (just to see whether it exists), organize the program to ensure that it never invokes that function. You can do this by calling it in another function that is never invoked. You can't do it by putting it after a call to `exit', because GCC version 2 knows that `exit' never returns and optimizes out any code that follows it in the same block. If you include any header files, make sure to call the functions relevant to them with the correct number of arguments, even if they are just 0, to avoid compilation errors due to prototypes. GCC version 2 has internal prototypes for several functions that it automatically inlines; for example, `memcpy'. To avoid errors when checking for them, either pass them the correct number of arguments or redeclare them with a different return type (such as `char').  File: autoconf.info, Node: Multiple Cases, Prev: Test Programs, Up: Writing Macros Multiple Cases ============== Some operations are accomplished in several possible ways, depending on the UNIX variant. Checking for them essentially requires a "case statement". Autoconf does not directly provide one; however, it is easy to simulate by using a shell variable to keep track of whether a way to perform the operation has been found yet. Here is an example excerpted from the `configure.in' for GNU `find'. It uses the shell variable `fstype' to keep track of whether the remaining cases need to be checked. There are several more cases which are not shown here but follow the same pattern. echo checking how to get filesystem type # SVR4. AC_TEST_CPP([#include #include ], AC_DEFINE(FSTYPE_STATVFS) fstype=1) if test -z "$fstype"; then # SVR3. AC_TEST_CPP([#include #include ], AC_DEFINE(FSTYPE_USG_STATFS) fstype=1) fi if test -z "$fstype"; then # AIX. AC_TEST_CPP([#include #include ], AC_DEFINE(FSTYPE_AIX_STATFS) fstype=1) fi  File: autoconf.info, Node: Makefiles, Next: Running configure Scripts, Prev: Writing Macros, Up: Top Makefiles ********* Each subdirectory in a distribution should come with a file `Makefile.in', from which `configure' will produce a `Makefile' in that directory. Most of the substitutions that `configure' does are simple: for each configuration variable that the package uses, it just replaces occurrences of `@VARIABLE@' with the value that `configure' has determined for that variable. Any occurrences of `@VARIABLE@' for variables that `configure' does not know about are passed through unchanged. There is no point in checking for the correct value to give a variable that is never used. Every variable that the `configure' script might set a value for should appear in a `@VARIABLE@' reference in at least one `Makefile.in'. If `AC_CONFIG_HEADER' is called, `configure' replaces `@DEFS@' with `-DHAVE_CONFIG_H', since the contents of `DEFS' would be redundant. *Note Makefile Conventions: (standards.info)Makefiles, for more information on what to put in Makefiles. *Note Sample Makefile.in::, for an example of a real `Makefile.in'. * Menu: * Predefined Variables:: Heavily used `make' variables. * Installation Prefixes:: A special variable substitution. * VPATH Substitutions:: Compiling in a different directory. * Automatic Remaking:: Makefile rules for configuring.  File: autoconf.info, Node: Predefined Variables, Next: Installation Prefixes, Up: Makefiles Predefined Variables ==================== Some `make' variables are predefined by the Autoconf macros. `AC_SUBST' is called for them automatically (*note Setting Variables::.), so in your `Makefile.in' files you can get their values by enclosing their names in `@' characters. *Note Makefiles::, for more information on `@' substitutions. The variables that are defined by the general purpose Autoconf macros are: - Variable: srcdir The directory that contains the source code for that `Makefile'. - Variable: DEFS `-D' options to pass to the C compiler. If `AC_CONFIG_HEADER' is called, `configure' replaces `@DEFS@' with `-DHAVE_CONFIG_H', since the contents of `DEFS' would be redundant. - Variable: LIBS `-l' and `-L' options to pass to the linker. - Variable: LIBOBJS Names of object files (ending in `.o'). Set by `AC_REPLACE_FUNCS' (*note General Tests::.).  File: autoconf.info, Node: Installation Prefixes, Next: VPATH Substitutions, Prev: Predefined Variables, Up: Makefiles Installation Prefixes ===================== If `configure' has figured out a value for the installation prefix, either by the user supplying one on the command line (*note Running configure Scripts::.) or with `AC_PREFIX', then it substitutes that value in `Makefile's that it creates. Wherever a `Makefile.in' contains lines like prefix = /usr/local exec_prefix = ${prefix} `configure' substitutes the value it figured out. The word `prefix' or `exec_prefix' must not be preceded by any other characters on the line. If `configure' has not figured out a value for the prefix, the value in the `Makefile.in' is left alone. There can be separate installation prefixes for architecture-specific files (`exec_prefix') and architecture-independent files (`prefix'). *Note Running configure Scripts::, for more information on setting them. Autoconf `configure' scripts replace these two variables without requiring them to be enclosed in `@' characters, and only if they have been set, because the Cygnus `configure' does so. In retrospect, being compatible in this way was a bad decision, because it created an inconsistency in Autoconf without giving significant benefits.  File: autoconf.info, Node: VPATH Substitutions, Next: Automatic Remaking, Prev: Installation Prefixes, Up: Makefiles `VPATH' Substitutions ===================== You might want to compile a software package in a different directory from the one that contains the source code. Doing this allows you to compile the package for several architectures simultaneously from the same copy of the source code and keep multiple sets of object files on disk. To support doing this, `make' uses the `VPATH' variable to find the files that are in the source directory. GNU `make' and most other recent `make' programs can do this. Older `make' programs do not support `VPATH'; when using them, the source code must be in the same directory as the object files. To support `VPATH', each `Makefile.in' should contain two lines that look like: srcdir = @srcdir@ VPATH = @srcdir@ Do not set `VPATH' to the value of another variable, for example `VPATH = $(srcdir)', because some versions of `make' do not do variable substitutions on the value of `VPATH'. `configure' substitutes in the correct value for `srcdir' when it produces `Makefile.in'. Do not use the `make' variable `$<', which expands to the pathname of the file in the source directory (found with `VPATH'), except in implicit rules. (An implicit rule is one such as `.c.o', which tells how to create a `.o' file from a `.c' file.) Some versions of `make' do not set `$<' in explicit rules; they expand it to an empty value. Instead, `Makefile' command lines should always refer to source files by prefixing them with `$(srcdir)/'. For example: time.info: time.texinfo $(MAKEINFO) $(srcdir)/time.texinfo  File: autoconf.info, Node: Automatic Remaking, Prev: VPATH Substitutions, Up: Makefiles Automatic Remaking ================== You can put rules like the following in the top-level `Makefile.in' for a package to automatically update the configuration information when you change the configuration files. # The next rule also takes care of making config.h from config.h.in. # If remaking config.h does not change it, its timestamp is untouched. Makefile: Makefile.in config.status $(SHELL) config.status config.status: configure $(SHELL) config.status --recheck configure: configure.in cd $(srcdir); autoconf config.h.in: configure.in cd $(srcdir); autoheader  File: autoconf.info, Node: Running configure Scripts, Next: Example, Prev: Makefiles, Up: Top Running `configure' Scripts *************************** A software package that uses a `configure' script should be distributed with a file `Makefile.in', but no `Makefile'; that way, the user has to properly configure the package for the local system before compiling it. Here is how to configure a package that uses a `configure' script. Normally, you just `cd' to the directory containing the package's source code and type `./configure'. If you're using `csh' on an old version of System V, you might need to type `sh configure' instead to prevent `csh' from trying to execute `configure' itself. Running `configure' takes a minute or two. While it is running, it prints some messages that tell what it is doing. If you don't want to see the messages, run `configure' with its standard output redirected to `/dev/null'; for example, `./configure >/dev/null'. To compile the package in a different directory from the one containing the source code, you must use a version of `make' that supports the `VPATH' variable, such as GNU `make'. `cd' to the directory where you want the object files and executables to go and run the `configure' script. `configure' automatically checks for the source code in the directory that `configure' is in and in `..'. If for some reason `configure' is not in the source code directory that you are configuring, then it will report that it can't find the source code. In that case, run `configure' with the option `--srcdir=DIR', where DIR is the directory that contains the source code. By default, `make install' will install the package's files in `/usr/local/bin', `/usr/local/man', etc. You can specify an installation prefix other than `/usr/local' by giving `configure' the option `--prefix=PATH'. Alternately, you can do so by consistently giving a value for the `prefix' variable when you run `make', e.g., make prefix=/usr/gnu make prefix=/usr/gnu install You can specify separate installation prefixes for architecture-specific files and architecture-independent files. If you give `configure' the option `--exec-prefix=PATH' or set the `make' variable `exec_prefix' to PATH, the package will use PATH as the prefix for installing programs and libraries. Data files and documentation will still use the regular prefix. Normally, all files are installed using the same prefix. Some packages pay attention to `--with-PACKAGE' options to `configure', where PACKAGE is something like `gnu-as' or `x' (for the X Window System). The README should mention any `--with-' options that the package recognizes. `configure' ignores any other arguments that you give it. * Menu: * Overriding variables:: Workarounds for unusual systems. * Running config.status:: Recreating a configuration.  File: autoconf.info, Node: Overriding variables, Next: Running config.status, Up: Running configure Scripts Overriding variables ==================== On systems that require unusual options for compilation or linking that the package's `configure' script does not know about, you can give `configure' initial values for variables by setting them in the environment. In Bourne-compatible shells, you can do that on the command line like this: CC='gcc -traditional' LIBS=-lposix ./configure Here are the `make' variables that you might want to override with environment variables when running `configure'. For these variables, any value given in the environment overrides the value that `configure' would choose: - Variable: CC C compiler program. The default is `cc'. - Variable: INSTALL Program to use to install files. The default is `install' if you have it, `cp' otherwise. For these variables, any value given in the environment is added to the value that `configure' chooses: - Variable: DEFS Configuration options, in the form `-Dfoo -Dbar...'. Do not use this variable in packages that create a configuration header file. - Variable: LIBS Libraries to link with, in the form `-lfoo -lbar...'. In the long term, most problems requiring manual intervention should be fixed by updating either the Autoconf macros or the `configure.in' file for that package. *Note Making configure Scripts::, for a discussion of that subject.  File: autoconf.info, Node: Running config.status, Prev: Overriding variables, Up: Running configure Scripts Recreating a Configuration ========================== The `configure' script creates a file named `config.status' which describes which configuration options were specified when the package was last configured. This file is a shell script which, if run, will recreate the same configuration. You can give `config.status' the `--recheck' option, which makes it re-run `configure' with the same arguments you used before. This option is useful if you change `configure', so that the results of some tests might be different from the previous run. `config.status' checks several optional environment variables that can alter its behavior: - Variable: CONFIG_SHELL The shell with which to run `configure' for the `--recheck' option. The default is `/bin/sh'. The following two variables provide one way for separately distributed packages to share the values computed by `configure'. Doing so can be useful if some of the packages need a superset of the features that one of them, perhaps a common library, does. These variables allow a `config.status' file to create files other than the ones that its `configure.in' specifies, so it can be used for a different package. - Variable: CONFIG_FILES The files in which to perform `@VARIABLE@' substitutions. The default is the arguments given to `AC_OUTPUT' in `configure.in'. - Variable: CONFIG_HEADERS The files in which to substitute C `#define' statements. The default is the arguments given to `AC_CONFIG_HEADER'; if that macro was not called, `config.status' ignores this variable.  File: autoconf.info, Node: Example, Next: Preprocessor Symbol Index, Prev: Running configure Scripts, Up: Top An Example ********** Here are sample `configure.in' and `Makefile.in' files, to give a real illustration of using Autoconf. They are from the GNU `cpio' package, which also includes the `mt' and `rmt' programs. This package does not use a configuration header file; it passes `-D' options to the C compiler on the command line. * Menu: * Sample configure.in:: An example of a `configure' template. * Sample Makefile.in:: An example of a `Makefile' template.  File: autoconf.info, Node: Sample configure.in, Next: Sample Makefile.in, Up: Example Sample `configure.in' ===================== Here is `configure.in' from GNU `cpio'. The `dnl' macro after `AC_SUBST' is suppresses an extra (though harmless) newline in the generated `configure' script (because the `AC_SUBST' macro does not produce any output where it is called). dnl Process this file with autoconf to produce a configure script. AC_INIT(cpio.h) PROGS="cpio" AC_SUBST(PROGS)dnl AC_PROG_CC AC_PROG_CPP AC_GCC_TRADITIONAL AC_PROG_INSTALL AC_AIX AC_MINIX AC_ISC_POSIX AC_RETSIGTYPE AC_MAJOR_HEADER AC_REMOTE_TAPE test -n "$have_mtio" && PROGS="$PROGS mt" AC_RSH AC_CONST AC_UID_T AC_STDC_HEADERS AC_UNISTD_H AC_HAVE_HEADERS(string.h fcntl.h utime.h sys/io/trioctl.h) AC_REPLACE_FUNCS(fnmatch bcopy mkdir strdup) AC_HAVE_FUNCS(strerror lchown) AC_VPRINTF AC_ALLOCA AC_XENIX_DIR AC_HAVE_LIBRARY(socket, [LIBS="$LIBS -lsocket"]) AC_HAVE_LIBRARY(nsl, [LIBS="$LIBS -lnsl"]) AC_OUTPUT(Makefile)  File: autoconf.info, Node: Sample Makefile.in, Prev: Sample configure.in, Up: Example Sample `Makefile.in' ==================== Here is `Makefile.in' from GNU `cpio', with some irrelevant lines omitted, for brevity. srcdir = @srcdir@ VPATH = @srcdir@ CC = @CC@ INSTALL = @INSTALL@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_DATA = @INSTALL_DATA@ DEFS = @DEFS@ LIBS = @LIBS@ RTAPELIB = @RTAPELIB@ CFLAGS = -g LDFLAGS = -g prefix = /usr/local exec_prefix = $(prefix) binprefix = manprefix = bindir = $(exec_prefix)/bin libdir = $(exec_prefix)/lib mandir = $(prefix)/man/man1 manext = 1 SHELL = /bin/sh SRCS = copyin.c copyout.c copypass.c defer.c dstring.c global.c \ main.c tar.c util.c error.c getopt.c getopt1.c filemode.c version.c \ rtapelib.c dirname.c idcache.c makepath.c xmalloc.c stripslash.c \ userspec.c xstrdup.c bcopy.c fnmatch.c mkdir.c strdup.c OBJS = copyin.o copyout.o copypass.o defer.o dstring.o global.o \ main.o tar.o util.o error.o getopt.o getopt1.o filemode.o version.o \ $(RTAPELIB) dirname.o idcache.o makepath.o xmalloc.o stripslash.o \ userspec.o xstrdup.o @LIBOBJS@ @ALLOCA@ # mt source files not shared with cpio. MT_SRCS = mt.c argmatch.c MT_OBJS = mt.o argmatch.o error.o getopt.o getopt1.o \ xmalloc.o version.o $(RTAPELIB) @ALLOCA@ HDRS = cpio.h cpiohdr.h tar.h tarhdr.h defer.h dstring.h extern.h filetypes.h \ system.h fnmatch.h getopt.h rmt.h DISTFILES = $(SRCS) $(HDRS) COPYING COPYING.LIB ChangeLog Makefile.in \ README NEWS INSTALL cpio.1 mt.1 makefile.pc makefile.os2 cpio.def \ configure configure.in mkinstalldirs $(MT_SRCS) rmt.c tcexparg.c alloca.c all: @PROGS@ .c.o: $(CC) -c $(CPPFLAGS) $(DEFS) -I$(srcdir) $(CFLAGS) $< install: installdirs all $(srcdir)/cpio.1 $(srcdir)/mt.1 $(INSTALL_PROGRAM) cpio $(bindir)/$(binprefix)cpio test ! -f mt || $(INSTALL_PROGRAM) mt $(bindir)/$(binprefix)mt -test ! -f rmt || $(INSTALL_PROGRAM) rmt $(libdir)/rmt $(INSTALL_DATA) $(srcdir)/cpio.1 $(mandir)/$(manprefix)cpio.$(manext) test ! -f mt || \ $(INSTALL_DATA) $(srcdir)/mt.1 $(mandir)/$(manprefix)mt.$(manext) installdirs: $(srcdir)/mkinstalldirs $(bindir) $(libdir) $(mandir) uninstall: cd $(bindir); rm -f $(binprefix)cpio $(binprefix)mt -rm -f $(libdir)/rmt cd $(mandir); rm -f $(manprefix)cpio.$(manext) $(manprefix)mt.$(manext) check: @echo No tests are supplied. cpio: $(OBJS) $(CC) $(LDFLAGS) -o $@ $(OBJS) $(LIBS) rmt: rmt.o $(CC) $(LDFLAGS) -o $@ rmt.o $(LIBS) mt: $(MT_OBJS) $(CC) $(LDFLAGS) -o $@ $(MT_OBJS) $(LIBS) Makefile: Makefile.in config.status $(SHELL) config.status config.status: configure $(SHELL) config.status --recheck configure: configure.in cd $(srcdir); autoconf TAGS: $(SRCS) etags $(SRCS) clean: rm -f cpio rmt mt *.o core mostlyclean: clean distclean: clean rm -f Makefile config.status realclean: distclean rm -f TAGS dist: $(DISTFILES) echo cpio-`sed -e '/version_string/!d' \ -e 's/[^0-9.]*\([0-9.]*\).*/\1/' -e q version.c` > .fname rm -rf `cat .fname` mkdir `cat .fname` -ln $(DISTFILES) `cat .fname` for file in $(DISTFILES); do \ test -r `cat .fname`/$$file || cp -p $$file `cat .fname`; \ done tar chzf `cat .fname`.tar.gz `cat .fname` rm -rf `cat .fname` .fname  File: autoconf.info, Node: Preprocessor Symbol Index, Next: Macro Index, Prev: Example, Up: Top Preprocessor Symbol Index ************************* This is an alphabetical list of the C preprocessor symbols that the Autoconf macros define. To work with Autoconf, C source code needs to use these names in `#if' directives. * Menu: * CC: Overriding variables. * CONFIG_FILES: Running config.status. * CONFIG_HEADERS: Running config.status. * CONFIG_SHELL: Running config.status. * const: Compiler Characteristics. * C_ALLOCA: Library Functions. * DECLARE_YYTEXT: Alternative Programs. * DEFS: Overriding variables. * DEFS: Predefined Variables. * DGUX: Library Functions. * DIRENT: Header Files. * GETGROUPS_T: Typedefs. * GETLODAVG_PRIVILEGED: Library Functions. * gid_t: Typedefs. * HAVE_FUNCTION: General Tests. * HAVE_HEADER: General Tests. * HAVE_ALLOCA_H: Library Functions. * HAVE_CONFIG_H: Setup. * HAVE_DOPRNT: Library Functions. * HAVE_LONG_DOUBLE: Compiler Characteristics. * HAVE_LONG_FILE_NAMES: System Services. * HAVE_NETDB_H: Alternative Programs. * HAVE_RESTARTABLE_SYSCALLS: System Services. * HAVE_STRCOLL: Library Functions. * HAVE_ST_BLKSIZE: Structures. * HAVE_ST_BLOCKS: Structures. * HAVE_ST_RDEV: Structures. * HAVE_SYS_MTIO_H: System Services. * HAVE_TM_ZONE: Structures. * HAVE_TZNAME: Structures. * HAVE_UNISTD_H: Header Files. * HAVE_UTIME_NULL: Library Functions. * HAVE_VFORK_H: Library Functions. * HAVE_VPRINTF: Library Functions. * HAVE_WAIT3: Library Functions. * inline: Compiler Characteristics. * INSTALL: Overriding variables. * INT_16_BITS: Compiler Characteristics. * LIBOBJS: Predefined Variables. * LIBS: Predefined Variables. * LIBS: Overriding variables. * LONG_64_BITS: Compiler Characteristics. * MAJOR_IN_MKDEV: Header Files. * MAJOR_IN_SYSMACROS: Header Files. * mode_t: Typedefs. * NDIR: Header Files. * NEED_MEMORY_H: Header Files. * NEED_SETGID: Library Functions. * NLIST_NAME_UNION: Library Functions. * NLIST_STRUCT: Library Functions. * NO_ARG_ARRAY: Compiler Characteristics. * NO_MINUS_C_MINUS_O: Alternative Programs. * NO_REMOTE: Alternative Programs. * off_t: Typedefs. * pid_t: Typedefs. * RETSIGTYPE: Typedefs. * SETVBUF_REVERSED: Library Functions. * size_t: Typedefs. * srcdir: Predefined Variables. * STDC_HEADERS: Header Files. * SVR4: Library Functions. * SYSDIR: Header Files. * SYSNDIR: Header Files. * TIME_WITH_SYS_TIME: Structures. * TM_IN_SYS_TIME: Structures. * uid_t: Typedefs. * UMAX: Library Functions. * UMAX4_3: Library Functions. * USG: Header Files. * vfork: Library Functions. * VOID_CLOSEDIR: Header Files. * VOID_CLOSEDIR: UNIX Variants. * WORDS_BIGENDIAN: Compiler Characteristics. * _ALL_SOURCE: UNIX Variants. * _MINIX: UNIX Variants. * _POSIX_1_SOURCE: UNIX Variants. * _POSIX_SOURCE: UNIX Variants. * _POSIX_SOURCE: UNIX Variants. * _POSIX_VERSION: Header Files. * __CHAR_UNSIGNED__: Compiler Characteristics.  File: autoconf.info, Node: Macro Index, Prev: Preprocessor Symbol Index, Up: Top Macro Index *********** This is an alphabetical list of the Autoconf macros. To make the list easier to use, the macros are listed without their preceding `AC_'. * Menu: * AIX: UNIX Variants. * ALLOCA: Library Functions. * ARG_ARRAY: Compiler Characteristics. * BEFORE: Macro Ordering. * CHAR_UNSIGNED: Compiler Characteristics. * COMPILE_CHECK: General Tests. * CONFIG_HEADER: Setup. * CONST: Compiler Characteristics. * CROSS_CHECK: Compiler Characteristics. * DECLARE_YYTEXT: Alternative Programs. * DEFINE: Setting Variables. * DEFINE_UNQUOTED: Setting Variables. * DIR_HEADER: Header Files. * DYNIX_SEQ: UNIX Variants. * FIND_X: System Services. * FUNC_CHECK: General Tests. * GCC_TRADITIONAL: Alternative Programs. * GETGROUPS_T: Typedefs. * GETLOADAVG: Library Functions. * HAVE_FUNCS: General Tests. * HAVE_HEADERS: General Tests. * HAVE_LIBRARY: General Tests. * HAVE_LONG_DOUBLE: Compiler Characteristics. * HAVE_POUNDBANG: System Services. * HEADER_CHECK: General Tests. * HEADER_EGREP: General Tests. * INIT: Setup. * INLINE: Compiler Characteristics. * INT_16_BITS: Compiler Characteristics. * IRIX_SUN: UNIX Variants. * ISC_POSIX: UNIX Variants. * LN_S: Alternative Programs. * LONG_64_BITS: Compiler Characteristics. * LONG_FILE_NAMES: System Services. * MAJOR_HEADER: Header Files. * MEMORY_H: Header Files. * MINIX: UNIX Variants. * MINUS_C_MINUS_O: Alternative Programs. * MODE_T: Typedefs. * OBSOLETE: Macro Ordering. * OFF_T: Typedefs. * OUTPUT: Setup. * PID_T: Typedefs. * PREFIX: General Tests. * PREPARE: Setup. * PROGRAMS_CHECK: General Tests. * PROGRAM_CHECK: General Tests. * PROGRAM_EGREP: General Tests. * PROG_AWK: Alternative Programs. * PROG_CC: Alternative Programs. * PROG_CPP: Alternative Programs. * PROG_INSTALL: Alternative Programs. * PROG_LEX: Alternative Programs. * PROG_RANLIB: Alternative Programs. * PROG_YACC: Alternative Programs. * PROVIDE: Macro Ordering. * REMOTE_TAPE: System Services. * REPLACE_FUNCS: General Tests. * REQUIRE: Macro Ordering. * RESTARTABLE_SYSCALLS: System Services. * RETSIGTYPE: Typedefs. * REVISION: Setup. * RSH: Alternative Programs. * SCO_INTL: UNIX Variants. * SETVBUF_REVERSED: Library Functions. * SIZE_T: Typedefs. * STAT_MACROS_BROKEN: Structures. * STAT_MACROS_BROKEN: Structures. * STDC_HEADERS: Header Files. * STRCOLL: Library Functions. * STRUCT_TM: Structures. * ST_BLKSIZE: Structures. * ST_BLOCKS: Structures. * ST_RDEV: Structures. * SUBST: Setting Variables. * TEST_CPP: General Tests. * TEST_PROGRAM: General Tests. * TIMEZONE: Structures. * TIME_WITH_SYS_TIME: Structures. * UID_T: Typedefs. * UNISTD_H: Header Files. * USG: Header Files. * UTIME_NULL: Library Functions. * VFORK: Library Functions. * VPRINTF: Library Functions. * WAIT3: Library Functions. * WITH: General Tests. * WORDS_BIGENDIAN: Compiler Characteristics. * XENIX_DIR: UNIX Variants.  Tag Table: Node: Top964 Node: Introduction4130 Node: Distributing7890 Node: Making configure Scripts8910 Node: Writing configure.in10858 Node: Invoking autoconf13283 Node: Invoking autoheader15132 Node: Specific Tests17150 Node: Alternative Programs18814 Node: Header Files21907 Node: Typedefs27143 Node: Library Functions28360 Node: Structures32925 Node: Compiler Characteristics35015 Node: System Services37564 Node: UNIX Variants39335 Node: General Purpose Macros41243 Node: Setup42169 Node: General Tests46404 Node: Setting Variables55160 Node: Macro Ordering58428 Node: Writing Macros61176 Node: Macro Format62672 Node: Quoting64044 Node: Dependencies Between Macros65502 Node: Prerequisite Macros66089 Node: Suggested Ordering67138 Node: Checking for Files68517 Node: Checking for Symbols69303 Node: Test Programs70773 Node: Guidelines72422 Node: Tricks73565 Node: Multiple Cases74568 Node: Makefiles75772 Node: Predefined Variables77180 Node: Installation Prefixes78195 Node: VPATH Substitutions79519 Node: Automatic Remaking81240 Node: Running configure Scripts81992 Node: Overriding variables84890 Node: Running config.status86396 Node: Example88105 Node: Sample configure.in88692 Node: Sample Makefile.in89837 Node: Preprocessor Symbol Index93822 Node: Macro Index98761  End Tag Table autoconf-1.7/standards.texi100666 21570 1 176552 5461536574 14426 0ustar djmdaemon\input texinfo @c -*-texinfo-*- @c %**start of header @setfilename standards.info @settitle GNU Coding Standards @c %**end of header @setchapternewpage off @ifinfo Copyright (C) 1992, 1993 Free Software Foundation Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. @ignore Permission is granted to process this file through TeX and print the results, provided the printed document carries copying permission notice identical to this one except for the removal of this paragraph (this paragraph not being relevant to the printed manual). @end ignore Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Free Software Foundation. @end ifinfo @titlepage @title GNU Coding Standards @author Richard Stallman @author last updated 9 September 1993 @c Note date also appears below. @page @vskip 0pt plus 1filll Copyright @copyright{} 1992, 1993 Free Software Foundation Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by Free Software Foundation. @end titlepage @ifinfo @node Top, Reading Non-Free Code, (dir), (dir) @top Version Last updated 9 September 1993. @c Note date also appears above. @end ifinfo @menu * Reading Non-Free Code:: Referring to Proprietary Programs * Contributions:: Accepting Contributions * Change Logs:: Recording Changes * Compatibility:: Compatibility with Other Implementations * Makefile Conventions:: Makefile Conventions * Configuration:: How Configuration Should Work * Source Language:: Using Languages Other Than C * Formatting:: Formatting Your Source Code * Comments:: Commenting Your Work * Syntactic Conventions:: Clean Use of C Constructs * Names:: Naming Variables and Functions * Using Extensions:: Using Non-standard Features * System Functions:: Portability and ``standard'' library functions * Semantics:: Program Behavior for All Programs * Errors:: Formatting Error Messages * Libraries:: Library Behavior * Portability:: Portability As It Applies to GNU * User Interfaces:: Standards for Command Line Interfaces * Documentation:: Documenting Programs * Releases:: Making Releases @end menu @node Reading Non-Free Code @chapter Referring to Proprietary Programs Don't in any circumstances refer to Unix source code for or during your work on GNU! (Or to any other proprietary programs.) If you have a vague recollection of the internals of a Unix program, this does not absolutely mean you can't write an imitation of it, but do try to organize the imitation internally along different lines, because this is likely to make the details of the Unix version irrelevant and dissimilar to your results. For example, Unix utilities were generally optimized to minimize memory use; if you go for speed instead, your program will be very different. You could keep the entire input file in core and scan it there instead of using stdio. Use a smarter algorithm discovered more recently than the Unix program. Eliminate use of temporary files. Do it in one pass instead of two (we did this in the assembler). Or, on the contrary, emphasize simplicity instead of speed. For some applications, the speed of today's computers makes simpler algorithms adequate. Or go for generality. For example, Unix programs often have static tables or fixed-size strings, which make for arbitrary limits; use dynamic allocation instead. Make sure your program handles NULs and other funny characters in the input files. Add a programming language for extensibility and write part of the program in that language. Or turn some parts of the program into independently usable libraries. Or use a simple garbage collector instead of tracking precisely when to free memory, or use a new GNU facility such as obstacks. @node Contributions @chapter Accepting Contributions If someone else sends you a piece of code to add to the program you are working on, we need legal papers to use it---the same sort of legal papers we will need to get from you. @emph{Each} significant contributor to a program must sign some sort of legal papers in order for us to have clear title to the program. The main author alone is not enough. So, before adding in any contributions from other people, tell us so we can arrange to get the papers. Then wait until we tell you that we have received the signed papers, before you actually use the contribution. This applies both before you release the program and afterward. If you receive diffs to fix a bug, and they make significant change, we need legal papers for it. You don't need papers for changes of a few lines here or there, since they are not significant for copyright purposes. Also, you don't need papers if all you get from the suggestion is some ideas, not actual code which you use. For example, if you write a different solution to the problem, you don't need to get papers. I know this is frustrating; it's frustrating for us as well. But if you don't wait, you are going out on a limb---for example, what if the contributor's employer won't sign a disclaimer? You might have to take that code out again! The very worst thing is if you forget to tell us about the other contributor. We could be very embarrassed in court some day as a result. @node Change Logs @chapter Change Logs Keep a change log for each directory, describing the changes made to source files in that directory. The purpose of this is so that people investigating bugs in the future will know about the changes that might have introduced the bug. Often a new bug can be found by looking at what was recently changed. More importantly, change logs can help eliminate conceptual inconsistencies between different parts of a program; they can give you a history of how the conflicting concepts arose. Use the Emacs command @kbd{M-x add-change} to start a new entry in the change log. An entry should have an asterisk, the name of the changed file, and then in parentheses the name of the changed functions, variables or whatever, followed by a colon. Then describe the changes you made to that function or variable. Separate unrelated entries with blank lines. When two entries represent parts of the same change, so that they work together, then don't put blank lines between them. Then you can omit the file name and the asterisk when successive entries are in the same file. Here are some examples: @example * register.el (insert-register): Return nil. (jump-to-register): Likewise. * sort.el (sort-subr): Return nil. * tex-mode.el (tex-bibtex-file, tex-file, tex-region): Restart the tex shell if process is gone or stopped. (tex-shell-running): New function. * expr.c (store_one_arg): Round size up for move_block_to_reg. (expand_call): Round up when emitting USE insns. * stmt.c (assign_parms): Round size up for move_block_from_reg. @end example It's important to name the changed function or variable in full. Don't abbreviate them; don't combine them. Subsequent maintainers will often search for a function name to find all the change log entries that pertain to it; if you abbreviate the name, they won't find it when they search. For example, some people are tempted to abbreviate groups of function names by writing @samp{* register.el (@{insert,jump-to@}-register)}; this is not a good idea, since searching for @code{jump-to-register} or @code{insert-register} would not find the entry. There's no need to describe the full purpose of the changes or how they work together. It is better to put such explanations in comments in the code. That's why just ``New function'' is enough; there is a comment with the function in the source to explain what it does. However, sometimes it is useful to write one line to describe the overall purpose of a large batch of changes. You can think of the change log as a conceptual ``undo list'' which explains how earlier versions were different from the current version. People can see the current version; they don't need the change log to tell them what is in it. What they want from a change log is a clear explanation of how the earlier version differed. When you change the calling sequence of a function in a simple fashion, and you change all the callers of the function, there is no need to make individual entries for all the callers. Just write in the entry for the function being called, ``All callers changed.'' When you change just comments or doc strings, it is enough to write an entry for the file, without mentioning the functions. Write just, ``Doc fix.'' There's no need to keep a change log for documentation files. This is because documentation is not susceptible to bugs that are hard to fix. Documentation does not consist of parts that must interact in a precisely engineered fashion; to correct an error, you need not know the history of the erroneous passage. @node Compatibility @chapter Compatibility with Other Implementations With certain exceptions, utility programs and libraries for GNU should be upward compatible with those in Berkeley Unix, and upward compatible with @sc{ANSI} C if @sc{ANSI} C specifies their behavior, and upward compatible with @sc{POSIX} if @sc{POSIX} specifies their behavior. When these standards conflict, it is useful to offer compatibility modes for each of them. @sc{ANSI} C and @sc{POSIX} prohibit many kinds of extensions. Feel free to make the extensions anyway, and include a @samp{--ansi} or @samp{--compatible} option to turn them off. However, if the extension has a significant chance of breaking any real programs or scripts, then it is not really upward compatible. Try to redesign its interface. When a feature is used only by users (not by programs or command files), and it is done poorly in Unix, feel free to replace it completely with something totally different and better. (For example, vi is replaced with Emacs.) But it is nice to offer a compatible feature as well. (There is a free vi clone, so we offer it.) Additional useful features not in Berkeley Unix are welcome. Additional programs with no counterpart in Unix may be useful, but our first priority is usually to duplicate what Unix already has. @comment The makefile standards are in a separate file that is also @comment included by make.texinfo. Done by roland@gnu.ai.mit.edu on 1/6/93. @include make-stds.texi @node Configuration @chapter How Configuration Should Work Each GNU distribution should come with a shell script named @code{configure}. This script is given arguments which describe the kind of machine and system you want to compile the program for. The @code{configure} script must record the configuration options so that they affect compilation. One way to do this is to make a link from a standard name such as @file{config.h} to the proper configuration file for the chosen system. If you use this technique, the distribution should @emph{not} contain a file named @file{config.h}. This is so that people won't be able to build the program without configuring it first. Another thing that @code{configure} can do is to edit the Makefile. If you do this, the distribution should @emph{not} contain a file named @file{Makefile}. Instead, include a file @file{Makefile.in} which contains the input used for editing. Once again, this is so that people won't be able to build the program without configuring it first. If @code{configure} does write the @file{Makefile}, then @file{Makefile} should have a target named @file{Makefile} which causes @code{configure} to be rerun, setting up the same configuration that was set up last time. The files that @code{configure} reads should be listed as dependencies of @file{Makefile}. All the files which are output from the @code{configure} script should have comments at the beginning explaining that they were generated automatically using @code{configure}. This is so that users won't think of trying to edit them by hand. The @code{configure} script should write a file named @file{config.status} which describes which configuration options were specified when the program was last configured. This file should be a shell script which, if run, will recreate the same configuration. The @code{configure} script should accept an option of the form @samp{--srcdir=@var{dirname}} to specify the directory where sources are found (if it is not the current directory). This makes it possible to build the program in a separate directory, so that the actual source directory is not modified. If the user does not specify @samp{--srcdir}, then @code{configure} should check both @file{.} and @file{..} to see if it can find the sources. If it finds the sources in one of these places, it should use them from there. Otherwise, it should report that it cannot find the sources, and should exit with nonzero status. Usually the easy way to support @samp{--srcdir} is by editing a definition of @code{VPATH} into the Makefile. Some rules may need to refer explicitly to the specified source directory. To make this possible, @code{configure} can add to the Makefile a variable named @code{srcdir} whose value is precisely the specified directory. The @code{configure} script should also take an argument which specifies the type of system to build the program for. This argument should look like this: @example @var{cpu}-@var{company}-@var{system} @end example For example, a Sun 3 might be @samp{m68k-sun-sunos4.1}. The @code{configure} script needs to be able to decode all plausible alternatives for how to describe a machine. Thus, @samp{sun3-sunos4.1} would be a valid alias. So would @samp{sun3-bsd4.2}, since SunOS is basically @sc{BSD} and no other @sc{BSD} system is used on a Sun. For many programs, @samp{vax-dec-ultrix} would be an alias for @samp{vax-dec-bsd}, simply because the differences between Ultrix and @sc{BSD} are rarely noticeable, but a few programs might need to distinguish them. There is a shell script called @file{config.sub} that you can use as a subroutine to validate system types and canonicalize aliases. Other options are permitted to specify in more detail the software or hardware are present on the machine: @table @samp @item --with-@var{package} The package @var{package} will be installed, so configure this package to work with @var{package}. Possible values of @var{package} include @samp{x}, @samp{gnu-as} (or @samp{gas}), @samp{gnu-ld}, @samp{gnu-libc}, and @samp{gdb}. @item --nfp The target machine has no floating point processor. @item --gas The target machine assembler is GAS, the GNU assembler. This is obsolete; use @samp{--with-gnu-as} instead. @item --x The target machine has the X Window System installed. This is obsolete; use @samp{--with-x} instead. @end table All @code{configure} scripts should accept all of these ``detail'' options, whether or not they make any difference to the particular package at hand. In particular, they should accept any option that starts with @samp{--with-}. This is so users will be able to configure an entire GNU source tree at once with a single set of options. Packages that perform part of compilation may support cross-compilation. In such a case, the host and target machines for the program may be different. The @code{configure} script should normally treat the specified type of system as both the host and the target, thus producing a program which works for the same type of machine that it runs on. The way to build a cross-compiler, cross-assembler, or what have you, is to specify the option @samp{--host=@var{hosttype}} when running @code{configure}. This specifies the host system without changing the type of target system. The syntax for @var{hosttype} is the same as described above. Programs for which cross-operation is not meaningful need not accept the @samp{--host} option, because configuring an entire operating system for cross-operation is not a meaningful thing. Some programs have ways of configuring themselves automatically. If your program is set up to do this, your @code{configure} script can simply ignore most of its arguments. @node Source Language @chapter Using Languages Other Than C Using a language other than C is like using a non-standard feature: it will cause trouble for users. Even if GCC supports the other language, users may find it inconvenient to have to install the compiler for that other language in order to build your program. So please write in C. There are three exceptions for this rule: @itemize @bullet @item It is okay to use a special language if the same program contains an interpreter for that language. Thus, it is not a problem that GNU Emacs contains code written in Emacs Lisp, because it comes with a Lisp interpreter. @item It is okay to use another language in a tool specifically intended for use with that language. This is okay because the only people who want to build the tool will be those who have installed the other language anyway. @item If an application is not of extremely widespread interest, then perhaps it's not important if the application is inconvenient to install. @end itemize @node Formatting @chapter Formatting Your Source Code It is important to put the open-brace that starts the body of a C function in column zero, and avoid putting any other open-brace or open-parenthesis or open-bracket in column zero. Several tools look for open-braces in column zero to find the beginnings of C functions. These tools will not work on code not formatted that way. It is also important for function definitions to start the name of the function in column zero. This helps people to search for function definitions, and may also help certain tools recognize them. Thus, the proper format is this: @example static char * concat (s1, s2) /* Name starts in column zero here */ char *s1, *s2; @{ /* Open brace in column zero here */ @dots{} @} @end example @noindent or, if you want to use @sc{ANSI} C, format the definition like this: @example static char * concat (char *s1, char *s2) @{ @dots{} @} @end example In @sc{ANSI} C, if the arguments don't fit nicely on one line, split it like this: @example int lots_of_args (int an_integer, long a_long, short a_short, double a_double, float a_float) @dots{} @end example For the body of the function, we prefer code formatted like this: @example if (x < foo (y, z)) haha = bar[4] + 5; else @{ while (z) @{ haha += foo (z, z); z--; @} return ++x + bar (); @} @end example We find it easier to read a program when it has spaces before the open-parentheses and after the commas. Especially after the commas. When you split an expression into multiple lines, split it before an operator, not after one. Here is the right way: @example if (foo_this_is_long && bar > win (x, y, z) && remaining_condition) @end example Try to avoid having two operators of different precedence at the same level of indentation. For example, don't write this: @example mode = (inmode[j] == VOIDmode || GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j]) ? outmode[j] : inmode[j]); @end example Instead, use extra parentheses so that the indentation shows the nesting: @example mode = ((inmode[j] == VOIDmode || (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j]))) ? outmode[j] : inmode[j]); @end example Insert extra parentheses so that Emacs will indent the code properly. For example, the following indentation looks nice if you do it by hand, but Emacs would mess it up: @example v = rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000; @end example But adding a set of parentheses solves the problem: @example v = (rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000); @end example Format do-while statements like this: @example do @{ a = foo (a); @} while (a > 0); @end example Please use formfeed characters (control-L) to divide the program into pages at logical places (but not within a function). It does not matter just how long the pages are, since they do not have to fit on a printed page. The formfeeds should appear alone on lines by themselves. @node Comments @chapter Commenting Your Work Every program should start with a comment saying briefly what it is for. Example: @samp{fmt - filter for simple filling of text}. Please put a comment on each function saying what the function does, what sorts of arguments it gets, and what the possible values of arguments mean and are used for. It is not necessary to duplicate in words the meaning of the C argument declarations, if a C type is being used in its customary fashion. If there is anything nonstandard about its use (such as an argument of type @code{char *} which is really the address of the second character of a string, not the first), or any possible values that would not work the way one would expect (such as, that strings containing newlines are not guaranteed to work), be sure to say so. Also explain the significance of the return value, if there is one. Please put two spaces after the end of a sentence in your comments, so that the Emacs sentence commands will work. Also, please write complete sentences and capitalize the first word. If a lower-case identifer comes at the beginning of a sentence, don't capitalize it! Changing the spelling makes it a different identifier. If you don't like starting a sentence with a lower case letter, write the sentence differently (e.g., ``The identifier lower-case is @dots{}''). The comment on a function is much clearer if you use the argument names to speak about the argument values. The variable name itself should be lower case, but write it in upper case when you are speaking about the value rather than the variable itself. Thus, ``the inode number NODE_NUM'' rather than ``an inode''. There is usually no purpose in restating the name of the function in the comment before it, because the reader can see that for himself. There might be an exception when the comment is so long that the function itself would be off the bottom of the screen. There should be a comment on each static variable as well, like this: @example /* Nonzero means truncate lines in the display; zero means continue them. */ int truncate_lines; @end example Every @samp{#endif} should have a comment, except in the case of short conditionals (just a few lines) that are not nested. The comment should state the condition of the conditional that is ending, @emph{including its sense}. @samp{#else} should have a comment describing the condition @emph{and sense} of the code that follows. For example: @example #ifdef foo @dots{} #else /* not foo */ @dots{} #endif /* not foo */ @end example @noindent but, by contrast, write the comments this way for a @samp{#ifndef}: @example #ifndef foo @dots{} #else /* foo */ @dots{} #endif /* foo */ @end example @node Syntactic Conventions @chapter Clean Use of C Constructs Please explicitly declare all arguments to functions. Don't omit them just because they are @code{int}s. Declarations of external functions and functions to appear later in the source file should all go in one place near the beginning of the file (somewhere before the first function definition in the file), or else should go in a header file. Don't put @code{extern} declarations inside functions. It used to be common practice to use the same local variables (with names like @code{tem}) over and over for different values within one function. Instead of doing this, it is better declare a separate local variable for each distinct purpose, and give it a name which is meaningful. This not only makes programs easier to understand, it also facilitates optimization by good compilers. You can also move the declaration of each local variable into the smallest scope that includes all its uses. This makes the program even cleaner. Don't use local variables or parameters that shadow global identifiers. Don't declare multiple variables in one declaration that spans lines. Start a new declaration on each line, instead. For example, instead of this: @example int foo, bar; @end example @noindent write either this: @example int foo, bar; @end example @noindent or this: @example int foo; int bar; @end example @noindent (If they are global variables, each should have a comment preceding it anyway.) When you have an @code{if}-@code{else} statement nested in another @code{if} statement, always put braces around the @code{if}-@code{else}. Thus, never write like this: @example if (foo) if (bar) win (); else lose (); @end example @noindent always like this: @example if (foo) @{ if (bar) win (); else lose (); @} @end example If you have an @code{if} statement nested inside of an @code{else} statement, either write @code{else if} on one line, like this, @example if (foo) @dots{} else if (bar) @dots{} @end example @noindent with its @code{then}-part indented like the preceding @code{then}-part, or write the nested @code{if} within braces like this: @example if (foo) @dots{} else @{ if (bar) @dots{} @} @end example Don't declare both a structure tag and variables or typedefs in the same declaration. Instead, declare the structure tag separately and then use it to declare the variables or typedefs. Try to avoid assignments inside @code{if}-conditions. For example, don't write this: @example if ((foo = (char *) malloc (sizeof *foo)) == 0) fatal ("virtual memory exhausted"); @end example @noindent instead, write this: @example foo = (char *) malloc (sizeof *foo); if (foo == 0) fatal ("virtual memory exhausted"); @end example Don't make the program ugly to placate @code{lint}. Please don't insert any casts to @code{void}. Zero without a cast is perfectly fine as a null pointer constant. @node Names @chapter Naming Variables and Functions Please use underscores to separate words in a name, so that the Emacs word commands can be useful within them. Stick to lower case; reserve upper case for macros and @code{enum} constants, and for name-prefixes that follow a uniform convention. For example, you should use names like @code{ignore_space_change_flag}; don't use names like @code{iCantReadThis}. Variables that indicate whether command-line options have been specified should be named after the meaning of the option, not after the option-letter. A comment should state both the exact meaning of the option and its letter. For example, @example /* Ignore changes in horizontal whitespace (-b). */ int ignore_space_change_flag; @end example When you want to define names with constant integer values, use @code{enum} rather than @samp{#define}. GDB knows about enumeration constants. Use file names of 14 characters or less, to avoid creating gratuitous problems on System V. @node Using Extensions @chapter Using Non-standard Features Many GNU facilities that already exist support a number of convenient extensions over the comparable Unix facilities. Whether to use these extensions in implementing your program is a difficult question. On the one hand, using the extensions can make a cleaner program. On the other hand, people will not be able to build the program unless the other GNU tools are available. This might cause the program to work on fewer kinds of machines. With some extensions, it might be easy to provide both alternatives. For example, you can define functions with a ``keyword'' @code{INLINE} and define that as a macro to expand into either @code{inline} or nothing, depending on the compiler. In general, perhaps it is best not to use the extensions if you can straightforwardly do without them, but to use the extensions if they are a big improvement. An exception to this rule are the large, established programs (such as Emacs) which run on a great variety of systems. Such programs would be broken by use of GNU extensions. Another exception is for programs that are used as part of compilation: anything that must be compiled with other compilers in order to bootstrap the GNU compilation facilities. If these require the GNU compiler, then no one can compile them without having them installed already. That would be no good. Since most computer systems do not yet implement @sc{ANSI} C, using the @sc{ANSI} C features is effectively using a GNU extension, so the same considerations apply. (Except for @sc{ANSI} features that we discourage, such as trigraphs---don't ever use them.) @node System Functions @chapter Calling System Functions C implementations differ substantially. ANSI C reduces but does not eliminate the incompatibilities; meanwhile, many users wish to compile GNU software with pre-ANSI compilers. This chapter gives recommendations for how to use the more or less standard C library functions to avoid unnecessary loss of portability. @itemize @bullet @item Don't use the value of @code{sprintf}. It returns the number of characters written on some systems, but not on all systems. @item Don't declare system functions explicitly. Almost any declaration for a system function is wrong on some system. To minimize conflicts, leave it to the system header files to declare system functions. If the headers don't declare a function, let it remain undeclared. While it may seem unclean to use a function without declaring it, in practice this works fine for most system library functions on the systems where this really happens. The problem is only theoretical. By contrast, actual declarations have frequently caused actual conflicts. @item If you must declare a system function, don't specify the argument types. Use an old-style declaration, not an ANSI prototype. The more you specify about the function, the more likely a conflict. @item In particular, don't unconditionally declare @code{malloc} or @code{realloc}. Most GNU programs use those functions just once, in functions conventionally named @code{xmalloc} and @code{xrealloc}. These functions call @code{malloc} and @code{realloc}, respectively, and check the results. Because @code{xmalloc} and @code{xrealloc} are defined in your program, you can declare them in other files without any risk of type conflict. On most systems, @code{int} is the same length as a pointer; thus, the calls to @code{malloc} and @code{realloc} work fine. For the few exceptional systems (mostly 64-bit machines), you can use @strong{conditionalized} declarations of @code{malloc} and @code{realloc}---or put these declarations in configuration files specific to those systems. @item The string functions require special treatment. Some Unix systems have a header file @file{string.h}; other have @file{strings.h}. Neither file name is portable. There are two things you can do: use Autoconf to figure out which file to include, or don't include either file. @item If you don't include either strings file, you can't get declarations for the string functions from the header file in the usual way. That causes less of a problem than you might think. The newer ANSI string functions are off-limits anyway because many systems still don't support them. The string functions you can use are these: @example strcpy strncpy strcat strncat strlen strcmp strncmp strchr strrchr @end example The copy and concatenate functions work fine without a declaration as long as you don't use their values. Using their values without a declaration fails on systems where the width of a pointer differs from the width of @code{int}, and perhaps in other cases. It is trivial to avoid using their values, so do that. The compare functions and @code{strlen} work fine without a declaration on most systems, possibly all the ones that GNU software runs on. You may find it necessary to declare them @strong{conditionally} on a few systems. The search functions must be declared to return @code{char *}. Luckily, there is no variation in the data type they return. But there is variation in their names. Some systems give these functions the names @code{index} and @code{rindex}; other systems use the names @code{strchr} and @code{strrchr}. Some systems support both pairs of names, but neither pair works on all systems. You should pick a single pair of names and use it throughout your program. (Nowadays, it is better to choose @code{strchr} and @code{strrchr}.) Declare both of those names as functions returning @code{char *}. On systems which don't support those names, define them as macros in terms of the other pair. For example, here is what to put at the beginning of your file (or in a header) if you want to use the names @code{strchr} and @code{strrchr} throughout: @example #ifndef HAVE_STRCHR #define strchr index #endif #ifndef HAVE_STRRCHR #define strrchr rindex #endif char *strchr (); char *strrchr (); @end example @end itemize Here we assume that @code{HAVE_STRCHR} and @code{HAVE_STRRCHR} are macros defined in systems where the corresponding functions exist. One way to get them properly defined is to use Autoconf. @node Semantics @chapter Program Behavior for All Programs Avoid arbitrary limits on the length or number of @emph{any} data structure, including filenames, lines, files, and symbols, by allocating all data structures dynamically. In most Unix utilities, ``long lines are silently truncated''. This is not acceptable in a GNU utility. Utilities reading files should not drop NUL characters, or any other nonprinting characters @emph{including those with codes above 0177}. The only sensible exceptions would be utilities specifically intended for interface to certain types of printers that can't handle those characters. Check every system call for an error return, unless you know you wish to ignore errors. Include the system error text (from @code{perror} or equivalent) in @emph{every} error message resulting from a failing system call, as well as the name of the file if any and the name of the utility. Just ``cannot open foo.c'' or ``stat failed'' is not sufficient. Check every call to @code{malloc} or @code{realloc} to see if it returned zero. Check @code{realloc} even if you are making the block smaller; in a system that rounds block sizes to a power of 2, @code{realloc} may get a different block if you ask for less space. In Unix, @code{realloc} can destroy the storage block if it returns zero. GNU @code{realloc} does not have this bug: if it fails, the original block is unchanged. Feel free to assume the bug is fixed. If you wish to run your program on Unix, and wish to avoid lossage in this case, you can use the GNU @code{malloc}. You must expect @code{free} to alter the contents of the block that was freed. Anything you want to fetch from the block, you must fetch before calling @code{free}. Use @code{getopt_long} to decode arguments, unless the argument syntax makes this unreasonable. When static storage is to be written in during program execution, use explicit C code to initialize it. Reserve C initialized declarations for data that will not be changed. Try to avoid low-level interfaces to obscure Unix data structures (such as file directories, utmp, or the layout of kernel memory), since these are less likely to work compatibly. If you need to find all the files in a directory, use @code{readdir} or some other high-level interface. These will be supported compatibly by GNU. By default, the GNU system will provide the signal handling functions of @sc{BSD} and of @sc{POSIX}. So GNU software should be written to use these. In error checks that detect ``impossible'' conditions, just abort. There is usually no point in printing any message. These checks indicate the existence of bugs. Whoever wants to fix the bugs will have to read the source code and run a debugger. So explain the problem with comments in the source. The relevant data will be in variables, which are easy to examine with the debugger, so there is no point moving them elsewhere. @node Errors @chapter Formatting Error Messages Error messages from compilers should look like this: @example @var{source-file-name}:@var{lineno}: @var{message} @end example Error messages from other noninteractive programs should look like this: @example @var{program}:@var{source-file-name}:@var{lineno}: @var{message} @end example @noindent when there is an appropriate source file, or like this: @example @var{program}: @var{message} @end example @noindent when there is no relevant source file. In an interactive program (one that is reading commands from a terminal), it is better not to include the program name in an error message. The place to indicate which program is running is in the prompt or with the screen layout. (When the same program runs with input from a source other than a terminal, it is not interactive and would do best to print error messages using the noninteractive style.) The string @var{message} should not begin with a capital letter when it follows a program name and/or filename. Also, it should not end with a period. Error messages from interactive programs, and other messages such as usage messages, should start with a capital letter. But they should not end with a period. @node Libraries @chapter Library Behavior Try to make library functions reentrant. If they need to do dynamic storage allocation, at least try to avoid any nonreentrancy aside from that of @code{malloc} itself. Here are certain name conventions for libraries, to avoid name conflicts. Choose a name prefix for the library, more than two characters long. All external function and variable names should start with this prefix. In addition, there should only be one of these in any given library member. This usually means putting each one in a separate source file. An exception can be made when two external symbols are always used together, so that no reasonable program could use one without the other; then they can both go in the same file. External symbols that are not documented entry points for the user should have names beginning with @samp{_}. They should also contain the chosen name prefix for the library, to prevent collisions with other libraries. These can go in the same files with user entry points if you like. Static functions and variables can be used as you like and need not fit any naming convention. @node Portability @chapter Portability As It Applies to GNU Much of what is called ``portability'' in the Unix world refers to porting to different Unix versions. This is a secondary consideration for GNU software, because its primary purpose is to run on top of one and only one kernel, the GNU kernel, compiled with one and only one C compiler, the GNU C compiler. The amount and kinds of variation among GNU systems on different cpu's will be like the variation among Berkeley 4.3 systems on different cpu's. All users today run GNU software on non-GNU systems. So supporting a variety of non-GNU systems is desirable; simply not paramount. The easiest way to achieve portability to a reasonable range of systems is to use Autoconf. It's unlikely that your program needs to know more information about the host machine than Autoconf can provide, simply because most of the programs that need such knowledge have already been written. It is difficult to be sure exactly what facilities the GNU kernel will provide, since it isn't finished yet. Therefore, assume you can use anything in 4.3; just avoid using the format of semi-internal data bases (e.g., directories) when there is a higher-level alternative (@code{readdir}). You can freely assume any reasonably standard facilities in the C language, libraries or kernel, because we will find it necessary to support these facilities in the full GNU system, whether or not we have already done so. The fact that there may exist kernels or C compilers that lack these facilities is irrelevant as long as the GNU kernel and C compiler support them. It remains necessary to worry about differences among cpu types, such as the difference in byte ordering and alignment restrictions. It's unlikely that 16-bit machines will ever be supported by GNU, so there is no point in spending any time to consider the possibility that an int will be less than 32 bits. You can assume that all pointers have the same format, regardless of the type they point to, and that this is really an integer. There are some weird machines where this isn't true, but they aren't important; don't waste time catering to them. Besides, eventually we will put function prototypes into all GNU programs, and that will probably make your program work even on weird machines. Since some important machines (including the 68000) are big-endian, it is important not to assume that the address of an @code{int} object is also the address of its least-significant byte. Thus, don't make the following mistake: @example int c; @dots{} while ((c = getchar()) != EOF) write(file_descriptor, &c, 1); @end example You can assume that it is reasonable to use a meg of memory. Don't strain to reduce memory usage unless it can get to that level. If your program creates complicated data structures, just make them in core and give a fatal error if malloc returns zero. If a program works by lines and could be applied to arbitrary user-supplied input files, it should keep only a line in memory, because this is not very hard and users will want to be able to operate on input files that are bigger than will fit in core all at once. @node User Interfaces @chapter Standards for Command Line Interfaces Please don't make the behavior of a utility depend on the name used to invoke it. It is useful sometimes to make a link to a utility with a different name, and that should not change what it does. Instead, use a run time option or a compilation switch or both to select among the alternate behaviors. Likewise, please don't make the behavior of the program depend on the type of output device it is used with. Device independence is an important principle of the system's design; do not compromise it merely to save someone from typing an option now and then. If you think one behavior is most useful when the output is to a terminal, and another is most useful when the output is a file or a pipe, then it is usually best to make the default behavior the one that is useful with output to a terminal, and have an option for the other behavior. Compatibility requires certain programs to depend on the type of output device. It would be disastrous if @code{ls} or @code{sh} did not do so in the way all users expect. In some of these cases, we supplement the program with a preferred alternate version that does not depend on the output device type. For example, we provide a @code{dir} program much like @code{ls} except that its default output format is always multi-column format. It is a good idea to follow the @sc{POSIX} guidelines for the command-line options of a program. The easiest way to do this is to use @code{getopt} to parse them. Note that the GNU version of @code{getopt} will normally permit options anywhere among the arguments unless the special argument @samp{--} is used. This is not what @sc{POSIX} specifies; it is a GNU extension. Please define long-named options that are equivalent to the single-letter Unix-style options. We hope to make GNU more user friendly this way. This is easy to do with the GNU function @code{getopt_long}. One of the advantages of long-named options is that they can be consistent from program to program. For example, users should be able to expect the ``verbose'' option of any GNU program which has one, to be spelled precisely @samp{--verbose}. To achieve this uniformity, look at the table of common long-option names when you choose the option names for your program. The table appears below. If you use names not already in the table, please send @samp{gnu@@prep.ai.mit.edu} a list of them, with their meanings, so we can update the table. It is usually a good idea for file names given as ordinary arguments to be input files only; any output files would be specified using options (preferably @samp{-o}). Even if you allow an output file name as an ordinary argument for compatibility, try to provide a suitable option as well. This will lead to more consistency among GNU utilities, so that there are fewer idiosyncracies for users to remember. Programs should support an option @samp{--version} which prints the program's version number on standard output and exits successfully, and an option @samp{--help} which prints option usage information on standard output and exits successfully. These options should inhibit the normal function of the command; they should do nothing except print the requested information. @table @samp @item after-date @samp{-N} in @code{tar}. @item all @samp{-a} in @code{du}, @code{ls}, @code{nm}, @code{stty}, @code{uname}, and @code{unexpand}. @item all-text @samp{-a} in @code{diff}. @item almost-all @samp{-A} in @code{ls}. @item append @samp{-a} in @code{etags}, @code{tee}, @code{time}; @samp{-r} in @code{tar}. @item archive @samp{-a} in @code{cp}. @item ascii @samp{-a} in @code{diff}. @item assume-new @samp{-W} in Make. @item assume-old @samp{-o} in Make. @item backward-search @samp{-B} in etags. @item batch Used in GDB. @item baud Used in GDB. @item before @samp{-b} in @code{tac}. @item binary @samp{-b} in @code{cpio} and @code{diff}. @item block-size Used in @code{cpio} and @code{tar}. @item blocks @samp{-b} in @code{head} and @code{tail}. @item brief Used in various programs to make output shorter. @item bytes @samp{-c} in @code{head}, @code{split}, and @code{tail}. @item c++ @samp{-C} in @code{etags}. @item catenate @samp{-A} in @code{tar}. @item cd Used in various programs to specify the directory to use. @item changes @samp{-c} in @code{chgrp} and @code{chown}. @item classify @samp{-F} in @code{ls}. @item command @samp{-c} in @code{su}; @samp{-x} in GDB. @item compare @samp{-d} in @code{tar}. @item compress @samp{-z} in @code{tar}. @item concatenate @samp{-A} in @code{tar}. @item confirmation @samp{-w} in @code{tar}. @item context Used in @code{diff}. @item core Used in GDB. @item count @samp{-q} in @code{who}. @item count-links @samp{-l} in @code{du}. @item create Used in @code{tar} and @code{cpio}. @item cxref @samp{-x} in @code{etags}. @item date @samp{-d} in @code{touch}. @item debug @samp{-d} in @code{make}; @samp{-t} in Bison. @item defines @samp{-d} in Bison and @code{etags}. @item delete @samp{-D} in @code{tar}. @item dereference @samp{-L} in @code{chgrp}, @code{chown}, @code{cpio}, @code{du}, @code{ls}, and @code{tar}. @item dereference-args @samp{-D} in @code{du}. @item dictionary-order @samp{-d} in @code{look}. @item diff @samp{-d} in @code{tar}. @item directory Specify the directory to use, in various programs. In @code{ls}, it means to show directories themselves rather than their contents. In @code{rm} and @code{ln}, it means to not treat links to directories specially. @item discard-all @samp{-x} in @code{strip}. @item discard-locals @samp{-X} in @code{strip}. @item dry-run @samp{-n} in Make. @item ed @samp{-e} in @code{diff}. @item entire-new-file @samp{-N} in @code{diff}. @item environment-overrides @samp{-e} in Make. @item eof @samp{-e} in @code{xargs}. @item epoch Used in GDB. @item error-limit Used in Makeinfo. @item escape @samp{-b} in @code{ls}. @item exclude-from @samp{-X} in @code{tar}. @item exec Used in GDB. @item exit @samp{-x} in @code{xargs}. @item expand-tabs @samp{-t} in @code{diff}. @item expression @samp{-e} in @code{sed}. @item extern-only @samp{-g} in @code{nm}. @item extract @samp{-i} in @code{cpio}; @samp{-x} in @code{tar}. @item faces @samp{-f} in @code{finger}. @item fast @samp{-f} in @code{su}. @item file @samp{-f} in @code{info}, Make, @code{mt}, and @code{tar}. @samp{-n} in @code{sed}; @samp{-r} in @code{touch}. @item file-prefix @samp{-b} in Bison. @item file-type @samp{-F} in @code{ls}. @item files-from @samp{-T} in @code{tar}. @item fill-column Used in Makeinfo. @item fixed-output-files @samp{-y} in Bison. @item follow @samp{-f} in @code{tail}. @item footnote-style Used in Makeinfo. @item force @samp{-f} in @code{cp}, @code{ln}, @code{mv}, and @code{rm}. @item format Used in @code{ls} and @code{time}. @item forward-search @samp{-F} in @code{etags}. @item fullname Used in GDB. @item get @samp{-x} in @code{tar}. @item graphic @samp{-i} in @code{ul}. @item group @samp{-g} in @code{install}. @item header @samp{-h} in @code{objdump}. @item heading @samp{-H} in @code{who}. @item help Used to ask for brief usage information. @item hide-control-chars @samp{-q} in @code{ls}. @item idle @samp{-u} in @code{who}. @item ifdef @samp{-D} in @code{diff}. @item ignore @samp{-I} in @code{ls}. @item ignore-all-space @samp{-w} in @code{diff}. @item ignore-backups @samp{-B} in @code{ls}. @item ignore-blank-lines @samp{-B} in @code{diff}. @item ignore-case @samp{-f} in @code{look}; @samp{-i} in @code{diff}. @item ignore-errors @samp{-i} in Make. @item ignore-indentation @samp{-S} in @code{etags}. @item ignore-init-file @samp{-f} in Oleo. @item ignore-interrupts @samp{-i} in @code{tee}. @item ignore-matching-lines @samp{-I} in @code{diff}. @item ignore-space-change @samp{-b} in @code{diff}. @item ignore-zeros @samp{-i} in @code{tar}. @item include @samp{-i} in @code{etags}. @item include-dir @samp{-I} in Make. @item incremental @samp{-G} in @code{tar}. @item info @samp{-i}, @samp{-l}, and @samp{-m} in Finger. @item initial @samp{-i} in @code{expand}. @item initial-tab @samp{-T} in @code{diff}. @item inode @samp{-i} in @code{ls}. @item interactive @samp{-i} in @code{cp}, @code{ln}, @code{mv}, @code{rm}; @samp{-p} in @code{xargs}; @samp{-w} in @code{tar}. @item jobs @samp{-j} in Make. @item just-print @samp{-n} in Make. @item keep-going @samp{-k} in Make. @item kilobytes @samp{-k} in @code{du} and @code{ls}. @item line-bytes @samp{-C} in @code{split}. @item lines Used in @code{split}, @code{head}, and @code{tail}. @item link @samp{-l} in @code{cpio}. @item list @samp{-t} in @code{cpio}. @item list @samp{-t} in @code{tar}. @item literal @samp{-N} in @code{ls}. @item load-average @samp{-l} in Make. @item login Used in @code{su}. @item machine @item mail @samp{-m} in @code{hello}. @samp{-m} in @code{uname}. @item make-directories @samp{-d} in @code{cpio}. @item makefile @samp{-f} in Make. @item mapped Used in GDB. @item max-args @samp{-n} in @code{xargs}. @item max-chars @samp{-n} in @code{xargs}. @item max-lines @samp{-l} in @code{xargs}. @item max-load @samp{-l} in Make. @item max-procs @samp{-P} in @code{xargs}. @item mesg @samp{-T} in @code{who}. @item message @samp{-T} in @code{who}. @item minimal @samp{-d} in @code{diff}. @item mode @samp{-m} in @code{install}, @code{mkdir}, and @code{mkfifo}. @item modification-time @samp{-m} in @code{tar}. @item multi-volume @samp{-M} in @code{tar}. @item name-prefix @samp{-a} in Bison. @item new-file @samp{-W} in Make. @item no-builtin-rules @samp{-r} in Make. @item no-create @samp{-c} in @code{touch}. @item no-defines @samp{-D} in @code{etags}. @item no-dereference @samp{-d} in @code{cp}. @item no-keep-going @samp{-S} in Make. @item no-lines @samp{-l} in Bison. @item no-prof @samp{-e} in @code{gprof}. @item no-sort @samp{-p} in @code{nm}. @item no-split Used in Makeinfo. @item no-static @samp{-a} in @code{gprof}. @item no-time @samp{-E} in @code{gprof}. @item no-validate Used in Makeinfo. @item no-warn Used in various programs to inhibit warnings. @item node @samp{-n} in @code{info}. @item nodename @samp{-n} in @code{uname}. @item nonmatching @samp{-f} in @code{cpio}. @item nstuff @samp{-n} in @code{objdump}. @item null @samp{-0} in @code{xargs}. @item number @samp{-n} in @code{cat}. @item number-nonblank @samp{-b} in @code{cat}. @item numeric-sort @samp{-n} in @code{nm}. @item numeric-uid-gid @samp{-n} in @code{cpio} and @code{ls}. @item nx Used in GDB. @item old-archive @samp{-o} in @code{tar}. @item old-file @samp{-o} in Make. @item one-file-system @samp{-l} in @code{tar}, @code{cp}, and @code{du}. @item only-prof @samp{-f} in @code{gprof}. @item only-time @samp{-F} in @code{gprof}. @item output In various programs, specify the output file name. @item override @samp{-o} in @code{rm}. @item owner @samp{-o} in @code{install}. @item paginate @samp{-l} in @code{diff}. @item paragraph-indent Used in Makeinfo. @item parents @samp{-p} in @code{mkdir} and @code{rmdir}. @item pass-all @samp{-p} in @code{ul}. @item pass-through @samp{-p} in @code{cpio}. @item port @samp{-P} in @code{finger}. @item portability @samp{-c} in @code{cpio} and @code{tar}. @item preserve Used in @code{tar} and @code{cp}. @item preserve-environment @samp{-p} in @code{su}. @item preserve-modification-time @samp{-m} in @code{cpio}. @item preserve-order @samp{-s} in @code{tar}. @item preserve-permissions @samp{-p} in @code{tar}. @item print @samp{-l} in @code{diff}. @item print-chars @samp{-L} in @code{cmp}. @item print-data-base @samp{-p} in Make. @item print-directory @samp{-w} in Make. @item print-file-name @samp{-o} in @code{nm}. @item print-symdefs @samp{-s} in @code{nm}. @item question @samp{-q} in Make. @item quiet Used in many programs to inhibit the usual output. @strong{Note:} every program accepting @samp{--quiet} should accept @samp{--silent} as a synonym. @item quote-name @samp{-Q} in @code{ls}. @item rcs @samp{-n} in @code{diff}. @item read-full-blocks @samp{-B} in @code{tar}. @item readnow Used in GDB. @item recon @samp{-n} in Make. @item record-number @samp{-R} in @code{tar}. @item recursive Used in @code{chgrp}, @code{chown}, @code{cp}, @code{ls}, @code{diff}, and @code{rm}. @item reference-limit Used in Makeinfo. @item regex @samp{-r} in @code{tac}. @item release @samp{-r} in @code{uname}. @item relocation @samp{-r} in @code{objdump}. @item rename @samp{-r} in @code{cpio}. @item replace @samp{-i} in @code{xargs}. @item report-identical-files @samp{-s} in @code{diff}. @item reset-access-time @samp{-a} in @code{cpio}. @item reverse @samp{-r} in @code{ls} and @code{nm}. @item reversed-ed @samp{-f} in @code{diff}. @item same-order @samp{-s} in @code{tar}. @item same-permissions @samp{-p} in @code{tar}. @item save @samp{-g} in @code{stty}. @item se Used in GDB. @item separate-dirs @samp{-S} in @code{du}. @item separator @samp{-s} in @code{tac}. @item shell @samp{-s} in @code{su}. @item show-all @samp{-A} in @code{cat}. @item show-c-function @samp{-p} in @code{diff}. @item show-ends @samp{-E} in @code{cat}. @item show-function-line @samp{-F} in @code{diff}. @item show-tabs @samp{-T} in @code{cat}. @item silent Used in many programs to inhibit the usual output. @strong{Note:} every program accepting @samp{--silent} should accept @samp{--quiet} as a synonym. @item size @samp{-s} in @code{ls}. @item sort Used in @code{ls}. @item sparse @samp{-S} in @code{tar}. @item speed-large-files @samp{-H} in @code{diff}. @item squeeze-blank @samp{-s} in @code{cat}. @item starting-file Used in @code{tar} and @code{diff} to specify which file within a directory to start processing with. @item stop @samp{-S} in Make. @item strip @samp{-s} in @code{install}. @item strip-all @samp{-s} in @code{strip}. @item strip-debug @samp{-S} in @code{strip}. @item suffix @samp{-S} in @code{cp}, @code{ln}, @code{mv}. @item sum @samp{-s} in @code{gprof}. @item summarize @samp{-s} in @code{du}. @item symbolic @samp{-s} in @code{ln}. @item symbols Used in GDB and @code{objdump}. @item sysname @samp{-s} in @code{uname}. @item tabs @samp{-t} in @code{expand} and @code{unexpand}. @item tabsize @samp{-T} in @code{ls}. @item terminal @samp{-T} in @code{tput} and @code{ul}. @item text @samp{-a} in @code{diff}. @item time Used in @code{ls} and @code{touch}. @item to-stdout @samp{-O} in @code{tar}. @item total @samp{-c} in @code{du}. @item touch @samp{-t} in Make and @code{ranlib}. @item traditional @samp{-t} in @code{hello}. @item tty Used in GDB. @item typedefs @samp{-t} in @code{etags}. @item typedefs-and-c++ @samp{-T} in @code{etags}. @item uncompress @samp{-z} in @code{tar}. @item unconditional @samp{-u} in @code{cpio}. @item undefined-only @samp{-u} in @code{nm}. @item update @samp{-u} in @code{cp}, @samp{etags}, @samp{mv}, @samp{tar}. @item verbose Print more information about progress. Many programs support this. @item verify @samp{-W} in @code{tar}. @item version Print the version number. @item version-control @samp{-V} in @code{cp}, @code{ln}, @code{mv}. @item vgrind @samp{-v} in @code{etags}. @item volume @samp{-V} in @code{tar}. @item what-if @samp{-W} in Make. @item width @samp{-w} in @code{ls}. @item writable @samp{-T} in @code{who}. @item zeros @samp{-z} in @code{gprof}. @end table @node Documentation @chapter Documenting Programs Please use Texinfo for documenting GNU programs. See the Texinfo manual, either the hardcopy or the version in the GNU Emacs Info subsystem (@kbd{C-h i}). See existing GNU Texinfo files (e.g., those under the @file{man/} directory in the GNU Emacs distribution) for examples. The title page of the manual should state the version of the program which the manual applies to. The Top node of the manual should also contain this information. If the manual is changing more frequently than or independent of the program, also state a version number for the manual in both of these places. The manual should document all command-line arguments and all commands. It should give examples of their use. But don't organize the manual as a list of features. Instead, organize it by the concepts a user will have before reaching that point in the manual. Address the goals that a user will have in mind, and explain how to accomplish them. Don't use Unix man pages as a model for how to write GNU documentation; they are a bad example to follow. The manual should have a node named @samp{@var{program} Invocation} or @samp{Invoking @var{program}}, where @var{program} stands for the name of the program being described, as you would type it in the shell to run the program. This node (together with its subnodes, if any) should describe the program's command line arguments and how to run it (the sort of information people would look in a man page for). Start with an @samp{@@example} containing a template for all the options and arguments that the program uses. Alternatively, put a menu item in some menu whose item name fits one of the above patterns. This identifies the node which that item points to as the node for this purpose, regardless of the node's actual name. There will be automatic features for specifying a program name and quickly reading just this part of its manual. If one manual describes several programs, it should have such a node for each program described. In addition to its manual, the package should have a file named @file{NEWS} which contains a list of user-visible changes worth mentioning. In each new release, add items to the front of the file and identify the version they pertain to. Don't discard old items; leave them in the file after the newer items. This way, a user upgrading from any previous version can see what is new. If the @file{NEWS} file gets very long, move some of the older items into a file named @file{ONEWS} and put a note at the end referring the user to that file. It is ok to supply a man page for the program as well as a Texinfo manual if you wish to. But keep in mind that supporting a man page requires continual effort, each time the program is changed. Any time you spend on the man page is time taken away from more useful things you could contribute. Thus, even if a user volunteers to donate a man page, you may find this gift costly to accept. Unless you have time on your hands, it may be better to refuse the man page unless the same volunteer agrees to take full responsibility for maintaining it---so that you can wash your hands of it entirely. If the volunteer ceases to do the job, then don't feel obliged to pick it up yourself; it may be better to withdraw the man page until another volunteer offers to carry on with it. Alternatively, if you expect the discrepancies to be small enough that the man page remains useful, put a prominent note near the beginning of the man page explaining that you don't maintain it and that the Texinfo manual is more authoritative, and describing how to access the Texinfo documentation. @node Releases @chapter Making Releases Package the distribution of Foo version 69.96 in a gzipped tar file named @file{foo-69.96.tar.gz}. It should unpack into a subdirectory named @file{foo-69.96}. Building and installing the program should never modify any of the files contained in the distribution. This means that all the files that form part of the program in any way must be classified into @dfn{source files} and @dfn{non-source files}. Source files are written by humans and never changed automatically; non-source files are produced from source files by programs under the control of the Makefile. Naturally, all the source files must be in the distribution. It is okay to include non-source files in the distribution, provided they are up-to-date and machine-independent, so that building the distribution normally will never modify them. We commonly included non-source files produced by Bison, Lex, @TeX{}, and Makeinfo; this helps avoid unnecessary dependencies between our distributions, so that users can install whichever packages they want to install. Non-source files that might actually be modified by building and installing the program should @strong{never} be included in the distribution. So if you do distribute non-source files, always make sure they are up to date when you make a new distribution. Make sure that the directory into which the distribution unpacks (as well as any subdirectories) are all world-writable (octal mode 777). This is so that old versions of @code{tar} which preserve the ownership and permissions of the files from the tar archive will be able to extract all the files even if the user is unprivileged. Make sure that no file name in the distribution is more than 14 characters long. Likewise, no file created by building the program should have a name longer than 14 characters. The reason for this is that some systems adhere to a foolish interpretation of the POSIX standard, and refuse to open a longer name, rather than truncating as they did in the past. Don't include any symbolic links in the distribution itself. If the tar file contains symbolic links, then people cannot even unpack it on systems that don't support symbolic links. Also, don't use multiple names for one file in different directories, because certain file systems cannot handle this and that prevents unpacking the distribution. Try to make sure that all the file names will be unique on MS-DOG. A name on MS-DOG consists of up to 8 characters, optionally followed by a period and up to three characters. MS-DOG will truncate extra characters both before and after the period. Thus, @file{foobarhacker.c} and @file{foobarhacker.o} are not ambiguous; they are truncated to @file{foobarha.c} and @file{foobarha.o}, which are distinct. Include in your distribution a copy of the @file{texinfo.tex} you used to test print any @file{*.texinfo} files. Likewise, if your program uses small GNU software packages like regex, getopt, obstack, or termcap, include them in the distribution file. Leaving them out would make the distribution file a little smaller at the expense of possible inconvenience to a user who doesn't know what other files to get. @contents @bye autoconf-1.7/make-stds.texi100666 21570 1 47721 5432524602 14271 0ustar djmdaemon@comment This file is included by both standards.texi and make.texinfo. @comment It was broken out of standards.texi on 1/6/93 by roland. @node Makefile Conventions @chapter Makefile Conventions @comment standards.texi does not print an index, but make.texinfo does. @cindex makefile, conventions for @cindex conventions for makefiles @cindex standards for makefiles This chapter describes conventions for writing the Makefiles for GNU programs. @menu * Makefile Basics:: * Utilities in Makefiles:: * Standard Targets:: * Command Variables:: * Directory Variables:: @end menu @node Makefile Basics @section General Conventions for Makefiles Every Makefile should contain this line: @example SHELL = /bin/sh @end example @noindent to avoid trouble on systems where the @code{SHELL} variable might be inherited from the environment. (This is never a problem with GNU @code{make}.) Don't assume that @file{.} is in the path for command execution. When you need to run programs that are a part of your package during the make, please make sure that it uses @file{./} if the program is built as part of the make or @file{$(srcdir)/} if the file is an unchanging part of the source code. Without one of these prefixes, the current search path is used. The distinction between @file{./} and @file{$(srcdir)/} is important when using the @samp{--srcdir} option to @file{configure}. A rule of the form: @smallexample foo.1 : foo.man sedscript sed -e sedscript foo.man > foo.1 @end smallexample @noindent will fail when the current directory is not the source directory, because @file{foo.man} and @file{sedscript} are not in the current directory. When using GNU @code{make}, relying on @samp{VPATH} to find the source file will work in the case where there is a single dependency file, since the @file{make} automatic variable @samp{$<} will represent the source file wherever it is. (Many versions of @code{make} set @samp{$<} only in implicit rules.) A makefile target like @smallexample foo.o : bar.c $(CC) -I. -I$(srcdir) $(CFLAGS) -c bar.c -o foo.o @end smallexample @noindent should instead be written as @smallexample foo.o : bar.c $(CC) $(CFLAGS) $< -o $@@ @end smallexample @noindent in order to allow @samp{VPATH} to work correctly. When the target has multiple dependencies, using an explicit @samp{$(srcdir)} is the easiest way to make the rule work well. For example, the target above for @file{foo.1} is best written as: @smallexample foo.1 : foo.man sedscript sed -s $(srcdir)/sedscript $(srcdir)/foo.man > foo.1 @end smallexample @node Utilities in Makefiles @section Utilities in Makefiles Write the Makefile commands (and any shell scripts, such as @code{configure}) to run in @code{sh}, not in @code{csh}. Don't use any special features of @code{ksh} or @code{bash}. The @code{configure} script and the Makefile rules for building and installation should not use any utilities directly except these: @example cat cmp cp echo egrep expr grep ln mkdir mv pwd rm rmdir sed test touch @end example Stick to the generally supported options for these programs. For example, don't use @samp{mkdir -p}, convenient as it may be, because most systems don't support it. The Makefile rules for building and installation can also use compilers and related programs, but should do so via @code{make} variables so that the user can substitute alternatives. Here are some of the programs we mean: @example ar bison cc flex install ld lex make makeinfo ranlib texi2dvi yacc @end example When you use @code{ranlib}, you should test whether it exists, and run it only if it exists, so that the distribution will work on systems that don't have @code{ranlib}. If you use symbolic links, you should implement a fallback for systems that don't have symbolic links. It is ok to use other utilities in Makefile portions (or scripts) intended only for particular systems where you know those utilities to exist. @node Standard Targets @section Standard Targets for Users All GNU programs should have the following targets in their Makefiles: @table @samp @item all Compile the entire program. This should be the default target. This target need not rebuild any documentation files; Info files should normally be included in the distribution, and DVI files should be made only when explicitly asked for. @item install Compile the program and copy the executables, libraries, and so on to the file names where they should reside for actual use. If there is a simple test to verify that a program is properly installed, this target should run that test. The commands should create all the directories in which files are to be installed, if they don't already exist. This includes the directories specified as the values of the variables @code{prefix} and @code{exec_prefix}, as well as all subdirectories that are needed. One way to do this is by means of an @code{installdirs} target as described below. Use @samp{-} before any command for installing a man page, so that @code{make} will ignore any errors. This is in case there are systems that don't have the Unix man page documentation system installed. The way to install Info files is to copy them into @file{$(infodir)} with @code{$(INSTALL_DATA)} (@pxref{Command Variables}), and then run the @code{install-info} program if it is present. @code{install-info} is a script that edits the Info @file{dir} file to add or update the menu entry for the given Info file; it will be part of the Texinfo package. Here is a sample rule to install an Info file: @comment This example has been carefully formatted for the Make manual. @comment Please do not reformat it without talking to roland@gnu.ai.mit.edu. @smallexample $(infodir)/foo.info: foo.info # There may be a newer info file in . than in srcdir. -if test -f foo.info; then d=.; \ else d=$(srcdir); fi; \ $(INSTALL_DATA) $$d/foo.info $@@; \ # Run install-info only if it exists. # Use `if' instead of just prepending `-' to the # line so we notice real errors from install-info. # We use `$(SHELL) -c' because some shells do not # fail gracefully when there is an unknown command. if $(SHELL) -c 'install-info --version' \ >/dev/null 2>&1; then \ install-info --infodir=$(infodir) $$d/foo.info; \ else true; fi @end smallexample @item uninstall Delete all the installed files that the @samp{install} target would create (but not the noninstalled files such as @samp{make all} would create). @comment The gratuitous blank line here is to make the table look better @comment in the printed Make manual. Please leave it in. @item clean Delete all files from the current directory that are normally created by building the program. Don't delete the files that record the configuration. Also preserve files that could be made by building, but normally aren't because the distribution comes with them. Delete @file{.dvi} files here if they are not part of the distribution. @item distclean Delete all files from the current directory that are created by configuring or building the program. If you have unpacked the source and built the program without creating any other files, @samp{make distclean} should leave only the files that were in the distribution. @item mostlyclean Like @samp{clean}, but may refrain from deleting a few files that people normally don't want to recompile. For example, the @samp{mostlyclean} target for GCC does not delete @file{libgcc.a}, because recompiling it is rarely necessary and takes a lot of time. @item realclean Delete everything from the current directory that can be reconstructed with this Makefile. This typically includes everything deleted by @code{distclean}, plus more: C source files produced by Bison, tags tables, Info files, and so on. One exception, however: @samp{make realclean} should not delete @file{configure} even if @file{configure} can be remade using a rule in the Makefile. More generally, @samp{make realclean} should not delete anything that needs to exist in order to run @file{configure} and then begin to build the program. @item TAGS Update a tags table for this program. @item info Generate any Info files needed. The best way to write the rules is as follows: @smallexample info: foo.info foo.info: foo.texi chap1.texi chap2.texi $(MAKEINFO) $(srcdir)/foo.texi @end smallexample @noindent You must define the variable @code{MAKEINFO} in the Makefile. It should run the @code{makeinfo} program, which is part of the Texinfo distribution. @item dvi Generate DVI files for all TeXinfo documentation. For example: @smallexample dvi: foo.dvi foo.dvi: foo.texi chap1.texi chap2.texi $(TEXI2DVI) $(srcdir)/foo.texi @end smallexample @noindent You must define the variable @code{TEXI2DVI} in the Makefile. It should run the program @code{texi2dvi}, which is part of the Texinfo distribution. Alternatively, write just the dependencies, and allow GNU Make to provide the command. @item dist Create a distribution tar file for this program. The tar file should be set up so that the file names in the tar file start with a subdirectory name which is the name of the package it is a distribution for. This name can include the version number. For example, the distribution tar file of GCC version 1.40 unpacks into a subdirectory named @file{gcc-1.40}. The easiest way to do this is to create a subdirectory appropriately named, use @code{ln} or @code{cp} to install the proper files in it, and then @code{tar} that subdirectory. The @code{dist} target should explicitly depend on all non-source files that are in the distribution, to make sure they are up to date in the distribution. @xref{Releases, , Making Releases, standards, GNU Coding Standards}. @item check Perform self-tests (if any). The user must build the program before running the tests, but need not install the program; you should write the self-tests so that they work when the program is built but not installed. @end table The following targets are suggested as conventional names, for programs in which they are useful. @table @code @item installcheck Perform installation tests (if any). The user must build and install the program before running the tests. You should not assume that @file{$(bindir)} is in the search path. @item installdirs It's useful to add a target named @samp{installdirs} to create the directories where files are installed, and their parent directories. There is a script called @file{mkinstalldirs} which is convenient for this; find it in the Texinfo package.@c It's in /gd/gnu/lib/mkinstalldirs. You can use a rule like this: @comment This has been carefully formatted to look decent in the Make manual. @comment Please be sure not to make it extend any further to the right.--roland @smallexample # Make sure all installation directories (e.g. $(bindir)) # actually exist by making them if necessary. installdirs: mkinstalldirs $(srcdir)/mkinstalldirs $(bindir) $(datadir) \ $(libdir) $(infodir) \ $(mandir) @end smallexample @end table @node Command Variables @section Variables for Specifying Commands Makefiles should provide variables for overriding certain commands, options, and so on. In particular, you should run most utility programs via variables. Thus, if you use Bison, have a variable named @code{BISON} whose default value is set with @samp{BISON = bison}, and refer to it with @code{$(BISON)} whenever you need to use Bison. File management utilities such as @code{ln}, @code{rm}, @code{mv}, and so on, need not be referred to through variables in this way, since users don't need to replace them with other programs. Each program-name variable should come with an options variable that is used to supply options to the program. Append @samp{FLAGS} to the program-name variable name to get the options variable name---for example, @code{BISONFLAGS}. (The name @code{CFLAGS} is an exception to this rule, but we keep it because it is standard.) Use @code{CPPFLAGS} in any compilation command that runs the preprocessor, and use @code{LDFLAGS} in any compilation command that does linking as well as in any direct use of @code{ld}. If there are C compiler options that @emph{must} be used for proper compilation of certain files, do not include them in @code{CFLAGS}. Users expect to be able to specify @code{CFLAGS} freely themselves. Instead, arrange to pass the necessary options to the C compiler independently of @code{CFLAGS}, by writing them explicitly in the compilation commands or by defining an implicit rule, like this: @smallexample CFLAGS = -g ALL_CFLAGS = -I. $(CFLAGS) .c.o: $(CC) -c $(CPPFLAGS) $(ALL_CFLAGS) $< @end smallexample Do include the @samp{-g} option in @code{CFLAGS}, because that is not @emph{required} for proper compilation. You can consider it a default that is only recommended. If the package is set up so that it is compiled with GCC by default, then you might as well include @samp{-O} in the default value of @code{CFLAGS} as well. Put @code{CFLAGS} last in the compilation command, after other variables containing compiler options, so the user can use @code{CFLAGS} to override the others. Every Makefile should define the variable @code{INSTALL}, which is the basic command for installing a file into the system. Every Makefile should also define the variables @code{INSTALL_PROGRAM} and @code{INSTALL_DATA}. (The default for each of these should be @code{$(INSTALL)}.) Then it should use those variables as the commands for actual installation, for executables and nonexecutables respectively. Use these variables as follows: @example $(INSTALL_PROGRAM) foo $(bindir)/foo $(INSTALL_DATA) libfoo.a $(libdir)/libfoo.a @end example @noindent Always use a file name, not a directory name, as the second argument of the installation commands. Use a separate command for each file to be installed. @node Directory Variables @section Variables for Installation Directories Installation directories should always be named by variables, so it is easy to install in a nonstandard place. The standard names for these variables are: @table @samp @item prefix A prefix used in constructing the default values of the variables listed below. The default value of @code{prefix} should be @file{/usr/local} (at least for now). @item exec_prefix A prefix used in constructing the default values of some of the variables listed below. The default value of @code{exec_prefix} should be @code{$(prefix)}. Generally, @code{$(exec_prefix)} is used for directories that contain machine-specific files (such as executables and subroutine libraries), while @code{$(prefix)} is used directly for other directories. @item bindir The directory for installing executable programs that users can run. This should normally be @file{/usr/local/bin}, but write it as @file{$(exec_prefix)/bin}. @item libdir The directory for installing executable files to be run by the program rather than by users. Object files and libraries of object code should also go in this directory. The idea is that this directory is used for files that pertain to a specific machine architecture, but need not be in the path for commands. The value of @code{libdir} should normally be @file{/usr/local/lib}, but write it as @file{$(exec_prefix)/lib}. @item datadir The directory for installing read-only data files which the programs refer to while they run. This directory is used for files which are independent of the type of machine being used. This should normally be @file{/usr/local/lib}, but write it as @file{$(prefix)/lib}. @item statedir The directory for installing data files which the programs modify while they run. These files should be independent of the type of machine being used, and it should be possible to share them among machines at a network installation. This should normally be @file{/usr/local/lib}, but write it as @file{$(prefix)/lib}. @item includedir @c rewritten to avoid overfull hbox --roland The directory for installing header files to be included by user programs with the C @samp{#include} preprocessor directive. This should normally be @file{/usr/local/include}, but write it as @file{$(prefix)/include}. Most compilers other than GCC do not look for header files in @file{/usr/local/include}. So installing the header files this way is only useful with GCC. Sometimes this is not a problem because some libraries are only really intended to work with GCC. But some libraries are intended to work with other compilers. They should install their header files in two places, one specified by @code{includedir} and one specified by @code{oldincludedir}. @item oldincludedir The directory for installing @samp{#include} header files for use with compilers other than GCC. This should normally be @file{/usr/include}. The Makefile commands should check whether the value of @code{oldincludedir} is empty. If it is, they should not try to use it; they should cancel the second installation of the header files. A package should not replace an existing header in this directory unless the header came from the same package. Thus, if your Foo package provides a header file @file{foo.h}, then it should install the header file in the @code{oldincludedir} directory if either (1) there is no @file{foo.h} there or (2) the @file{foo.h} that exists came from the Foo package. To tell whether @file{foo.h} came from the Foo package, put a magic string in the file---part of a comment---and grep for that string. @item mandir The directory for installing the man pages (if any) for this package. It should include the suffix for the proper section of the manual---usually @samp{1} for a utility. It will normally be @file{/usr/local/man/man1}, but you should write it as @file{$(prefix)/man/man1}. @item man1dir The directory for installing section 1 man pages. @item man2dir The directory for installing section 2 man pages. @item @dots{} Use these names instead of @samp{mandir} if the package needs to install man pages in more than one section of the manual. @strong{Don't make the primary documentation for any GNU software be a man page. Write a manual in Texinfo instead. Man pages are just for the sake of people running GNU software on Unix, which is a secondary application only.} @item manext The file name extension for the installed man page. This should contain a period followed by the appropriate digit; it should normally be @samp{.1}. @item man1ext The file name extension for installed section 1 man pages. @item man2ext The file name extension for installed section 2 man pages. @item @dots{} Use these names instead of @samp{manext} if the package needs to install man pages in more than one section of the manual. @item infodir The directory for installing the Info files for this package. By default, it should be @file{/usr/local/info}, but it should be written as @file{$(prefix)/info}. @item srcdir The directory for the sources being compiled. The value of this variable is normally inserted by the @code{configure} shell script. @end table For example: @smallexample @c I have changed some of the comments here slightly to fix an overfull @c hbox, so the make manual can format correctly. --roland # Common prefix for installation directories. # NOTE: This directory must exist when you start the install. prefix = /usr/local exec_prefix = $(prefix) # Where to put the executable for the command `gcc'. bindir = $(exec_prefix)/bin # Where to put the directories used by the compiler. libdir = $(exec_prefix)/lib # Where to put the Info files. infodir = $(prefix)/info @end smallexample If your program installs a large number of files into one of the standard user-specified directories, it might be useful to group them into a subdirectory particular to that program. If you do this, you should write the @code{install} rule to create these subdirectories. Do not expect the user to include the subdirectory name in the value of any of the variables listed above. The idea of having a uniform set of variable names for installation directories is to enable the user to specify the exact same values for several different GNU packages. In order for this to be useful, all the packages must be designed so that they will work sensibly when the user does so. autoconf-1.7/standards.info100666 74720 14 242771 5463274733 15023 0ustar eggertguestThis is Info file standards.info, produced by Makeinfo-1.54 from the input file ./standards.texi. Copyright (C) 1992, 1993 Free Software Foundation Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Free Software Foundation.  File: standards.info, Node: Top, Next: Reading Non-Free Code, Prev: (dir), Up: (dir) Version ******* Last updated 9 September 1993. * Menu: * Reading Non-Free Code:: Referring to Proprietary Programs * Contributions:: Accepting Contributions * Change Logs:: Recording Changes * Compatibility:: Compatibility with Other Implementations * Makefile Conventions:: Makefile Conventions * Configuration:: How Configuration Should Work * Source Language:: Using Languages Other Than C * Formatting:: Formatting Your Source Code * Comments:: Commenting Your Work * Syntactic Conventions:: Clean Use of C Constructs * Names:: Naming Variables and Functions * Using Extensions:: Using Non-standard Features * System Functions:: Portability and "standard" library functions * Semantics:: Program Behavior for All Programs * Errors:: Formatting Error Messages * Libraries:: Library Behavior * Portability:: Portability As It Applies to GNU * User Interfaces:: Standards for Command Line Interfaces * Documentation:: Documenting Programs * Releases:: Making Releases  File: standards.info, Node: Reading Non-Free Code, Next: Contributions, Prev: Top, Up: Top Referring to Proprietary Programs ********************************* Don't in any circumstances refer to Unix source code for or during your work on GNU! (Or to any other proprietary programs.) If you have a vague recollection of the internals of a Unix program, this does not absolutely mean you can't write an imitation of it, but do try to organize the imitation internally along different lines, because this is likely to make the details of the Unix version irrelevant and dissimilar to your results. For example, Unix utilities were generally optimized to minimize memory use; if you go for speed instead, your program will be very different. You could keep the entire input file in core and scan it there instead of using stdio. Use a smarter algorithm discovered more recently than the Unix program. Eliminate use of temporary files. Do it in one pass instead of two (we did this in the assembler). Or, on the contrary, emphasize simplicity instead of speed. For some applications, the speed of today's computers makes simpler algorithms adequate. Or go for generality. For example, Unix programs often have static tables or fixed-size strings, which make for arbitrary limits; use dynamic allocation instead. Make sure your program handles NULs and other funny characters in the input files. Add a programming language for extensibility and write part of the program in that language. Or turn some parts of the program into independently usable libraries. Or use a simple garbage collector instead of tracking precisely when to free memory, or use a new GNU facility such as obstacks.  File: standards.info, Node: Contributions, Next: Change Logs, Prev: Reading Non-Free Code, Up: Top Accepting Contributions *********************** If someone else sends you a piece of code to add to the program you are working on, we need legal papers to use it--the same sort of legal papers we will need to get from you. *Each* significant contributor to a program must sign some sort of legal papers in order for us to have clear title to the program. The main author alone is not enough. So, before adding in any contributions from other people, tell us so we can arrange to get the papers. Then wait until we tell you that we have received the signed papers, before you actually use the contribution. This applies both before you release the program and afterward. If you receive diffs to fix a bug, and they make significant change, we need legal papers for it. You don't need papers for changes of a few lines here or there, since they are not significant for copyright purposes. Also, you don't need papers if all you get from the suggestion is some ideas, not actual code which you use. For example, if you write a different solution to the problem, you don't need to get papers. I know this is frustrating; it's frustrating for us as well. But if you don't wait, you are going out on a limb--for example, what if the contributor's employer won't sign a disclaimer? You might have to take that code out again! The very worst thing is if you forget to tell us about the other contributor. We could be very embarrassed in court some day as a result.  File: standards.info, Node: Change Logs, Next: Compatibility, Prev: Contributions, Up: Top Change Logs *********** Keep a change log for each directory, describing the changes made to source files in that directory. The purpose of this is so that people investigating bugs in the future will know about the changes that might have introduced the bug. Often a new bug can be found by looking at what was recently changed. More importantly, change logs can help eliminate conceptual inconsistencies between different parts of a program; they can give you a history of how the conflicting concepts arose. Use the Emacs command `M-x add-change' to start a new entry in the change log. An entry should have an asterisk, the name of the changed file, and then in parentheses the name of the changed functions, variables or whatever, followed by a colon. Then describe the changes you made to that function or variable. Separate unrelated entries with blank lines. When two entries represent parts of the same change, so that they work together, then don't put blank lines between them. Then you can omit the file name and the asterisk when successive entries are in the same file. Here are some examples: * register.el (insert-register): Return nil. (jump-to-register): Likewise. * sort.el (sort-subr): Return nil. * tex-mode.el (tex-bibtex-file, tex-file, tex-region): Restart the tex shell if process is gone or stopped. (tex-shell-running): New function. * expr.c (store_one_arg): Round size up for move_block_to_reg. (expand_call): Round up when emitting USE insns. * stmt.c (assign_parms): Round size up for move_block_from_reg. It's important to name the changed function or variable in full. Don't abbreviate them; don't combine them. Subsequent maintainers will often search for a function name to find all the change log entries that pertain to it; if you abbreviate the name, they won't find it when they search. For example, some people are tempted to abbreviate groups of function names by writing `* register.el ({insert,jump-to}-register)'; this is not a good idea, since searching for `jump-to-register' or `insert-register' would not find the entry. There's no need to describe the full purpose of the changes or how they work together. It is better to put such explanations in comments in the code. That's why just "New function" is enough; there is a comment with the function in the source to explain what it does. However, sometimes it is useful to write one line to describe the overall purpose of a large batch of changes. You can think of the change log as a conceptual "undo list" which explains how earlier versions were different from the current version. People can see the current version; they don't need the change log to tell them what is in it. What they want from a change log is a clear explanation of how the earlier version differed. When you change the calling sequence of a function in a simple fashion, and you change all the callers of the function, there is no need to make individual entries for all the callers. Just write in the entry for the function being called, "All callers changed." When you change just comments or doc strings, it is enough to write an entry for the file, without mentioning the functions. Write just, "Doc fix." There's no need to keep a change log for documentation files. This is because documentation is not susceptible to bugs that are hard to fix. Documentation does not consist of parts that must interact in a precisely engineered fashion; to correct an error, you need not know the history of the erroneous passage.  File: standards.info, Node: Compatibility, Next: Makefile Conventions, Prev: Change Logs, Up: Top Compatibility with Other Implementations **************************************** With certain exceptions, utility programs and libraries for GNU should be upward compatible with those in Berkeley Unix, and upward compatible with ANSI C if ANSI C specifies their behavior, and upward compatible with POSIX if POSIX specifies their behavior. When these standards conflict, it is useful to offer compatibility modes for each of them. ANSI C and POSIX prohibit many kinds of extensions. Feel free to make the extensions anyway, and include a `--ansi' or `--compatible' option to turn them off. However, if the extension has a significant chance of breaking any real programs or scripts, then it is not really upward compatible. Try to redesign its interface. When a feature is used only by users (not by programs or command files), and it is done poorly in Unix, feel free to replace it completely with something totally different and better. (For example, vi is replaced with Emacs.) But it is nice to offer a compatible feature as well. (There is a free vi clone, so we offer it.) Additional useful features not in Berkeley Unix are welcome. Additional programs with no counterpart in Unix may be useful, but our first priority is usually to duplicate what Unix already has.  File: standards.info, Node: Makefile Conventions, Next: Configuration, Prev: Compatibility, Up: Top Makefile Conventions ******************** This chapter describes conventions for writing the Makefiles for GNU programs. * Menu: * Makefile Basics:: * Utilities in Makefiles:: * Standard Targets:: * Command Variables:: * Directory Variables::  File: standards.info, Node: Makefile Basics, Next: Utilities in Makefiles, Up: Makefile Conventions General Conventions for Makefiles ================================= Every Makefile should contain this line: SHELL = /bin/sh to avoid trouble on systems where the `SHELL' variable might be inherited from the environment. (This is never a problem with GNU `make'.) Don't assume that `.' is in the path for command execution. When you need to run programs that are a part of your package during the make, please make sure that it uses `./' if the program is built as part of the make or `$(srcdir)/' if the file is an unchanging part of the source code. Without one of these prefixes, the current search path is used. The distinction between `./' and `$(srcdir)/' is important when using the `--srcdir' option to `configure'. A rule of the form: foo.1 : foo.man sedscript sed -e sedscript foo.man > foo.1 will fail when the current directory is not the source directory, because `foo.man' and `sedscript' are not in the current directory. When using GNU `make', relying on `VPATH' to find the source file will work in the case where there is a single dependency file, since the `make' automatic variable `$<' will represent the source file wherever it is. (Many versions of `make' set `$<' only in implicit rules.) A makefile target like foo.o : bar.c $(CC) -I. -I$(srcdir) $(CFLAGS) -c bar.c -o foo.o should instead be written as foo.o : bar.c $(CC) $(CFLAGS) $< -o $@ in order to allow `VPATH' to work correctly. When the target has multiple dependencies, using an explicit `$(srcdir)' is the easiest way to make the rule work well. For example, the target above for `foo.1' is best written as: foo.1 : foo.man sedscript sed -s $(srcdir)/sedscript $(srcdir)/foo.man > foo.1  File: standards.info, Node: Utilities in Makefiles, Next: Standard Targets, Prev: Makefile Basics, Up: Makefile Conventions Utilities in Makefiles ====================== Write the Makefile commands (and any shell scripts, such as `configure') to run in `sh', not in `csh'. Don't use any special features of `ksh' or `bash'. The `configure' script and the Makefile rules for building and installation should not use any utilities directly except these: cat cmp cp echo egrep expr grep ln mkdir mv pwd rm rmdir sed test touch Stick to the generally supported options for these programs. For example, don't use `mkdir -p', convenient as it may be, because most systems don't support it. The Makefile rules for building and installation can also use compilers and related programs, but should do so via `make' variables so that the user can substitute alternatives. Here are some of the programs we mean: ar bison cc flex install ld lex make makeinfo ranlib texi2dvi yacc When you use `ranlib', you should test whether it exists, and run it only if it exists, so that the distribution will work on systems that don't have `ranlib'. If you use symbolic links, you should implement a fallback for systems that don't have symbolic links. It is ok to use other utilities in Makefile portions (or scripts) intended only for particular systems where you know those utilities to exist.  File: standards.info, Node: Standard Targets, Next: Command Variables, Prev: Utilities in Makefiles, Up: Makefile Conventions Standard Targets for Users ========================== All GNU programs should have the following targets in their Makefiles: `all' Compile the entire program. This should be the default target. This target need not rebuild any documentation files; Info files should normally be included in the distribution, and DVI files should be made only when explicitly asked for. `install' Compile the program and copy the executables, libraries, and so on to the file names where they should reside for actual use. If there is a simple test to verify that a program is properly installed, this target should run that test. The commands should create all the directories in which files are to be installed, if they don't already exist. This includes the directories specified as the values of the variables `prefix' and `exec_prefix', as well as all subdirectories that are needed. One way to do this is by means of an `installdirs' target as described below. Use `-' before any command for installing a man page, so that `make' will ignore any errors. This is in case there are systems that don't have the Unix man page documentation system installed. The way to install Info files is to copy them into `$(infodir)' with `$(INSTALL_DATA)' (*note Command Variables::.), and then run the `install-info' program if it is present. `install-info' is a script that edits the Info `dir' file to add or update the menu entry for the given Info file; it will be part of the Texinfo package. Here is a sample rule to install an Info file: $(infodir)/foo.info: foo.info # There may be a newer info file in . than in srcdir. -if test -f foo.info; then d=.; \ else d=$(srcdir); fi; \ $(INSTALL_DATA) $$d/foo.info $@; \ # Run install-info only if it exists. # Use `if' instead of just prepending `-' to the # line so we notice real errors from install-info. # We use `$(SHELL) -c' because some shells do not # fail gracefully when there is an unknown command. if $(SHELL) -c 'install-info --version' \ >/dev/null 2>&1; then \ install-info --infodir=$(infodir) $$d/foo.info; \ else true; fi `uninstall' Delete all the installed files that the `install' target would create (but not the noninstalled files such as `make all' would create). `clean' Delete all files from the current directory that are normally created by building the program. Don't delete the files that record the configuration. Also preserve files that could be made by building, but normally aren't because the distribution comes with them. Delete `.dvi' files here if they are not part of the distribution. `distclean' Delete all files from the current directory that are created by configuring or building the program. If you have unpacked the source and built the program without creating any other files, `make distclean' should leave only the files that were in the distribution. `mostlyclean' Like `clean', but may refrain from deleting a few files that people normally don't want to recompile. For example, the `mostlyclean' target for GCC does not delete `libgcc.a', because recompiling it is rarely necessary and takes a lot of time. `realclean' Delete everything from the current directory that can be reconstructed with this Makefile. This typically includes everything deleted by `distclean', plus more: C source files produced by Bison, tags tables, Info files, and so on. One exception, however: `make realclean' should not delete `configure' even if `configure' can be remade using a rule in the Makefile. More generally, `make realclean' should not delete anything that needs to exist in order to run `configure' and then begin to build the program. `TAGS' Update a tags table for this program. `info' Generate any Info files needed. The best way to write the rules is as follows: info: foo.info foo.info: foo.texi chap1.texi chap2.texi $(MAKEINFO) $(srcdir)/foo.texi You must define the variable `MAKEINFO' in the Makefile. It should run the `makeinfo' program, which is part of the Texinfo distribution. `dvi' Generate DVI files for all TeXinfo documentation. For example: dvi: foo.dvi foo.dvi: foo.texi chap1.texi chap2.texi $(TEXI2DVI) $(srcdir)/foo.texi You must define the variable `TEXI2DVI' in the Makefile. It should run the program `texi2dvi', which is part of the Texinfo distribution. Alternatively, write just the dependencies, and allow GNU Make to provide the command. `dist' Create a distribution tar file for this program. The tar file should be set up so that the file names in the tar file start with a subdirectory name which is the name of the package it is a distribution for. This name can include the version number. For example, the distribution tar file of GCC version 1.40 unpacks into a subdirectory named `gcc-1.40'. The easiest way to do this is to create a subdirectory appropriately named, use `ln' or `cp' to install the proper files in it, and then `tar' that subdirectory. The `dist' target should explicitly depend on all non-source files that are in the distribution, to make sure they are up to date in the distribution. *Note Making Releases: (standards)Releases. `check' Perform self-tests (if any). The user must build the program before running the tests, but need not install the program; you should write the self-tests so that they work when the program is built but not installed. The following targets are suggested as conventional names, for programs in which they are useful. `installcheck' Perform installation tests (if any). The user must build and install the program before running the tests. You should not assume that `$(bindir)' is in the search path. `installdirs' It's useful to add a target named `installdirs' to create the directories where files are installed, and their parent directories. There is a script called `mkinstalldirs' which is convenient for this; find it in the Texinfo package.You can use a rule like this: # Make sure all installation directories (e.g. $(bindir)) # actually exist by making them if necessary. installdirs: mkinstalldirs $(srcdir)/mkinstalldirs $(bindir) $(datadir) \ $(libdir) $(infodir) \ $(mandir)  File: standards.info, Node: Command Variables, Next: Directory Variables, Prev: Standard Targets, Up: Makefile Conventions Variables for Specifying Commands ================================= Makefiles should provide variables for overriding certain commands, options, and so on. In particular, you should run most utility programs via variables. Thus, if you use Bison, have a variable named `BISON' whose default value is set with `BISON = bison', and refer to it with `$(BISON)' whenever you need to use Bison. File management utilities such as `ln', `rm', `mv', and so on, need not be referred to through variables in this way, since users don't need to replace them with other programs. Each program-name variable should come with an options variable that is used to supply options to the program. Append `FLAGS' to the program-name variable name to get the options variable name--for example, `BISONFLAGS'. (The name `CFLAGS' is an exception to this rule, but we keep it because it is standard.) Use `CPPFLAGS' in any compilation command that runs the preprocessor, and use `LDFLAGS' in any compilation command that does linking as well as in any direct use of `ld'. If there are C compiler options that *must* be used for proper compilation of certain files, do not include them in `CFLAGS'. Users expect to be able to specify `CFLAGS' freely themselves. Instead, arrange to pass the necessary options to the C compiler independently of `CFLAGS', by writing them explicitly in the compilation commands or by defining an implicit rule, like this: CFLAGS = -g ALL_CFLAGS = -I. $(CFLAGS) .c.o: $(CC) -c $(CPPFLAGS) $(ALL_CFLAGS) $< Do include the `-g' option in `CFLAGS', because that is not *required* for proper compilation. You can consider it a default that is only recommended. If the package is set up so that it is compiled with GCC by default, then you might as well include `-O' in the default value of `CFLAGS' as well. Put `CFLAGS' last in the compilation command, after other variables containing compiler options, so the user can use `CFLAGS' to override the others. Every Makefile should define the variable `INSTALL', which is the basic command for installing a file into the system. Every Makefile should also define the variables `INSTALL_PROGRAM' and `INSTALL_DATA'. (The default for each of these should be `$(INSTALL)'.) Then it should use those variables as the commands for actual installation, for executables and nonexecutables respectively. Use these variables as follows: $(INSTALL_PROGRAM) foo $(bindir)/foo $(INSTALL_DATA) libfoo.a $(libdir)/libfoo.a Always use a file name, not a directory name, as the second argument of the installation commands. Use a separate command for each file to be installed.  File: standards.info, Node: Directory Variables, Prev: Command Variables, Up: Makefile Conventions Variables for Installation Directories ====================================== Installation directories should always be named by variables, so it is easy to install in a nonstandard place. The standard names for these variables are: `prefix' A prefix used in constructing the default values of the variables listed below. The default value of `prefix' should be `/usr/local' (at least for now). `exec_prefix' A prefix used in constructing the default values of some of the variables listed below. The default value of `exec_prefix' should be `$(prefix)'. Generally, `$(exec_prefix)' is used for directories that contain machine-specific files (such as executables and subroutine libraries), while `$(prefix)' is used directly for other directories. `bindir' The directory for installing executable programs that users can run. This should normally be `/usr/local/bin', but write it as `$(exec_prefix)/bin'. `libdir' The directory for installing executable files to be run by the program rather than by users. Object files and libraries of object code should also go in this directory. The idea is that this directory is used for files that pertain to a specific machine architecture, but need not be in the path for commands. The value of `libdir' should normally be `/usr/local/lib', but write it as `$(exec_prefix)/lib'. `datadir' The directory for installing read-only data files which the programs refer to while they run. This directory is used for files which are independent of the type of machine being used. This should normally be `/usr/local/lib', but write it as `$(prefix)/lib'. `statedir' The directory for installing data files which the programs modify while they run. These files should be independent of the type of machine being used, and it should be possible to share them among machines at a network installation. This should normally be `/usr/local/lib', but write it as `$(prefix)/lib'. `includedir' The directory for installing header files to be included by user programs with the C `#include' preprocessor directive. This should normally be `/usr/local/include', but write it as `$(prefix)/include'. Most compilers other than GCC do not look for header files in `/usr/local/include'. So installing the header files this way is only useful with GCC. Sometimes this is not a problem because some libraries are only really intended to work with GCC. But some libraries are intended to work with other compilers. They should install their header files in two places, one specified by `includedir' and one specified by `oldincludedir'. `oldincludedir' The directory for installing `#include' header files for use with compilers other than GCC. This should normally be `/usr/include'. The Makefile commands should check whether the value of `oldincludedir' is empty. If it is, they should not try to use it; they should cancel the second installation of the header files. A package should not replace an existing header in this directory unless the header came from the same package. Thus, if your Foo package provides a header file `foo.h', then it should install the header file in the `oldincludedir' directory if either (1) there is no `foo.h' there or (2) the `foo.h' that exists came from the Foo package. To tell whether `foo.h' came from the Foo package, put a magic string in the file--part of a comment--and grep for that string. `mandir' The directory for installing the man pages (if any) for this package. It should include the suffix for the proper section of the manual--usually `1' for a utility. It will normally be `/usr/local/man/man1', but you should write it as `$(prefix)/man/man1'. `man1dir' The directory for installing section 1 man pages. `man2dir' The directory for installing section 2 man pages. `...' Use these names instead of `mandir' if the package needs to install man pages in more than one section of the manual. *Don't make the primary documentation for any GNU software be a man page. Write a manual in Texinfo instead. Man pages are just for the sake of people running GNU software on Unix, which is a secondary application only.* `manext' The file name extension for the installed man page. This should contain a period followed by the appropriate digit; it should normally be `.1'. `man1ext' The file name extension for installed section 1 man pages. `man2ext' The file name extension for installed section 2 man pages. `...' Use these names instead of `manext' if the package needs to install man pages in more than one section of the manual. `infodir' The directory for installing the Info files for this package. By default, it should be `/usr/local/info', but it should be written as `$(prefix)/info'. `srcdir' The directory for the sources being compiled. The value of this variable is normally inserted by the `configure' shell script. For example: # Common prefix for installation directories. # NOTE: This directory must exist when you start the install. prefix = /usr/local exec_prefix = $(prefix) # Where to put the executable for the command `gcc'. bindir = $(exec_prefix)/bin # Where to put the directories used by the compiler. libdir = $(exec_prefix)/lib # Where to put the Info files. infodir = $(prefix)/info If your program installs a large number of files into one of the standard user-specified directories, it might be useful to group them into a subdirectory particular to that program. If you do this, you should write the `install' rule to create these subdirectories. Do not expect the user to include the subdirectory name in the value of any of the variables listed above. The idea of having a uniform set of variable names for installation directories is to enable the user to specify the exact same values for several different GNU packages. In order for this to be useful, all the packages must be designed so that they will work sensibly when the user does so.  File: standards.info, Node: Configuration, Next: Source Language, Prev: Makefile Conventions, Up: Top How Configuration Should Work ***************************** Each GNU distribution should come with a shell script named `configure'. This script is given arguments which describe the kind of machine and system you want to compile the program for. The `configure' script must record the configuration options so that they affect compilation. One way to do this is to make a link from a standard name such as `config.h' to the proper configuration file for the chosen system. If you use this technique, the distribution should *not* contain a file named `config.h'. This is so that people won't be able to build the program without configuring it first. Another thing that `configure' can do is to edit the Makefile. If you do this, the distribution should *not* contain a file named `Makefile'. Instead, include a file `Makefile.in' which contains the input used for editing. Once again, this is so that people won't be able to build the program without configuring it first. If `configure' does write the `Makefile', then `Makefile' should have a target named `Makefile' which causes `configure' to be rerun, setting up the same configuration that was set up last time. The files that `configure' reads should be listed as dependencies of `Makefile'. All the files which are output from the `configure' script should have comments at the beginning explaining that they were generated automatically using `configure'. This is so that users won't think of trying to edit them by hand. The `configure' script should write a file named `config.status' which describes which configuration options were specified when the program was last configured. This file should be a shell script which, if run, will recreate the same configuration. The `configure' script should accept an option of the form `--srcdir=DIRNAME' to specify the directory where sources are found (if it is not the current directory). This makes it possible to build the program in a separate directory, so that the actual source directory is not modified. If the user does not specify `--srcdir', then `configure' should check both `.' and `..' to see if it can find the sources. If it finds the sources in one of these places, it should use them from there. Otherwise, it should report that it cannot find the sources, and should exit with nonzero status. Usually the easy way to support `--srcdir' is by editing a definition of `VPATH' into the Makefile. Some rules may need to refer explicitly to the specified source directory. To make this possible, `configure' can add to the Makefile a variable named `srcdir' whose value is precisely the specified directory. The `configure' script should also take an argument which specifies the type of system to build the program for. This argument should look like this: CPU-COMPANY-SYSTEM For example, a Sun 3 might be `m68k-sun-sunos4.1'. The `configure' script needs to be able to decode all plausible alternatives for how to describe a machine. Thus, `sun3-sunos4.1' would be a valid alias. So would `sun3-bsd4.2', since SunOS is basically BSD and no other BSD system is used on a Sun. For many programs, `vax-dec-ultrix' would be an alias for `vax-dec-bsd', simply because the differences between Ultrix and BSD are rarely noticeable, but a few programs might need to distinguish them. There is a shell script called `config.sub' that you can use as a subroutine to validate system types and canonicalize aliases. Other options are permitted to specify in more detail the software or hardware are present on the machine: `--with-PACKAGE' The package PACKAGE will be installed, so configure this package to work with PACKAGE. Possible values of PACKAGE include `x', `gnu-as' (or `gas'), `gnu-ld', `gnu-libc', and `gdb'. `--nfp' The target machine has no floating point processor. `--gas' The target machine assembler is GAS, the GNU assembler. This is obsolete; use `--with-gnu-as' instead. `--x' The target machine has the X Window System installed. This is obsolete; use `--with-x' instead. All `configure' scripts should accept all of these "detail" options, whether or not they make any difference to the particular package at hand. In particular, they should accept any option that starts with `--with-'. This is so users will be able to configure an entire GNU source tree at once with a single set of options. Packages that perform part of compilation may support cross-compilation. In such a case, the host and target machines for the program may be different. The `configure' script should normally treat the specified type of system as both the host and the target, thus producing a program which works for the same type of machine that it runs on. The way to build a cross-compiler, cross-assembler, or what have you, is to specify the option `--host=HOSTTYPE' when running `configure'. This specifies the host system without changing the type of target system. The syntax for HOSTTYPE is the same as described above. Programs for which cross-operation is not meaningful need not accept the `--host' option, because configuring an entire operating system for cross-operation is not a meaningful thing. Some programs have ways of configuring themselves automatically. If your program is set up to do this, your `configure' script can simply ignore most of its arguments.  File: standards.info, Node: Source Language, Next: Formatting, Prev: Configuration, Up: Top Using Languages Other Than C **************************** Using a language other than C is like using a non-standard feature: it will cause trouble for users. Even if GCC supports the other language, users may find it inconvenient to have to install the compiler for that other language in order to build your program. So please write in C. There are three exceptions for this rule: * It is okay to use a special language if the same program contains an interpreter for that language. Thus, it is not a problem that GNU Emacs contains code written in Emacs Lisp, because it comes with a Lisp interpreter. * It is okay to use another language in a tool specifically intended for use with that language. This is okay because the only people who want to build the tool will be those who have installed the other language anyway. * If an application is not of extremely widespread interest, then perhaps it's not important if the application is inconvenient to install.  File: standards.info, Node: Formatting, Next: Comments, Prev: Source Language, Up: Top Formatting Your Source Code *************************** It is important to put the open-brace that starts the body of a C function in column zero, and avoid putting any other open-brace or open-parenthesis or open-bracket in column zero. Several tools look for open-braces in column zero to find the beginnings of C functions. These tools will not work on code not formatted that way. It is also important for function definitions to start the name of the function in column zero. This helps people to search for function definitions, and may also help certain tools recognize them. Thus, the proper format is this: static char * concat (s1, s2) /* Name starts in column zero here */ char *s1, *s2; { /* Open brace in column zero here */ ... } or, if you want to use ANSI C, format the definition like this: static char * concat (char *s1, char *s2) { ... } In ANSI C, if the arguments don't fit nicely on one line, split it like this: int lots_of_args (int an_integer, long a_long, short a_short, double a_double, float a_float) ... For the body of the function, we prefer code formatted like this: if (x < foo (y, z)) haha = bar[4] + 5; else { while (z) { haha += foo (z, z); z--; } return ++x + bar (); } We find it easier to read a program when it has spaces before the open-parentheses and after the commas. Especially after the commas. When you split an expression into multiple lines, split it before an operator, not after one. Here is the right way: if (foo_this_is_long && bar > win (x, y, z) && remaining_condition) Try to avoid having two operators of different precedence at the same level of indentation. For example, don't write this: mode = (inmode[j] == VOIDmode || GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j]) ? outmode[j] : inmode[j]); Instead, use extra parentheses so that the indentation shows the nesting: mode = ((inmode[j] == VOIDmode || (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j]))) ? outmode[j] : inmode[j]); Insert extra parentheses so that Emacs will indent the code properly. For example, the following indentation looks nice if you do it by hand, but Emacs would mess it up: v = rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000; But adding a set of parentheses solves the problem: v = (rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000); Format do-while statements like this: do { a = foo (a); } while (a > 0); Please use formfeed characters (control-L) to divide the program into pages at logical places (but not within a function). It does not matter just how long the pages are, since they do not have to fit on a printed page. The formfeeds should appear alone on lines by themselves.  File: standards.info, Node: Comments, Next: Syntactic Conventions, Prev: Formatting, Up: Top Commenting Your Work ******************** Every program should start with a comment saying briefly what it is for. Example: `fmt - filter for simple filling of text'. Please put a comment on each function saying what the function does, what sorts of arguments it gets, and what the possible values of arguments mean and are used for. It is not necessary to duplicate in words the meaning of the C argument declarations, if a C type is being used in its customary fashion. If there is anything nonstandard about its use (such as an argument of type `char *' which is really the address of the second character of a string, not the first), or any possible values that would not work the way one would expect (such as, that strings containing newlines are not guaranteed to work), be sure to say so. Also explain the significance of the return value, if there is one. Please put two spaces after the end of a sentence in your comments, so that the Emacs sentence commands will work. Also, please write complete sentences and capitalize the first word. If a lower-case identifer comes at the beginning of a sentence, don't capitalize it! Changing the spelling makes it a different identifier. If you don't like starting a sentence with a lower case letter, write the sentence differently (e.g., "The identifier lower-case is ..."). The comment on a function is much clearer if you use the argument names to speak about the argument values. The variable name itself should be lower case, but write it in upper case when you are speaking about the value rather than the variable itself. Thus, "the inode number NODE_NUM" rather than "an inode". There is usually no purpose in restating the name of the function in the comment before it, because the reader can see that for himself. There might be an exception when the comment is so long that the function itself would be off the bottom of the screen. There should be a comment on each static variable as well, like this: /* Nonzero means truncate lines in the display; zero means continue them. */ int truncate_lines; Every `#endif' should have a comment, except in the case of short conditionals (just a few lines) that are not nested. The comment should state the condition of the conditional that is ending, *including its sense*. `#else' should have a comment describing the condition *and sense* of the code that follows. For example: #ifdef foo ... #else /* not foo */ ... #endif /* not foo */ but, by contrast, write the comments this way for a `#ifndef': #ifndef foo ... #else /* foo */ ... #endif /* foo */  File: standards.info, Node: Syntactic Conventions, Next: Names, Prev: Comments, Up: Top Clean Use of C Constructs ************************* Please explicitly declare all arguments to functions. Don't omit them just because they are `int's. Declarations of external functions and functions to appear later in the source file should all go in one place near the beginning of the file (somewhere before the first function definition in the file), or else should go in a header file. Don't put `extern' declarations inside functions. It used to be common practice to use the same local variables (with names like `tem') over and over for different values within one function. Instead of doing this, it is better declare a separate local variable for each distinct purpose, and give it a name which is meaningful. This not only makes programs easier to understand, it also facilitates optimization by good compilers. You can also move the declaration of each local variable into the smallest scope that includes all its uses. This makes the program even cleaner. Don't use local variables or parameters that shadow global identifiers. Don't declare multiple variables in one declaration that spans lines. Start a new declaration on each line, instead. For example, instead of this: int foo, bar; write either this: int foo, bar; or this: int foo; int bar; (If they are global variables, each should have a comment preceding it anyway.) When you have an `if'-`else' statement nested in another `if' statement, always put braces around the `if'-`else'. Thus, never write like this: if (foo) if (bar) win (); else lose (); always like this: if (foo) { if (bar) win (); else lose (); } If you have an `if' statement nested inside of an `else' statement, either write `else if' on one line, like this, if (foo) ... else if (bar) ... with its `then'-part indented like the preceding `then'-part, or write the nested `if' within braces like this: if (foo) ... else { if (bar) ... } Don't declare both a structure tag and variables or typedefs in the same declaration. Instead, declare the structure tag separately and then use it to declare the variables or typedefs. Try to avoid assignments inside `if'-conditions. For example, don't write this: if ((foo = (char *) malloc (sizeof *foo)) == 0) fatal ("virtual memory exhausted"); instead, write this: foo = (char *) malloc (sizeof *foo); if (foo == 0) fatal ("virtual memory exhausted"); Don't make the program ugly to placate `lint'. Please don't insert any casts to `void'. Zero without a cast is perfectly fine as a null pointer constant.  File: standards.info, Node: Names, Next: Using Extensions, Prev: Syntactic Conventions, Up: Top Naming Variables and Functions ****************************** Please use underscores to separate words in a name, so that the Emacs word commands can be useful within them. Stick to lower case; reserve upper case for macros and `enum' constants, and for name-prefixes that follow a uniform convention. For example, you should use names like `ignore_space_change_flag'; don't use names like `iCantReadThis'. Variables that indicate whether command-line options have been specified should be named after the meaning of the option, not after the option-letter. A comment should state both the exact meaning of the option and its letter. For example, /* Ignore changes in horizontal whitespace (-b). */ int ignore_space_change_flag; When you want to define names with constant integer values, use `enum' rather than `#define'. GDB knows about enumeration constants. Use file names of 14 characters or less, to avoid creating gratuitous problems on System V.  File: standards.info, Node: Using Extensions, Next: System Functions, Prev: Names, Up: Top Using Non-standard Features *************************** Many GNU facilities that already exist support a number of convenient extensions over the comparable Unix facilities. Whether to use these extensions in implementing your program is a difficult question. On the one hand, using the extensions can make a cleaner program. On the other hand, people will not be able to build the program unless the other GNU tools are available. This might cause the program to work on fewer kinds of machines. With some extensions, it might be easy to provide both alternatives. For example, you can define functions with a "keyword" `INLINE' and define that as a macro to expand into either `inline' or nothing, depending on the compiler. In general, perhaps it is best not to use the extensions if you can straightforwardly do without them, but to use the extensions if they are a big improvement. An exception to this rule are the large, established programs (such as Emacs) which run on a great variety of systems. Such programs would be broken by use of GNU extensions. Another exception is for programs that are used as part of compilation: anything that must be compiled with other compilers in order to bootstrap the GNU compilation facilities. If these require the GNU compiler, then no one can compile them without having them installed already. That would be no good. Since most computer systems do not yet implement ANSI C, using the ANSI C features is effectively using a GNU extension, so the same considerations apply. (Except for ANSI features that we discourage, such as trigraphs--don't ever use them.)  File: standards.info, Node: System Functions, Next: Semantics, Prev: Using Extensions, Up: Top Calling System Functions ************************ C implementations differ substantially. ANSI C reduces but does not eliminate the incompatibilities; meanwhile, many users wish to compile GNU software with pre-ANSI compilers. This chapter gives recommendations for how to use the more or less standard C library functions to avoid unnecessary loss of portability. * Don't use the value of `sprintf'. It returns the number of characters written on some systems, but not on all systems. * Don't declare system functions explicitly. Almost any declaration for a system function is wrong on some system. To minimize conflicts, leave it to the system header files to declare system functions. If the headers don't declare a function, let it remain undeclared. While it may seem unclean to use a function without declaring it, in practice this works fine for most system library functions on the systems where this really happens. The problem is only theoretical. By contrast, actual declarations have frequently caused actual conflicts. * If you must declare a system function, don't specify the argument types. Use an old-style declaration, not an ANSI prototype. The more you specify about the function, the more likely a conflict. * In particular, don't unconditionally declare `malloc' or `realloc'. Most GNU programs use those functions just once, in functions conventionally named `xmalloc' and `xrealloc'. These functions call `malloc' and `realloc', respectively, and check the results. Because `xmalloc' and `xrealloc' are defined in your program, you can declare them in other files without any risk of type conflict. On most systems, `int' is the same length as a pointer; thus, the calls to `malloc' and `realloc' work fine. For the few exceptional systems (mostly 64-bit machines), you can use *conditionalized* declarations of `malloc' and `realloc'--or put these declarations in configuration files specific to those systems. * The string functions require special treatment. Some Unix systems have a header file `string.h'; other have `strings.h'. Neither file name is portable. There are two things you can do: use Autoconf to figure out which file to include, or don't include either file. * If you don't include either strings file, you can't get declarations for the string functions from the header file in the usual way. That causes less of a problem than you might think. The newer ANSI string functions are off-limits anyway because many systems still don't support them. The string functions you can use are these: strcpy strncpy strcat strncat strlen strcmp strncmp strchr strrchr The copy and concatenate functions work fine without a declaration as long as you don't use their values. Using their values without a declaration fails on systems where the width of a pointer differs from the width of `int', and perhaps in other cases. It is trivial to avoid using their values, so do that. The compare functions and `strlen' work fine without a declaration on most systems, possibly all the ones that GNU software runs on. You may find it necessary to declare them *conditionally* on a few systems. The search functions must be declared to return `char *'. Luckily, there is no variation in the data type they return. But there is variation in their names. Some systems give these functions the names `index' and `rindex'; other systems use the names `strchr' and `strrchr'. Some systems support both pairs of names, but neither pair works on all systems. You should pick a single pair of names and use it throughout your program. (Nowadays, it is better to choose `strchr' and `strrchr'.) Declare both of those names as functions returning `char *'. On systems which don't support those names, define them as macros in terms of the other pair. For example, here is what to put at the beginning of your file (or in a header) if you want to use the names `strchr' and `strrchr' throughout: #ifndef HAVE_STRCHR #define strchr index #endif #ifndef HAVE_STRRCHR #define strrchr rindex #endif char *strchr (); char *strrchr (); Here we assume that `HAVE_STRCHR' and `HAVE_STRRCHR' are macros defined in systems where the corresponding functions exist. One way to get them properly defined is to use Autoconf.  File: standards.info, Node: Semantics, Next: Errors, Prev: System Functions, Up: Top Program Behavior for All Programs ********************************* Avoid arbitrary limits on the length or number of *any* data structure, including filenames, lines, files, and symbols, by allocating all data structures dynamically. In most Unix utilities, "long lines are silently truncated". This is not acceptable in a GNU utility. Utilities reading files should not drop NUL characters, or any other nonprinting characters *including those with codes above 0177*. The only sensible exceptions would be utilities specifically intended for interface to certain types of printers that can't handle those characters. Check every system call for an error return, unless you know you wish to ignore errors. Include the system error text (from `perror' or equivalent) in *every* error message resulting from a failing system call, as well as the name of the file if any and the name of the utility. Just "cannot open foo.c" or "stat failed" is not sufficient. Check every call to `malloc' or `realloc' to see if it returned zero. Check `realloc' even if you are making the block smaller; in a system that rounds block sizes to a power of 2, `realloc' may get a different block if you ask for less space. In Unix, `realloc' can destroy the storage block if it returns zero. GNU `realloc' does not have this bug: if it fails, the original block is unchanged. Feel free to assume the bug is fixed. If you wish to run your program on Unix, and wish to avoid lossage in this case, you can use the GNU `malloc'. You must expect `free' to alter the contents of the block that was freed. Anything you want to fetch from the block, you must fetch before calling `free'. Use `getopt_long' to decode arguments, unless the argument syntax makes this unreasonable. When static storage is to be written in during program execution, use explicit C code to initialize it. Reserve C initialized declarations for data that will not be changed. Try to avoid low-level interfaces to obscure Unix data structures (such as file directories, utmp, or the layout of kernel memory), since these are less likely to work compatibly. If you need to find all the files in a directory, use `readdir' or some other high-level interface. These will be supported compatibly by GNU. By default, the GNU system will provide the signal handling functions of BSD and of POSIX. So GNU software should be written to use these. In error checks that detect "impossible" conditions, just abort. There is usually no point in printing any message. These checks indicate the existence of bugs. Whoever wants to fix the bugs will have to read the source code and run a debugger. So explain the problem with comments in the source. The relevant data will be in variables, which are easy to examine with the debugger, so there is no point moving them elsewhere.  File: standards.info, Node: Errors, Next: Libraries, Prev: Semantics, Up: Top Formatting Error Messages ************************* Error messages from compilers should look like this: SOURCE-FILE-NAME:LINENO: MESSAGE Error messages from other noninteractive programs should look like this: PROGRAM:SOURCE-FILE-NAME:LINENO: MESSAGE when there is an appropriate source file, or like this: PROGRAM: MESSAGE when there is no relevant source file. In an interactive program (one that is reading commands from a terminal), it is better not to include the program name in an error message. The place to indicate which program is running is in the prompt or with the screen layout. (When the same program runs with input from a source other than a terminal, it is not interactive and would do best to print error messages using the noninteractive style.) The string MESSAGE should not begin with a capital letter when it follows a program name and/or filename. Also, it should not end with a period. Error messages from interactive programs, and other messages such as usage messages, should start with a capital letter. But they should not end with a period.  File: standards.info, Node: Libraries, Next: Portability, Prev: Errors, Up: Top Library Behavior **************** Try to make library functions reentrant. If they need to do dynamic storage allocation, at least try to avoid any nonreentrancy aside from that of `malloc' itself. Here are certain name conventions for libraries, to avoid name conflicts. Choose a name prefix for the library, more than two characters long. All external function and variable names should start with this prefix. In addition, there should only be one of these in any given library member. This usually means putting each one in a separate source file. An exception can be made when two external symbols are always used together, so that no reasonable program could use one without the other; then they can both go in the same file. External symbols that are not documented entry points for the user should have names beginning with `_'. They should also contain the chosen name prefix for the library, to prevent collisions with other libraries. These can go in the same files with user entry points if you like. Static functions and variables can be used as you like and need not fit any naming convention.  File: standards.info, Node: Portability, Next: User Interfaces, Prev: Libraries, Up: Top Portability As It Applies to GNU ******************************** Much of what is called "portability" in the Unix world refers to porting to different Unix versions. This is a secondary consideration for GNU software, because its primary purpose is to run on top of one and only one kernel, the GNU kernel, compiled with one and only one C compiler, the GNU C compiler. The amount and kinds of variation among GNU systems on different cpu's will be like the variation among Berkeley 4.3 systems on different cpu's. All users today run GNU software on non-GNU systems. So supporting a variety of non-GNU systems is desirable; simply not paramount. The easiest way to achieve portability to a reasonable range of systems is to use Autoconf. It's unlikely that your program needs to know more information about the host machine than Autoconf can provide, simply because most of the programs that need such knowledge have already been written. It is difficult to be sure exactly what facilities the GNU kernel will provide, since it isn't finished yet. Therefore, assume you can use anything in 4.3; just avoid using the format of semi-internal data bases (e.g., directories) when there is a higher-level alternative (`readdir'). You can freely assume any reasonably standard facilities in the C language, libraries or kernel, because we will find it necessary to support these facilities in the full GNU system, whether or not we have already done so. The fact that there may exist kernels or C compilers that lack these facilities is irrelevant as long as the GNU kernel and C compiler support them. It remains necessary to worry about differences among cpu types, such as the difference in byte ordering and alignment restrictions. It's unlikely that 16-bit machines will ever be supported by GNU, so there is no point in spending any time to consider the possibility that an int will be less than 32 bits. You can assume that all pointers have the same format, regardless of the type they point to, and that this is really an integer. There are some weird machines where this isn't true, but they aren't important; don't waste time catering to them. Besides, eventually we will put function prototypes into all GNU programs, and that will probably make your program work even on weird machines. Since some important machines (including the 68000) are big-endian, it is important not to assume that the address of an `int' object is also the address of its least-significant byte. Thus, don't make the following mistake: int c; ... while ((c = getchar()) != EOF) write(file_descriptor, &c, 1); You can assume that it is reasonable to use a meg of memory. Don't strain to reduce memory usage unless it can get to that level. If your program creates complicated data structures, just make them in core and give a fatal error if malloc returns zero. If a program works by lines and could be applied to arbitrary user-supplied input files, it should keep only a line in memory, because this is not very hard and users will want to be able to operate on input files that are bigger than will fit in core all at once.  File: standards.info, Node: User Interfaces, Next: Documentation, Prev: Portability, Up: Top Standards for Command Line Interfaces ************************************* Please don't make the behavior of a utility depend on the name used to invoke it. It is useful sometimes to make a link to a utility with a different name, and that should not change what it does. Instead, use a run time option or a compilation switch or both to select among the alternate behaviors. Likewise, please don't make the behavior of the program depend on the type of output device it is used with. Device independence is an important principle of the system's design; do not compromise it merely to save someone from typing an option now and then. If you think one behavior is most useful when the output is to a terminal, and another is most useful when the output is a file or a pipe, then it is usually best to make the default behavior the one that is useful with output to a terminal, and have an option for the other behavior. Compatibility requires certain programs to depend on the type of output device. It would be disastrous if `ls' or `sh' did not do so in the way all users expect. In some of these cases, we supplement the program with a preferred alternate version that does not depend on the output device type. For example, we provide a `dir' program much like `ls' except that its default output format is always multi-column format. It is a good idea to follow the POSIX guidelines for the command-line options of a program. The easiest way to do this is to use `getopt' to parse them. Note that the GNU version of `getopt' will normally permit options anywhere among the arguments unless the special argument `--' is used. This is not what POSIX specifies; it is a GNU extension. Please define long-named options that are equivalent to the single-letter Unix-style options. We hope to make GNU more user friendly this way. This is easy to do with the GNU function `getopt_long'. One of the advantages of long-named options is that they can be consistent from program to program. For example, users should be able to expect the "verbose" option of any GNU program which has one, to be spelled precisely `--verbose'. To achieve this uniformity, look at the table of common long-option names when you choose the option names for your program. The table appears below. If you use names not already in the table, please send `gnu@prep.ai.mit.edu' a list of them, with their meanings, so we can update the table. It is usually a good idea for file names given as ordinary arguments to be input files only; any output files would be specified using options (preferably `-o'). Even if you allow an output file name as an ordinary argument for compatibility, try to provide a suitable option as well. This will lead to more consistency among GNU utilities, so that there are fewer idiosyncracies for users to remember. Programs should support an option `--version' which prints the program's version number on standard output and exits successfully, and an option `--help' which prints option usage information on standard output and exits successfully. These options should inhibit the normal function of the command; they should do nothing except print the requested information. `after-date' `-N' in `tar'. `all' `-a' in `du', `ls', `nm', `stty', `uname', and `unexpand'. `all-text' `-a' in `diff'. `almost-all' `-A' in `ls'. `append' `-a' in `etags', `tee', `time'; `-r' in `tar'. `archive' `-a' in `cp'. `ascii' `-a' in `diff'. `assume-new' `-W' in Make. `assume-old' `-o' in Make. `backward-search' `-B' in etags. `batch' Used in GDB. `baud' Used in GDB. `before' `-b' in `tac'. `binary' `-b' in `cpio' and `diff'. `block-size' Used in `cpio' and `tar'. `blocks' `-b' in `head' and `tail'. `brief' Used in various programs to make output shorter. `bytes' `-c' in `head', `split', and `tail'. `c++' `-C' in `etags'. `catenate' `-A' in `tar'. `cd' Used in various programs to specify the directory to use. `changes' `-c' in `chgrp' and `chown'. `classify' `-F' in `ls'. `command' `-c' in `su'; `-x' in GDB. `compare' `-d' in `tar'. `compress' `-z' in `tar'. `concatenate' `-A' in `tar'. `confirmation' `-w' in `tar'. `context' Used in `diff'. `core' Used in GDB. `count' `-q' in `who'. `count-links' `-l' in `du'. `create' Used in `tar' and `cpio'. `cxref' `-x' in `etags'. `date' `-d' in `touch'. `debug' `-d' in `make'; `-t' in Bison. `defines' `-d' in Bison and `etags'. `delete' `-D' in `tar'. `dereference' `-L' in `chgrp', `chown', `cpio', `du', `ls', and `tar'. `dereference-args' `-D' in `du'. `dictionary-order' `-d' in `look'. `diff' `-d' in `tar'. `directory' Specify the directory to use, in various programs. In `ls', it means to show directories themselves rather than their contents. In `rm' and `ln', it means to not treat links to directories specially. `discard-all' `-x' in `strip'. `discard-locals' `-X' in `strip'. `dry-run' `-n' in Make. `ed' `-e' in `diff'. `entire-new-file' `-N' in `diff'. `environment-overrides' `-e' in Make. `eof' `-e' in `xargs'. `epoch' Used in GDB. `error-limit' Used in Makeinfo. `escape' `-b' in `ls'. `exclude-from' `-X' in `tar'. `exec' Used in GDB. `exit' `-x' in `xargs'. `expand-tabs' `-t' in `diff'. `expression' `-e' in `sed'. `extern-only' `-g' in `nm'. `extract' `-i' in `cpio'; `-x' in `tar'. `faces' `-f' in `finger'. `fast' `-f' in `su'. `file' `-f' in `info', Make, `mt', and `tar'. `-n' in `sed'; `-r' in `touch'. `file-prefix' `-b' in Bison. `file-type' `-F' in `ls'. `files-from' `-T' in `tar'. `fill-column' Used in Makeinfo. `fixed-output-files' `-y' in Bison. `follow' `-f' in `tail'. `footnote-style' Used in Makeinfo. `force' `-f' in `cp', `ln', `mv', and `rm'. `format' Used in `ls' and `time'. `forward-search' `-F' in `etags'. `fullname' Used in GDB. `get' `-x' in `tar'. `graphic' `-i' in `ul'. `group' `-g' in `install'. `header' `-h' in `objdump'. `heading' `-H' in `who'. `help' Used to ask for brief usage information. `hide-control-chars' `-q' in `ls'. `idle' `-u' in `who'. `ifdef' `-D' in `diff'. `ignore' `-I' in `ls'. `ignore-all-space' `-w' in `diff'. `ignore-backups' `-B' in `ls'. `ignore-blank-lines' `-B' in `diff'. `ignore-case' `-f' in `look'; `-i' in `diff'. `ignore-errors' `-i' in Make. `ignore-indentation' `-S' in `etags'. `ignore-init-file' `-f' in Oleo. `ignore-interrupts' `-i' in `tee'. `ignore-matching-lines' `-I' in `diff'. `ignore-space-change' `-b' in `diff'. `ignore-zeros' `-i' in `tar'. `include' `-i' in `etags'. `include-dir' `-I' in Make. `incremental' `-G' in `tar'. `info' `-i', `-l', and `-m' in Finger. `initial' `-i' in `expand'. `initial-tab' `-T' in `diff'. `inode' `-i' in `ls'. `interactive' `-i' in `cp', `ln', `mv', `rm'; `-p' in `xargs'; `-w' in `tar'. `jobs' `-j' in Make. `just-print' `-n' in Make. `keep-going' `-k' in Make. `kilobytes' `-k' in `du' and `ls'. `line-bytes' `-C' in `split'. `lines' Used in `split', `head', and `tail'. `link' `-l' in `cpio'. `list' `-t' in `cpio'. `list' `-t' in `tar'. `literal' `-N' in `ls'. `load-average' `-l' in Make. `login' Used in `su'. `machine' `mail' `-m' in `hello'. `-m' in `uname'. `make-directories' `-d' in `cpio'. `makefile' `-f' in Make. `mapped' Used in GDB. `max-args' `-n' in `xargs'. `max-chars' `-n' in `xargs'. `max-lines' `-l' in `xargs'. `max-load' `-l' in Make. `max-procs' `-P' in `xargs'. `mesg' `-T' in `who'. `message' `-T' in `who'. `minimal' `-d' in `diff'. `mode' `-m' in `install', `mkdir', and `mkfifo'. `modification-time' `-m' in `tar'. `multi-volume' `-M' in `tar'. `name-prefix' `-a' in Bison. `new-file' `-W' in Make. `no-builtin-rules' `-r' in Make. `no-create' `-c' in `touch'. `no-defines' `-D' in `etags'. `no-dereference' `-d' in `cp'. `no-keep-going' `-S' in Make. `no-lines' `-l' in Bison. `no-prof' `-e' in `gprof'. `no-sort' `-p' in `nm'. `no-split' Used in Makeinfo. `no-static' `-a' in `gprof'. `no-time' `-E' in `gprof'. `no-validate' Used in Makeinfo. `no-warn' Used in various programs to inhibit warnings. `node' `-n' in `info'. `nodename' `-n' in `uname'. `nonmatching' `-f' in `cpio'. `nstuff' `-n' in `objdump'. `null' `-0' in `xargs'. `number' `-n' in `cat'. `number-nonblank' `-b' in `cat'. `numeric-sort' `-n' in `nm'. `numeric-uid-gid' `-n' in `cpio' and `ls'. `nx' Used in GDB. `old-archive' `-o' in `tar'. `old-file' `-o' in Make. `one-file-system' `-l' in `tar', `cp', and `du'. `only-prof' `-f' in `gprof'. `only-time' `-F' in `gprof'. `output' In various programs, specify the output file name. `override' `-o' in `rm'. `owner' `-o' in `install'. `paginate' `-l' in `diff'. `paragraph-indent' Used in Makeinfo. `parents' `-p' in `mkdir' and `rmdir'. `pass-all' `-p' in `ul'. `pass-through' `-p' in `cpio'. `port' `-P' in `finger'. `portability' `-c' in `cpio' and `tar'. `preserve' Used in `tar' and `cp'. `preserve-environment' `-p' in `su'. `preserve-modification-time' `-m' in `cpio'. `preserve-order' `-s' in `tar'. `preserve-permissions' `-p' in `tar'. `print' `-l' in `diff'. `print-chars' `-L' in `cmp'. `print-data-base' `-p' in Make. `print-directory' `-w' in Make. `print-file-name' `-o' in `nm'. `print-symdefs' `-s' in `nm'. `question' `-q' in Make. `quiet' Used in many programs to inhibit the usual output. *Note:* every program accepting `--quiet' should accept `--silent' as a synonym. `quote-name' `-Q' in `ls'. `rcs' `-n' in `diff'. `read-full-blocks' `-B' in `tar'. `readnow' Used in GDB. `recon' `-n' in Make. `record-number' `-R' in `tar'. `recursive' Used in `chgrp', `chown', `cp', `ls', `diff', and `rm'. `reference-limit' Used in Makeinfo. `regex' `-r' in `tac'. `release' `-r' in `uname'. `relocation' `-r' in `objdump'. `rename' `-r' in `cpio'. `replace' `-i' in `xargs'. `report-identical-files' `-s' in `diff'. `reset-access-time' `-a' in `cpio'. `reverse' `-r' in `ls' and `nm'. `reversed-ed' `-f' in `diff'. `same-order' `-s' in `tar'. `same-permissions' `-p' in `tar'. `save' `-g' in `stty'. `se' Used in GDB. `separate-dirs' `-S' in `du'. `separator' `-s' in `tac'. `shell' `-s' in `su'. `show-all' `-A' in `cat'. `show-c-function' `-p' in `diff'. `show-ends' `-E' in `cat'. `show-function-line' `-F' in `diff'. `show-tabs' `-T' in `cat'. `silent' Used in many programs to inhibit the usual output. *Note:* every program accepting `--silent' should accept `--quiet' as a synonym. `size' `-s' in `ls'. `sort' Used in `ls'. `sparse' `-S' in `tar'. `speed-large-files' `-H' in `diff'. `squeeze-blank' `-s' in `cat'. `starting-file' Used in `tar' and `diff' to specify which file within a directory to start processing with. `stop' `-S' in Make. `strip' `-s' in `install'. `strip-all' `-s' in `strip'. `strip-debug' `-S' in `strip'. `suffix' `-S' in `cp', `ln', `mv'. `sum' `-s' in `gprof'. `summarize' `-s' in `du'. `symbolic' `-s' in `ln'. `symbols' Used in GDB and `objdump'. `sysname' `-s' in `uname'. `tabs' `-t' in `expand' and `unexpand'. `tabsize' `-T' in `ls'. `terminal' `-T' in `tput' and `ul'. `text' `-a' in `diff'. `time' Used in `ls' and `touch'. `to-stdout' `-O' in `tar'. `total' `-c' in `du'. `touch' `-t' in Make and `ranlib'. `traditional' `-t' in `hello'. `tty' Used in GDB. `typedefs' `-t' in `etags'. `typedefs-and-c++' `-T' in `etags'. `uncompress' `-z' in `tar'. `unconditional' `-u' in `cpio'. `undefined-only' `-u' in `nm'. `update' `-u' in `cp', `etags', `mv', `tar'. `verbose' Print more information about progress. Many programs support this. `verify' `-W' in `tar'. `version' Print the version number. `version-control' `-V' in `cp', `ln', `mv'. `vgrind' `-v' in `etags'. `volume' `-V' in `tar'. `what-if' `-W' in Make. `width' `-w' in `ls'. `writable' `-T' in `who'. `zeros' `-z' in `gprof'.  File: standards.info, Node: Documentation, Next: Releases, Prev: User Interfaces, Up: Top Documenting Programs ******************** Please use Texinfo for documenting GNU programs. See the Texinfo manual, either the hardcopy or the version in the GNU Emacs Info subsystem (`C-h i'). See existing GNU Texinfo files (e.g., those under the `man/' directory in the GNU Emacs distribution) for examples. The title page of the manual should state the version of the program which the manual applies to. The Top node of the manual should also contain this information. If the manual is changing more frequently than or independent of the program, also state a version number for the manual in both of these places. The manual should document all command-line arguments and all commands. It should give examples of their use. But don't organize the manual as a list of features. Instead, organize it by the concepts a user will have before reaching that point in the manual. Address the goals that a user will have in mind, and explain how to accomplish them. Don't use Unix man pages as a model for how to write GNU documentation; they are a bad example to follow. The manual should have a node named `PROGRAM Invocation' or `Invoking PROGRAM', where PROGRAM stands for the name of the program being described, as you would type it in the shell to run the program. This node (together with its subnodes, if any) should describe the program's command line arguments and how to run it (the sort of information people would look in a man page for). Start with an `@example' containing a template for all the options and arguments that the program uses. Alternatively, put a menu item in some menu whose item name fits one of the above patterns. This identifies the node which that item points to as the node for this purpose, regardless of the node's actual name. There will be automatic features for specifying a program name and quickly reading just this part of its manual. If one manual describes several programs, it should have such a node for each program described. In addition to its manual, the package should have a file named `NEWS' which contains a list of user-visible changes worth mentioning. In each new release, add items to the front of the file and identify the version they pertain to. Don't discard old items; leave them in the file after the newer items. This way, a user upgrading from any previous version can see what is new. If the `NEWS' file gets very long, move some of the older items into a file named `ONEWS' and put a note at the end referring the user to that file. It is ok to supply a man page for the program as well as a Texinfo manual if you wish to. But keep in mind that supporting a man page requires continual effort, each time the program is changed. Any time you spend on the man page is time taken away from more useful things you could contribute. Thus, even if a user volunteers to donate a man page, you may find this gift costly to accept. Unless you have time on your hands, it may be better to refuse the man page unless the same volunteer agrees to take full responsibility for maintaining it--so that you can wash your hands of it entirely. If the volunteer ceases to do the job, then don't feel obliged to pick it up yourself; it may be better to withdraw the man page until another volunteer offers to carry on with it. Alternatively, if you expect the discrepancies to be small enough that the man page remains useful, put a prominent note near the beginning of the man page explaining that you don't maintain it and that the Texinfo manual is more authoritative, and describing how to access the Texinfo documentation.  File: standards.info, Node: Releases, Prev: Documentation, Up: Top Making Releases *************** Package the distribution of Foo version 69.96 in a gzipped tar file named `foo-69.96.tar.gz'. It should unpack into a subdirectory named `foo-69.96'. Building and installing the program should never modify any of the files contained in the distribution. This means that all the files that form part of the program in any way must be classified into "source files" and "non-source files". Source files are written by humans and never changed automatically; non-source files are produced from source files by programs under the control of the Makefile. Naturally, all the source files must be in the distribution. It is okay to include non-source files in the distribution, provided they are up-to-date and machine-independent, so that building the distribution normally will never modify them. We commonly included non-source files produced by Bison, Lex, TeX, and Makeinfo; this helps avoid unnecessary dependencies between our distributions, so that users can install whichever packages they want to install. Non-source files that might actually be modified by building and installing the program should *never* be included in the distribution. So if you do distribute non-source files, always make sure they are up to date when you make a new distribution. Make sure that the directory into which the distribution unpacks (as well as any subdirectories) are all world-writable (octal mode 777). This is so that old versions of `tar' which preserve the ownership and permissions of the files from the tar archive will be able to extract all the files even if the user is unprivileged. Make sure that no file name in the distribution is more than 14 characters long. Likewise, no file created by building the program should have a name longer than 14 characters. The reason for this is that some systems adhere to a foolish interpretation of the POSIX standard, and refuse to open a longer name, rather than truncating as they did in the past. Don't include any symbolic links in the distribution itself. If the tar file contains symbolic links, then people cannot even unpack it on systems that don't support symbolic links. Also, don't use multiple names for one file in different directories, because certain file systems cannot handle this and that prevents unpacking the distribution. Try to make sure that all the file names will be unique on MS-DOG. A name on MS-DOG consists of up to 8 characters, optionally followed by a period and up to three characters. MS-DOG will truncate extra characters both before and after the period. Thus, `foobarhacker.c' and `foobarhacker.o' are not ambiguous; they are truncated to `foobarha.c' and `foobarha.o', which are distinct. Include in your distribution a copy of the `texinfo.tex' you used to test print any `*.texinfo' files. Likewise, if your program uses small GNU software packages like regex, getopt, obstack, or termcap, include them in the distribution file. Leaving them out would make the distribution file a little smaller at the expense of possible inconvenience to a user who doesn't know what other files to get.  Tag Table: Node: Top823 Node: Reading Non-Free Code1927 Node: Contributions3653 Node: Change Logs5251 Node: Compatibility8967 Node: Makefile Conventions10373 Node: Makefile Basics10730 Node: Utilities in Makefiles12621 Node: Standard Targets14057 Node: Command Variables21166 Node: Directory Variables23996 Node: Configuration30468 Node: Source Language36035 Node: Formatting37165 Node: Comments40455 Node: Syntactic Conventions43241 Node: Names46133 Node: Using Extensions47227 Node: System Functions48968 Node: Semantics53771 Node: Errors56737 Node: Libraries57940 Node: Portability59164 Node: User Interfaces62451 Node: Documentation75766 Node: Releases79507  End Tag Table autoconf-1.7/texinfo.tex100444 74577 13 403572 5452114003 14013 0ustar peschuser%% TeX macros to handle texinfo files % Copyright (C) 1985, 86, 88, 90, 91, 92, 1993 Free Software Foundation, Inc. %This texinfo.tex file is free software; you can redistribute it and/or %modify it under the terms of the GNU General Public License as %published by the Free Software Foundation; either version 2, or (at %your option) any later version. %This texinfo.tex file is distributed in the hope that it will be %useful, but WITHOUT ANY WARRANTY; without even the implied warranty %of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU %General Public License for more details. %You should have received a copy of the GNU General Public License %along with this texinfo.tex file; see the file COPYING. If not, write %to the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, %USA. %In other words, you are welcome to use, share and improve this program. %You are forbidden to forbid anyone else to use, share and improve %what you give them. Help stamp out software-hoarding! \def\texinfoversion{2.115} \message{Loading texinfo package [Version \texinfoversion]:} % Print the version number if in a .fmt file. \everyjob{\message{[Texinfo version \texinfoversion]}\message{}} % Save some parts of plain tex whose names we will redefine. \let\ptexlbrace=\{ \let\ptexrbrace=\} \let\ptexdots=\dots \let\ptexdot=\. \let\ptexstar=\* \let\ptexend=\end \let\ptexbullet=\bullet \let\ptexb=\b \let\ptexc=\c \let\ptexi=\i \let\ptext=\t \let\ptexl=\l \let\ptexL=\L \def\tie{\penalty 10000\ } % Save plain tex definition of ~. \message{Basics,} \chardef\other=12 % If this character appears in an error message or help string, it % starts a new line in the output. \newlinechar = `^^J % Ignore a token. % \def\gobble#1{} \hyphenation{ap-pen-dix} \hyphenation{mini-buf-fer mini-buf-fers} \hyphenation{eshell} % Margin to add to right of even pages, to left of odd pages. \newdimen \bindingoffset \bindingoffset=0pt \newdimen \normaloffset \normaloffset=\hoffset \newdimen\pagewidth \newdimen\pageheight \pagewidth=\hsize \pageheight=\vsize % Sometimes it is convenient to have everything in the transcript file % and nothing on the terminal. We don't just call \tracingall here, % since that produces some useless output on the terminal. % \def\gloggingall{\begingroup \globaldefs = 1 \loggingall \endgroup}% \def\loggingall{\tracingcommands2 \tracingstats2 \tracingpages1 \tracingoutput1 \tracinglostchars1 \tracingmacros2 \tracingparagraphs1 \tracingrestores1 \showboxbreadth\maxdimen\showboxdepth\maxdimen }% %---------------------Begin change----------------------- % %%%% For @cropmarks command. % Dimensions to add cropmarks at corners Added by P. A. MacKay, 12 Nov. 1986 % \newdimen\cornerlong \newdimen\cornerthick \newdimen \topandbottommargin \newdimen \outerhsize \newdimen \outervsize \cornerlong=1pc\cornerthick=.3pt % These set size of cropmarks \outerhsize=7in %\outervsize=9.5in % Alternative @smallbook page size is 9.25in \outervsize=9.25in \topandbottommargin=.75in % %---------------------End change----------------------- % \onepageout takes a vbox as an argument. Note that \pagecontents % does insertions itself, but you have to call it yourself. \chardef\PAGE=255 \output={\onepageout{\pagecontents\PAGE}} \def\onepageout#1{\hoffset=\normaloffset \ifodd\pageno \advance\hoffset by \bindingoffset \else \advance\hoffset by -\bindingoffset\fi {\escapechar=`\\\relax % makes sure backslash is used in output files. \shipout\vbox{{\let\hsize=\pagewidth \makeheadline} \pagebody{#1}% {\let\hsize=\pagewidth \makefootline}}}% \advancepageno \ifnum\outputpenalty>-20000 \else\dosupereject\fi} %%%% For @cropmarks command %%%% % Here is a modification of the main output routine for Near East Publications % This provides right-angle cropmarks at all four corners. % The contents of the page are centerlined into the cropmarks, % and any desired binding offset is added as an \hskip on either % site of the centerlined box. (P. A. MacKay, 12 November, 1986) % \def\croppageout#1{\hoffset=0pt % make sure this doesn't mess things up {\escapechar=`\\\relax % makes sure backslash is used in output files. \shipout \vbox to \outervsize{\hsize=\outerhsize \vbox{\line{\ewtop\hfill\ewtop}} \nointerlineskip \line{\vbox{\moveleft\cornerthick\nstop} \hfill \vbox{\moveright\cornerthick\nstop}} \vskip \topandbottommargin \centerline{\ifodd\pageno\hskip\bindingoffset\fi \vbox{ {\let\hsize=\pagewidth \makeheadline} \pagebody{#1} {\let\hsize=\pagewidth \makefootline}} \ifodd\pageno\else\hskip\bindingoffset\fi} \vskip \topandbottommargin plus1fill minus1fill \boxmaxdepth\cornerthick \line{\vbox{\moveleft\cornerthick\nsbot} \hfill \vbox{\moveright\cornerthick\nsbot}} \nointerlineskip \vbox{\line{\ewbot\hfill\ewbot}} }} \advancepageno \ifnum\outputpenalty>-20000 \else\dosupereject\fi} % % Do @cropmarks to get crop marks \def\cropmarks{\let\onepageout=\croppageout } \def\pagebody#1{\vbox to\pageheight{\boxmaxdepth=\maxdepth #1}} {\catcode`\@ =11 \gdef\pagecontents#1{\ifvoid\topins\else\unvbox\topins\fi \dimen@=\dp#1 \unvbox#1 \ifvoid\footins\else\vskip\skip\footins\footnoterule \unvbox\footins\fi \ifr@ggedbottom \kern-\dimen@ \vfil \fi} } % % Here are the rules for the cropmarks. Note that they are % offset so that the space between them is truly \outerhsize or \outervsize % (P. A. MacKay, 12 November, 1986) % \def\ewtop{\vrule height\cornerthick depth0pt width\cornerlong} \def\nstop{\vbox {\hrule height\cornerthick depth\cornerlong width\cornerthick}} \def\ewbot{\vrule height0pt depth\cornerthick width\cornerlong} \def\nsbot{\vbox {\hrule height\cornerlong depth\cornerthick width\cornerthick}} % Parse an argument, then pass it to #1. The argument is the rest of % the input line (except we remove a trailing comment). #1 should be a % macro which expects an ordinary undelimited TeX argument. % \def\parsearg#1{% \let\next = #1% \begingroup \obeylines \futurelet\temp\parseargx } % If the next token is an obeyed space (from an @example environment or % the like), remove it and recurse. Otherwise, we're done. \def\parseargx{% % \obeyedspace is defined far below, after the definition of \sepspaces. \ifx\obeyedspace\temp \expandafter\parseargdiscardspace \else \expandafter\parseargline \fi } % Remove a single space (as the delimiter token to the macro call). {\obeyspaces % \gdef\parseargdiscardspace {\futurelet\temp\parseargx}} {\obeylines % \gdef\parseargline#1^^M{% \endgroup % End of the group started in \parsearg. % % First remove any @c comment, then any @comment. % Result of each macro is put in \toks0. \argremovec #1\c\relax % \expandafter\argremovecomment \the\toks0 \comment\relax % % % Call the caller's macro, saved as \next in \parsearg. \expandafter\next\expandafter{\the\toks0}% }% } % Since all \c{,omment} does is throw away the argument, we can let TeX % do that for us. The \relax here is matched by the \relax in the call % in \parseargline; it could be more or less anything, its purpose is % just to delimit the argument to the \c. \def\argremovec#1\c#2\relax{\toks0 = {#1}} \def\argremovecomment#1\comment#2\relax{\toks0 = {#1}} % \argremovec{,omment} might leave us with trailing spaces, though; e.g., % @end itemize @c foo % will have two active spaces as part of the argument with the % `itemize'. Here we remove all active spaces from #1, and assign the % result to \toks0. % % This loses if there are any *other* active characters besides spaces % in the argument -- _ ^ +, for example -- since they get expanded. % Fortunately, Texinfo does not define any such commands. (If it ever % does, the catcode of the characters in questionwill have to be changed % here.) But this means we cannot call \removeactivespaces as part of % \argremovec{,omment}, since @c uses \parsearg, and thus the argument % that \parsearg gets might well have any character at all in it. % \def\removeactivespaces#1{% \begingroup \ignoreactivespaces \edef\temp{#1}% \global\toks0 = \expandafter{\temp}% \endgroup } % Change the active space to expand to nothing. % \begingroup \obeyspaces \gdef\ignoreactivespaces{\obeyspaces\let =\empty} \endgroup \def\flushcr{\ifx\par\lisppar \def\next##1{}\else \let\next=\relax \fi \next} %% These are used to keep @begin/@end levels from running away %% Call \inENV within environments (after a \begingroup) \newif\ifENV \ENVfalse \def\inENV{\ifENV\relax\else\ENVtrue\fi} \def\ENVcheck{% \ifENV\errmessage{Still within an environment. Type Return to continue.} \endgroup\fi} % This is not perfect, but it should reduce lossage % @begin foo is the same as @foo, for now. \newhelp\EMsimple{Type to continue.} \outer\def\begin{\parsearg\beginxxx} \def\beginxxx #1{% \expandafter\ifx\csname #1\endcsname\relax {\errhelp=\EMsimple \errmessage{Undefined command @begin #1}}\else \csname #1\endcsname\fi} % @end foo executes the definition of \Efoo. % \def\end{\parsearg\endxxx} \def\endxxx #1{% \removeactivespaces{#1}% \edef\endthing{\the\toks0}% % \expandafter\ifx\csname E\endthing\endcsname\relax \expandafter\ifx\csname \endthing\endcsname\relax % There's no \foo, i.e., no ``environment'' foo. \errhelp = \EMsimple \errmessage{Undefined command `@end \endthing'}% \else \unmatchedenderror\endthing \fi \else % Everything's ok; the right environment has been started. \csname E\endthing\endcsname \fi } % There is an environment #1, but it hasn't been started. Give an error. % \def\unmatchedenderror#1{% \errhelp = \EMsimple \errmessage{This `@end #1' doesn't have a matching `@#1'}% } % Define the control sequence \E#1 to give an unmatched @end error. % \def\defineunmatchedend#1{% \expandafter\def\csname E#1\endcsname{\unmatchedenderror{#1}}% } % Single-spacing is done by various environments (specifically, in % \nonfillstart and \quotations). \newskip\singlespaceskip \singlespaceskip = \baselineskip \def\singlespace{% % Why was this kern here? It messes up equalizing space above and below % environments. --karl, 6may93 %{\advance \baselineskip by -\singlespaceskip %\kern \baselineskip}% \baselineskip=\singlespaceskip } %% Simple single-character @ commands % @@ prints an @ % Kludge this until the fonts are right (grr). \def\@{{\tt \char '100}} % This is turned off because it was never documented % and you can use @w{...} around a quote to suppress ligatures. %% Define @` and @' to be the same as ` and ' %% but suppressing ligatures. %\def\`{{`}} %\def\'{{'}} % Used to generate quoted braces. \def\mylbrace {{\tt \char '173}} \def\myrbrace {{\tt \char '175}} \let\{=\mylbrace \let\}=\myrbrace % @: forces normal size whitespace following. \def\:{\spacefactor=1000 } % @* forces a line break. \def\*{\hfil\break\hbox{}\ignorespaces} % @. is an end-of-sentence period. \def\.{.\spacefactor=3000 } % @w prevents a word break. Without the \leavevmode, @w at the % beginning of a paragraph, when TeX is still in vertical mode, would % produce a whole line of output instead of starting the paragraph. \def\w#1{\leavevmode\hbox{#1}} % @group ... @end group forces ... to be all on one page, by enclosing % it in a TeX vbox. We use \vtop instead of \vbox to construct the box % to keep its height that of a normal line. According to the rules for % \topskip (p.114 of the TeXbook), the glue inserted is % max (\topskip - \ht (first item), 0). If that height is large, % therefore, no glue is inserted, and the space between the headline and % the text is small, which looks bad. % \def\group{\begingroup \ifnum\catcode13=\active \else \errhelp = \groupinvalidhelp \errmessage{@group invalid in context where filling is enabled}% \fi % % The \vtop we start below produces a box with normal height and large % depth; thus, TeX puts \baselineskip glue before it, and (when the % next line of text is done) \lineskip glue after it. (See p.82 of % the TeXbook.) Thus, space below is not quite equal to space % above. But it's pretty close. \def\Egroup{% \egroup % End the \vtop. \endgroup % End the \group. }% % \vtop\bgroup % We have to put a strut on the last line in case the @group is in % the midst of an example, rather than completely enclosing it. % Otherwise, the interline space between the last line of the group % and the first line afterwards is too small. But we can't put the % strut in \Egroup, since there it would be on a line by itself. % Hence this just inserts a strut at the beginning of each line. \everypar = {\strut}% % % Since we have a strut on every line, we don't need any of TeX's % normal interline spacing. \offinterlineskip % % OK, but now we have to do something about blank % lines in the input in @example-like environments, which normally % just turn into \lisppar, which will insert no space now that we've % turned off the interline space. Simplest is to make them be an % empty paragraph. \ifx\par\lisppar \edef\par{\leavevmode \par}% % % Reset ^^M's definition to new definition of \par. \obeylines \fi % % We do @comment here in case we are called inside an environment, % such as @example, where each end-of-line in the input causes an % end-of-line in the output. We don't want the end-of-line after % the `@group' to put extra space in the output. Since @group % should appear on a line by itself (according to the Texinfo % manual), we don't worry about eating any user text. \comment } % % TeX puts in an \escapechar (i.e., `@') at the beginning of the help % message, so this ends up printing `@group can only ...'. % \newhelp\groupinvalidhelp{% group can only be used in environments such as @example,^^J% where each line of input produces a line of output.} % @need space-in-mils % forces a page break if there is not space-in-mils remaining. \newdimen\mil \mil=0.001in \def\need{\parsearg\needx} % Old definition--didn't work. %\def\needx #1{\par % %% This method tries to make TeX break the page naturally %% if the depth of the box does not fit. %{\baselineskip=0pt% %\vtop to #1\mil{\vfil}\kern -#1\mil\penalty 10000 %\prevdepth=-1000pt %}} \def\needx#1{% % Go into vertical mode, so we don't make a big box in the middle of a % paragraph. \par % % Don't add any leading before our big empty box, but allow a page % break, since the best break might be right here. \allowbreak \nointerlineskip \vtop to #1\mil{\vfil}% % % TeX does not even consider page breaks if a penalty added to the % main vertical list is 10000 or more. But in order to see if the % empty box we just added fits on the page, we must make it consider % page breaks. On the other hand, we don't want to actually break the % page after the empty box. So we use a penalty of 9999. % % There is an extremely small chance that TeX will actually break the % page at this \penalty, if there are no other feasible breakpoints in % sight. (If the user is using lots of big @group commands, which % almost-but-not-quite fill up a page, TeX will have a hard time doing % good page breaking, for example.) However, I could not construct an % example where a page broke at this \penalty; if it happens in a real % document, then we can reconsider our strategy. \penalty9999 % % Back up by the size of the box, whether we did a page break or not. \kern -#1\mil % % Do not allow a page break right after this kern. \nobreak } % @br forces paragraph break \let\br = \par % @dots{} output some dots \def\dots{$\ldots$} % @page forces the start of a new page \def\page{\par\vfill\supereject} % @exdent text.... % outputs text on separate line in roman font, starting at standard page margin % This records the amount of indent in the innermost environment. % That's how much \exdent should take out. \newskip\exdentamount % This defn is used inside fill environments such as @defun. \def\exdent{\parsearg\exdentyyy} \def\exdentyyy #1{{\hfil\break\hbox{\kern -\exdentamount{\rm#1}}\hfil\break}} % This defn is used inside nofill environments such as @example. \def\nofillexdent{\parsearg\nofillexdentyyy} \def\nofillexdentyyy #1{{\advance \leftskip by -\exdentamount \leftline{\hskip\leftskip{\rm#1}}}} %\hbox{{\rm#1}}\hfil\break}} % @include file insert text of that file as input. \def\include{\parsearg\includezzz} %Use \input\thisfile to avoid blank after \input, which may be an active %char (in which case the blank would become the \input argument). %The grouping keeps the value of \thisfile correct even when @include %is nested. \def\includezzz #1{\begingroup \def\thisfile{#1}\input\thisfile \endgroup} \def\thisfile{} % @center line outputs that line, centered \def\center{\parsearg\centerzzz} \def\centerzzz #1{{\advance\hsize by -\leftskip \advance\hsize by -\rightskip \centerline{#1}}} % @sp n outputs n lines of vertical space \def\sp{\parsearg\spxxx} \def\spxxx #1{\par \vskip #1\baselineskip} % @comment ...line which is ignored... % @c is the same as @comment % @ignore ... @end ignore is another way to write a comment \def\comment{\catcode 64=\other \catcode 123=\other \catcode 125=\other% \parsearg \commentxxx} \def\commentxxx #1{\catcode 64=0 \catcode 123=1 \catcode 125=2 } \let\c=\comment % Prevent errors for section commands. % Used in @ignore and in failing conditionals. \def\ignoresections{% \let\chapter=\relax \let\unnumbered=\relax \let\top=\relax \let\unnumberedsec=\relax \let\unnumberedsection=\relax \let\unnumberedsubsec=\relax \let\unnumberedsubsection=\relax \let\unnumberedsubsubsec=\relax \let\unnumberedsubsubsection=\relax \let\section=\relax \let\subsec=\relax \let\subsubsec=\relax \let\subsection=\relax \let\subsubsection=\relax \let\appendix=\relax \let\appendixsec=\relax \let\appendixsection=\relax \let\appendixsubsec=\relax \let\appendixsubsection=\relax \let\appendixsubsubsec=\relax \let\appendixsubsubsection=\relax \let\contents=\relax \let\smallbook=\relax \let\titlepage=\relax } % Used in nested conditionals, where we have to parse the Texinfo source % and so want to turn off most commands, in case they are used % incorrectly. % \def\ignoremorecommands{% \let\defcv = \relax \let\deffn = \relax \let\deffnx = \relax \let\defindex = \relax \let\defivar = \relax \let\defmac = \relax \let\defmethod = \relax \let\defop = \relax \let\defopt = \relax \let\defspec = \relax \let\deftp = \relax \let\deftypefn = \relax \let\deftypefun = \relax \let\deftypevar = \relax \let\deftypevr = \relax \let\defun = \relax \let\defvar = \relax \let\defvr = \relax \let\ref = \relax \let\xref = \relax \let\printindex = \relax \let\pxref = \relax \let\settitle = \relax \let\include = \relax \let\lowersections = \relax \let\down = \relax \let\raisesections = \relax \let\up = \relax \let\set = \relax \let\clear = \relax } % Ignore @ignore ... @end ignore. % \def\ignore{\doignore{ignore}} % Also ignore @ifinfo, @menu, and @direntry text. % \def\ifinfo{\doignore{ifinfo}} \def\menu{\doignore{menu}} \def\direntry{\doignore{direntry}} % Ignore text until a line `@end #1'. % \def\doignore#1{\begingroup % Don't complain about control sequences we have declared \outer. \ignoresections % % Define a command to swallow text until we reach `@end #1'. \long\def\doignoretext##1\end #1{\enddoignore}% % % Make sure that spaces turn into tokens that match what \doignoretext wants. \catcode32 = 10 % % And now expand that command. \doignoretext } % What we do to finish off ignored text. % \def\enddoignore{\endgroup\ignorespaces}% \newif\ifwarnedobs\warnedobsfalse \def\obstexwarn{% \ifwarnedobs\relax\else % We need to warn folks that they may have trouble with TeX 3.0. % This uses \immediate\write16 rather than \message to get newlines. \immediate\write16{} \immediate\write16{***WARNING*** for users of Unix TeX 3.0!} \immediate\write16{This manual trips a bug in TeX version 3.0 (tex hangs).} \immediate\write16{If you are running another version of TeX, relax.} \immediate\write16{If you are running Unix TeX 3.0, kill this TeX process.} \immediate\write16{ Then upgrade your TeX installation if you can.} \immediate\write16{If you are stuck with version 3.0, run the} \immediate\write16{ script ``tex3patch'' from the Texinfo distribution} \immediate\write16{ to use a workaround.} \immediate\write16{} \warnedobstrue \fi } % **In TeX 3.0, setting text in \nullfont hangs tex. For a % workaround (which requires the file ``dummy.tfm'' to be installed), % uncomment the following line: %%%%%\font\nullfont=dummy\let\obstexwarn=\relax % Ignore text, except that we keep track of conditional commands for % purposes of nesting, up to an `@end #1' command. % \def\nestedignore#1{% \obstexwarn % We must actually expand the ignored text to look for the @end % command, so that nested ignore constructs work. Thus, we put the % text into a \vbox and then do nothing with the result. To minimize % the change of memory overflow, we follow the approach outlined on % page 401 of the TeXbook: make the current font be a dummy font. % \setbox0 = \vbox\bgroup % Don't complain about control sequences we have declared \outer. \ignoresections % % Define `@end #1' to end the box, which will in turn undefine the % @end command again. \expandafter\def\csname E#1\endcsname{\egroup\ignorespaces}% % % We are going to be parsing Texinfo commands. Most cause no % trouble when they are used incorrectly, but some commands do % complicated argument parsing or otherwise get confused, so we % undefine them. % % We can't do anything about stray @-signs, unfortunately; % they'll produce `undefined control sequence' errors. \ignoremorecommands % % Set the current font to be \nullfont, a TeX primitive, and define % all the font commands to also use \nullfont. We don't use % dummy.tfm, as suggested in the TeXbook, because not all sites % might have that installed. Therefore, math mode will still % produce output, but that should be an extremely small amount of % stuff compared to the main input. % \nullfont \let\tenrm = \nullfont \let\tenit = \nullfont \let\tensl = \nullfont \let\tenbf = \nullfont \let\tentt = \nullfont \let\smallcaps = \nullfont \let\tensf = \nullfont % Similarly for index fonts (mostly for their use in % smallexample) \let\indrm = \nullfont \let\indit = \nullfont \let\indsl = \nullfont \let\indbf = \nullfont \let\indtt = \nullfont \let\indsc = \nullfont \let\indsf = \nullfont % % Don't complain when characters are missing from the fonts. \tracinglostchars = 0 % % Don't bother to do space factor calculations. \frenchspacing % % Don't report underfull hboxes. \hbadness = 10000 % % Do minimal line-breaking. \pretolerance = 10000 % % Do not execute instructions in @tex \def\tex{\doignore{tex}} } % @set VAR sets the variable VAR to an empty value. % @set VAR REST-OF-LINE sets VAR to the value REST-OF-LINE. % % Since we want to separate VAR from REST-OF-LINE (which might be % empty), we can't just use \parsearg; we have to insert a space of our % own to delimit the rest of the line, and then take it out again if we % didn't need it. % \def\set{\parsearg\setxxx} \def\setxxx#1{\setyyy#1 \endsetyyy} \def\setyyy#1 #2\endsetyyy{% \def\temp{#2}% \ifx\temp\empty \global\expandafter\let\csname SET#1\endcsname = \empty \else \setzzz{#1}#2\endsetzzz % Remove the trailing space \setxxx inserted. \fi } \def\setzzz#1#2 \endsetzzz{\expandafter\xdef\csname SET#1\endcsname{#2}} % @clear VAR clears (i.e., unsets) the variable VAR. % \def\clear{\parsearg\clearxxx} \def\clearxxx#1{\global\expandafter\let\csname SET#1\endcsname=\relax} % @value{foo} gets the text saved in variable foo. % \def\value#1{\expandafter \ifx\csname SET#1\endcsname\relax {\{No value for ``#1''\}} \else \csname SET#1\endcsname \fi} % @ifset VAR ... @end ifset reads the `...' iff VAR has been defined % with @set. % \def\ifset{\parsearg\ifsetxxx} \def\ifsetxxx #1{% \expandafter\ifx\csname SET#1\endcsname\relax \expandafter\ifsetfail \else \expandafter\ifsetsucceed \fi } \def\ifsetsucceed{\conditionalsucceed{ifset}} \def\ifsetfail{\nestedignore{ifset}} \defineunmatchedend{ifset} % @ifclear VAR ... @end ifclear reads the `...' iff VAR has never been % defined with @set, or has been undefined with @clear. % \def\ifclear{\parsearg\ifclearxxx} \def\ifclearxxx #1{% \expandafter\ifx\csname SET#1\endcsname\relax \expandafter\ifclearsucceed \else \expandafter\ifclearfail \fi } \def\ifclearsucceed{\conditionalsucceed{ifclear}} \def\ifclearfail{\nestedignore{ifclear}} \defineunmatchedend{ifclear} % @iftex always succeeds; we read the text following, through @end % iftex). But `@end iftex' should be valid only after an @iftex. % \def\iftex{\conditionalsucceed{iftex}} \defineunmatchedend{iftex} % We can't just want to start a group at @iftex (for example) and end it % at @end iftex, since then @set commands inside the conditional have no % effect (they'd get reverted at the end of the group). So we must % define \Eiftex to redefine itself to be its previous value. (We can't % just define it to fail again with an ``unmatched end'' error, since % the @ifset might be nested.) % \def\conditionalsucceed#1{% \edef\temp{% % Remember the current value of \E#1. \let\nece{prevE#1} = \nece{E#1}% % % At the `@end #1', redefine \E#1 to be its previous value. \def\nece{E#1}{\let\nece{E#1} = \nece{prevE#1}}% }% \temp } % We need to expand lots of \csname's, but we don't want to expand the % control sequences after we've constructed them. % \def\nece#1{\expandafter\noexpand\csname#1\endcsname} % @asis just yields its argument. Used with @table, for example. % \def\asis#1{#1} % @math means output in math mode. % We don't use $'s directly in the definition of \math because control % sequences like \math are expanded when the toc file is written. Then, % we read the toc file back, the $'s will be normal characters (as they % should be, according to the definition of Texinfo). So we must use a % control sequence to switch into and out of math mode. % % This isn't quite enough for @math to work properly in indices, but it % seems unlikely it will ever be needed there. % \let\implicitmath = $ \def\math#1{\implicitmath #1\implicitmath} % @bullet and @minus need the same treatment as @math, just above. \def\bullet{\implicitmath\ptexbullet\implicitmath} \def\minus{\implicitmath-\implicitmath} \def\node{\ENVcheck\parsearg\nodezzz} \def\nodezzz#1{\nodexxx [#1,]} \def\nodexxx[#1,#2]{\gdef\lastnode{#1}} \let\nwnode=\node \let\lastnode=\relax \def\donoderef{\ifx\lastnode\relax\else \expandafter\expandafter\expandafter\setref{\lastnode}\fi \let\lastnode=\relax} \def\unnumbnoderef{\ifx\lastnode\relax\else \expandafter\expandafter\expandafter\unnumbsetref{\lastnode}\fi \let\lastnode=\relax} \def\appendixnoderef{\ifx\lastnode\relax\else \expandafter\expandafter\expandafter\appendixsetref{\lastnode}\fi \let\lastnode=\relax} \let\refill=\relax % @setfilename is done at the beginning of every texinfo file. % So open here the files we need to have open while reading the input. % This makes it possible to make a .fmt file for texinfo. \def\setfilename{% \readauxfile \opencontents \openindices \fixbackslash % Turn off hack to swallow `\input texinfo'. \global\let\setfilename=\comment % Ignore extra @setfilename cmds. \comment % Ignore the actual filename. } \outer\def\bye{\pagealignmacro\tracingstats=1\ptexend} \def\inforef #1{\inforefzzz #1,,,,**} \def\inforefzzz #1,#2,#3,#4**{See Info file \file{\ignorespaces #3{}}, node \samp{\ignorespaces#1{}}} \message{fonts,} % Font-change commands. % Texinfo supports the sans serif font style, which plain TeX does not. % So we set up a \sf analogous to plain's \rm, etc. \newfam\sffam \def\sf{\fam=\sffam \tensf} \let\li = \sf % Sometimes we call it \li, not \sf. %% Try out Computer Modern fonts at \magstephalf \let\mainmagstep=\magstephalf \ifx\bigger\relax \let\mainmagstep=\magstep1 \font\textrm=cmr12 \font\texttt=cmtt12 \else \font\textrm=cmr10 scaled \mainmagstep \font\texttt=cmtt10 scaled \mainmagstep \fi % Instead of cmb10, you many want to use cmbx10. % cmbx10 is a prettier font on its own, but cmb10 % looks better when embedded in a line with cmr10. \font\textbf=cmb10 scaled \mainmagstep \font\textit=cmti10 scaled \mainmagstep \font\textsl=cmsl10 scaled \mainmagstep \font\textsf=cmss10 scaled \mainmagstep \font\textsc=cmcsc10 scaled \mainmagstep \font\texti=cmmi10 scaled \mainmagstep \font\textsy=cmsy10 scaled \mainmagstep % A few fonts for @defun, etc. \font\defbf=cmbx10 scaled \magstep1 %was 1314 \font\deftt=cmtt10 scaled \magstep1 \def\df{\let\tentt=\deftt \let\tenbf = \defbf \bf} % Fonts for indices and small examples. % We actually use the slanted font rather than the italic, % because texinfo normally uses the slanted fonts for that. % Do not make many font distinctions in general in the index, since they % aren't very useful. \font\ninett=cmtt9 \font\indrm=cmr9 \font\indit=cmsl9 \let\indsl=\indit \let\indtt=\ninett \let\indsf=\indrm \let\indbf=\indrm \let\indsc=\indrm \font\indi=cmmi9 \font\indsy=cmsy9 % Fonts for headings \font\chaprm=cmbx12 scaled \magstep2 \font\chapit=cmti12 scaled \magstep2 \font\chapsl=cmsl12 scaled \magstep2 \font\chaptt=cmtt12 scaled \magstep2 \font\chapsf=cmss12 scaled \magstep2 \let\chapbf=\chaprm \font\chapsc=cmcsc10 scaled\magstep3 \font\chapi=cmmi12 scaled \magstep2 \font\chapsy=cmsy10 scaled \magstep3 \font\secrm=cmbx12 scaled \magstep1 \font\secit=cmti12 scaled \magstep1 \font\secsl=cmsl12 scaled \magstep1 \font\sectt=cmtt12 scaled \magstep1 \font\secsf=cmss12 scaled \magstep1 \font\secbf=cmbx12 scaled \magstep1 \font\secsc=cmcsc10 scaled\magstep2 \font\seci=cmmi12 scaled \magstep1 \font\secsy=cmsy10 scaled \magstep2 % \font\ssecrm=cmbx10 scaled \magstep1 % This size an font looked bad. % \font\ssecit=cmti10 scaled \magstep1 % The letters were too crowded. % \font\ssecsl=cmsl10 scaled \magstep1 % \font\ssectt=cmtt10 scaled \magstep1 % \font\ssecsf=cmss10 scaled \magstep1 %\font\ssecrm=cmb10 scaled 1315 % Note the use of cmb rather than cmbx. %\font\ssecit=cmti10 scaled 1315 % Also, the size is a little larger than %\font\ssecsl=cmsl10 scaled 1315 % being scaled magstep1. %\font\ssectt=cmtt10 scaled 1315 %\font\ssecsf=cmss10 scaled 1315 %\let\ssecbf=\ssecrm \font\ssecrm=cmbx12 scaled \magstephalf \font\ssecit=cmti12 scaled \magstephalf \font\ssecsl=cmsl12 scaled \magstephalf \font\ssectt=cmtt12 scaled \magstephalf \font\ssecsf=cmss12 scaled \magstephalf \font\ssecbf=cmbx12 scaled \magstephalf \font\ssecsc=cmcsc10 scaled \magstep1 \font\sseci=cmmi12 scaled \magstephalf \font\ssecsy=cmsy10 scaled \magstep1 % The smallcaps and symbol fonts should actually be scaled \magstep1.5, % but that is not a standard magnification. % Fonts for title page: \font\titlerm = cmbx12 scaled \magstep3 \let\authorrm = \secrm % In order for the font changes to affect most math symbols and letters, % we have to define the \textfont of the standard families. Since % texinfo doesn't allow for producing subscripts and superscripts, we % don't bother to reset \scriptfont and \scriptscriptfont (which would % also require loading a lot more fonts). % \def\resetmathfonts{% \textfont0 = \tenrm \textfont1 = \teni \textfont2 = \tensy \textfont\itfam = \tenit \textfont\slfam = \tensl \textfont\bffam = \tenbf \textfont\ttfam = \tentt \textfont\sffam = \tensf } % The font-changing commands redefine the meanings of \tenSTYLE, instead % of just \STYLE. We do this so that font changes will continue to work % in math mode, where it is the current \fam that is relevant in most % cases, not the current. Plain TeX does, for example, % \def\bf{\fam=\bffam \tenbf} By redefining \tenbf, we obviate the need % to redefine \bf itself. \def\textfonts{% \let\tenrm=\textrm \let\tenit=\textit \let\tensl=\textsl \let\tenbf=\textbf \let\tentt=\texttt \let\smallcaps=\textsc \let\tensf=\textsf \let\teni=\texti \let\tensy=\textsy \resetmathfonts} \def\chapfonts{% \let\tenrm=\chaprm \let\tenit=\chapit \let\tensl=\chapsl \let\tenbf=\chapbf \let\tentt=\chaptt \let\smallcaps=\chapsc \let\tensf=\chapsf \let\teni=\chapi \let\tensy=\chapsy \resetmathfonts} \def\secfonts{% \let\tenrm=\secrm \let\tenit=\secit \let\tensl=\secsl \let\tenbf=\secbf \let\tentt=\sectt \let\smallcaps=\secsc \let\tensf=\secsf \let\teni=\seci \let\tensy=\secsy \resetmathfonts} \def\subsecfonts{% \let\tenrm=\ssecrm \let\tenit=\ssecit \let\tensl=\ssecsl \let\tenbf=\ssecbf \let\tentt=\ssectt \let\smallcaps=\ssecsc \let\tensf=\ssecsf \let\teni=\sseci \let\tensy=\ssecsy \resetmathfonts} \def\indexfonts{% \let\tenrm=\indrm \let\tenit=\indit \let\tensl=\indsl \let\tenbf=\indbf \let\tentt=\indtt \let\smallcaps=\indsc \let\tensf=\indsf \let\teni=\indi \let\tensy=\indsy \resetmathfonts} % Set up the default fonts, so we can use them for creating boxes. % \textfonts % Count depth in font-changes, for error checks \newcount\fontdepth \fontdepth=0 % Fonts for short table of contents. \font\shortcontrm=cmr12 \font\shortcontbf=cmbx12 \font\shortcontsl=cmsl12 %% Add scribe-like font environments, plus @l for inline lisp (usually sans %% serif) and @ii for TeX italic % \smartitalic{ARG} outputs arg in italics, followed by an italic correction % unless the following character is such as not to need one. \def\smartitalicx{\ifx\next,\else\ifx\next-\else\ifx\next.\else\/\fi\fi\fi} \def\smartitalic#1{{\sl #1}\futurelet\next\smartitalicx} \let\i=\smartitalic \let\var=\smartitalic \let\dfn=\smartitalic \let\emph=\smartitalic \let\cite=\smartitalic \def\b#1{{\bf #1}} \let\strong=\b % We can't just use \exhyphenpenalty, because that only has effect at % the end of a paragraph. Restore normal hyphenation at the end of the % group within which \nohyphenation is presumably called. % \def\nohyphenation{\hyphenchar\font = -1 \aftergroup\restorehyphenation} \def\restorehyphenation{\hyphenchar\font = `- } \def\t#1{% {\tt \nohyphenation \rawbackslash \frenchspacing #1}% \null } \let\ttfont = \t %\def\samp #1{`{\tt \rawbackslash \frenchspacing #1}'\null} \def\samp #1{`\tclose{#1}'\null} \def\key #1{{\tt \nohyphenation \uppercase{#1}}\null} \def\ctrl #1{{\tt \rawbackslash \hat}#1} \let\file=\samp % @code is a modification of @t, % which makes spaces the same size as normal in the surrounding text. \def\tclose#1{% {% % Change normal interword space to be same as for the current font. \spaceskip = \fontdimen2\font % % Switch to typewriter. \tt % % But `\ ' produces the large typewriter interword space. \def\ {{\spaceskip = 0pt{} }}% % % Turn off hyphenation. \nohyphenation % \rawbackslash \frenchspacing #1% }% \null } % We *must* turn on hyphenation at `-' and `_' in \code. % Otherwise, it is too hard to avoid overful hboxes % in the Emacs manual, the Library manual, etc. % Unfortunately, TeX uses one parameter (\hyphenchar) to control % both hyphenation at - and hyphenation within words. % We must therefore turn them both off (\tclose does that) % and arrange explicitly to hyphenate an a dash. % -- rms. { \catcode`\-=\active \catcode`\_=\active \global\def\code{\begingroup \catcode`\-=\active \let-\codedash \catcode`\_=\active \let_\codeunder \codex} % The following is used by \doprintindex to insure that long function names % wrap around. It is necessary for - and _ to be active before the index is % read from the file, as \entry parses the arguments long before \code is % ever called. -- mycroft \global\def\indexbreaks{\catcode`\-=\active \let-\realdash \catcode`\_=\active \let_\realunder} } \def\realdash{-} \def\realunder{_} \def\codedash{-\discretionary{}{}{}} \def\codeunder{\normalunderscore\discretionary{}{}{}} \def\codex #1{\tclose{#1}\endgroup} %\let\exp=\tclose %Was temporary % @kbd is like @code, except that if the argument is just one @key command, % then @kbd has no effect. \def\xkey{\key} \def\kbdfoo#1#2#3\par{\def\one{#1}\def\three{#3}\def\threex{??}% \ifx\one\xkey\ifx\threex\three \key{#2}% \else\tclose{\look}\fi \else\tclose{\look}\fi} % Typeset a dimension, e.g., `in' or `pt'. The only reason for the % argument is to make the input look right: @dmn{pt} instead of % @dmn{}pt. % \def\dmn#1{\thinspace #1} \def\kbd#1{\def\look{#1}\expandafter\kbdfoo\look??\par} \def\l#1{{\li #1}\null} % \def\r#1{{\rm #1}} % roman font % Use of \lowercase was suggested. \def\sc#1{{\smallcaps#1}} % smallcaps font \def\ii#1{{\it #1}} % italic font \message{page headings,} \newskip\titlepagetopglue \titlepagetopglue = 1.5in \newskip\titlepagebottomglue \titlepagebottomglue = 2pc % First the title page. Must do @settitle before @titlepage. \def\titlefont#1{{\titlerm #1}} \newif\ifseenauthor \newif\iffinishedtitlepage \def\shorttitlepage{\parsearg\shorttitlepagezzz} \def\shorttitlepagezzz #1{\begingroup\hbox{}\vskip 1.5in \chaprm \centerline{#1}% \endgroup\page\hbox{}\page} \def\titlepage{\begingroup \parindent=0pt \textfonts \let\subtitlerm=\tenrm % I deinstalled the following change because \cmr12 is undefined. % This change was not in the ChangeLog anyway. --rms. % \let\subtitlerm=\cmr12 \def\subtitlefont{\subtitlerm \normalbaselineskip = 13pt \normalbaselines}% % \def\authorfont{\authorrm \normalbaselineskip = 16pt \normalbaselines}% % % Leave some space at the very top of the page. \vglue\titlepagetopglue % % Now you can print the title using @title. \def\title{\parsearg\titlezzz}% \def\titlezzz##1{\leftline{\titlefont{##1}} % print a rule at the page bottom also. \finishedtitlepagefalse \vskip4pt \hrule height 4pt width \hsize \vskip4pt}% % No rule at page bottom unless we print one at the top with @title. \finishedtitlepagetrue % % Now you can put text using @subtitle. \def\subtitle{\parsearg\subtitlezzz}% \def\subtitlezzz##1{{\subtitlefont \rightline{##1}}}% % % @author should come last, but may come many times. \def\author{\parsearg\authorzzz}% \def\authorzzz##1{\ifseenauthor\else\vskip 0pt plus 1filll\seenauthortrue\fi {\authorfont \leftline{##1}}}% % % Most title ``pages'' are actually two pages long, with space % at the top of the second. We don't want the ragged left on the second. \let\oldpage = \page \def\page{% \iffinishedtitlepage\else \finishtitlepage \fi \oldpage \let\page = \oldpage \hbox{}}% % \def\page{\oldpage \hbox{}} } \def\Etitlepage{% \iffinishedtitlepage\else \finishtitlepage \fi % It is important to do the page break before ending the group, % because the headline and footline are only empty inside the group. % If we use the new definition of \page, we always get a blank page % after the title page, which we certainly don't want. \oldpage \endgroup \HEADINGSon } \def\finishtitlepage{% \vskip4pt \hrule height 2pt width \hsize \vskip\titlepagebottomglue \finishedtitlepagetrue } %%% Set up page headings and footings. \let\thispage=\folio \newtoks \evenheadline % Token sequence for heading line of even pages \newtoks \oddheadline % Token sequence for heading line of odd pages \newtoks \evenfootline % Token sequence for footing line of even pages \newtoks \oddfootline % Token sequence for footing line of odd pages % Now make Tex use those variables \headline={{\textfonts\rm \ifodd\pageno \the\oddheadline \else \the\evenheadline \fi}} \footline={{\textfonts\rm \ifodd\pageno \the\oddfootline \else \the\evenfootline \fi}\HEADINGShook} \let\HEADINGShook=\relax % Commands to set those variables. % For example, this is what @headings on does % @evenheading @thistitle|@thispage|@thischapter % @oddheading @thischapter|@thispage|@thistitle % @evenfooting @thisfile|| % @oddfooting ||@thisfile \def\evenheading{\parsearg\evenheadingxxx} \def\oddheading{\parsearg\oddheadingxxx} \def\everyheading{\parsearg\everyheadingxxx} \def\evenfooting{\parsearg\evenfootingxxx} \def\oddfooting{\parsearg\oddfootingxxx} \def\everyfooting{\parsearg\everyfootingxxx} {\catcode`\@=0 % \gdef\evenheadingxxx #1{\evenheadingyyy #1@|@|@|@|\finish} \gdef\evenheadingyyy #1@|#2@|#3@|#4\finish{% \global\evenheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \gdef\oddheadingxxx #1{\oddheadingyyy #1@|@|@|@|\finish} \gdef\oddheadingyyy #1@|#2@|#3@|#4\finish{% \global\oddheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \gdef\everyheadingxxx #1{\everyheadingyyy #1@|@|@|@|\finish} \gdef\everyheadingyyy #1@|#2@|#3@|#4\finish{% \global\evenheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}} \global\oddheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \gdef\evenfootingxxx #1{\evenfootingyyy #1@|@|@|@|\finish} \gdef\evenfootingyyy #1@|#2@|#3@|#4\finish{% \global\evenfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \gdef\oddfootingxxx #1{\oddfootingyyy #1@|@|@|@|\finish} \gdef\oddfootingyyy #1@|#2@|#3@|#4\finish{% \global\oddfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \gdef\everyfootingxxx #1{\everyfootingyyy #1@|@|@|@|\finish} \gdef\everyfootingyyy #1@|#2@|#3@|#4\finish{% \global\evenfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}} \global\oddfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} % }% unbind the catcode of @. % @headings double turns headings on for double-sided printing. % @headings single turns headings on for single-sided printing. % @headings off turns them off. % @headings on same as @headings double, retained for compatibility. % @headings after turns on double-sided headings after this page. % @headings doubleafter turns on double-sided headings after this page. % @headings singleafter turns on single-sided headings after this page. % By default, they are off. \def\headings #1 {\csname HEADINGS#1\endcsname} \def\HEADINGSoff{ \global\evenheadline={\hfil} \global\evenfootline={\hfil} \global\oddheadline={\hfil} \global\oddfootline={\hfil}} \HEADINGSoff % When we turn headings on, set the page number to 1. % For double-sided printing, put current file name in lower left corner, % chapter name on inside top of right hand pages, document % title on inside top of left hand pages, and page numbers on outside top % edge of all pages. \def\HEADINGSdouble{ %\pagealignmacro \global\pageno=1 \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\folio\hfil\thistitle}} \global\oddheadline={\line{\thischapter\hfil\folio}} } % For single-sided printing, chapter title goes across top left of page, % page number on top right. \def\HEADINGSsingle{ %\pagealignmacro \global\pageno=1 \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\thischapter\hfil\folio}} \global\oddheadline={\line{\thischapter\hfil\folio}} } \def\HEADINGSon{\HEADINGSdouble} \def\HEADINGSafter{\let\HEADINGShook=\HEADINGSdoublex} \let\HEADINGSdoubleafter=\HEADINGSafter \def\HEADINGSdoublex{% \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\folio\hfil\thistitle}} \global\oddheadline={\line{\thischapter\hfil\folio}} } \def\HEADINGSsingleafter{\let\HEADINGShook=\HEADINGSsinglex} \def\HEADINGSsinglex{% \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\thischapter\hfil\folio}} \global\oddheadline={\line{\thischapter\hfil\folio}} } % Subroutines used in generating headings % Produces Day Month Year style of output. \def\today{\number\day\space \ifcase\month\or January\or February\or March\or April\or May\or June\or July\or August\or September\or October\or November\or December\fi \space\number\year} % Use this if you want the Month Day, Year style of output. %\def\today{\ifcase\month\or %January\or February\or March\or April\or May\or June\or %July\or August\or September\or October\or November\or December\fi %\space\number\day, \number\year} % @settitle line... specifies the title of the document, for headings % It generates no output of its own \def\thistitle{No Title} \def\settitle{\parsearg\settitlezzz} \def\settitlezzz #1{\gdef\thistitle{#1}} \message{tables,} % @tabs -- simple alignment % These don't work. For one thing, \+ is defined as outer. % So these macros cannot even be defined. %\def\tabs{\parsearg\tabszzz} %\def\tabszzz #1{\settabs\+#1\cr} %\def\tabline{\parsearg\tablinezzz} %\def\tablinezzz #1{\+#1\cr} %\def\&{&} % Tables -- @table, @ftable, @vtable, @item(x), @kitem(x), @xitem(x). % default indentation of table text \newdimen\tableindent \tableindent=.8in % default indentation of @itemize and @enumerate text \newdimen\itemindent \itemindent=.3in % margin between end of table item and start of table text. \newdimen\itemmargin \itemmargin=.1in % used internally for \itemindent minus \itemmargin \newdimen\itemmax % Note @table, @vtable, and @vtable define @item, @itemx, etc., with % these defs. % They also define \itemindex % to index the item name in whatever manner is desired (perhaps none). \newif\ifitemxneedsnegativevskip \def\itemxpar{\par\ifitemxneedsnegativevskip\vskip-\parskip\nobreak\fi} \def\internalBitem{\smallbreak \parsearg\itemzzz} \def\internalBitemx{\itemxpar \parsearg\itemzzz} \def\internalBxitem "#1"{\def\xitemsubtopix{#1} \smallbreak \parsearg\xitemzzz} \def\internalBxitemx "#1"{\def\xitemsubtopix{#1} \itemxpar \parsearg\xitemzzz} \def\internalBkitem{\smallbreak \parsearg\kitemzzz} \def\internalBkitemx{\itemxpar \parsearg\kitemzzz} \def\kitemzzz #1{\dosubind {kw}{\code{#1}}{for {\bf \lastfunction}}% \itemzzz {#1}} \def\xitemzzz #1{\dosubind {kw}{\code{#1}}{for {\bf \xitemsubtopic}}% \itemzzz {#1}} \def\itemzzz #1{\begingroup % \advance\hsize by -\rightskip \advance\hsize by -\tableindent \setbox0=\hbox{\itemfont{#1}}% \itemindex{#1}% \nobreak % This prevents a break before @itemx. % % Be sure we are not still in the middle of a paragraph. %{\parskip = 0in %\par %}% % % If the item text does not fit in the space we have, put it on a line % by itself, and do not allow a page break either before or after that % line. We do not start a paragraph here because then if the next % command is, e.g., @kindex, the whatsit would get put into the % horizontal list on a line by itself, resulting in extra blank space. \ifdim \wd0>\itemmax % % Make this a paragraph so we get the \parskip glue and wrapping, % but leave it ragged-right. \begingroup \advance\leftskip by-\tableindent \advance\hsize by\tableindent \advance\rightskip by0pt plus1fil \leavevmode\unhbox0\par \endgroup % % We're going to be starting a paragraph, but we don't want the % \parskip glue -- logically it's part of the @item we just started. \nobreak \vskip-\parskip % % Stop a page break at the \parskip glue coming up. Unfortunately % we can't prevent a possible page break at the following % \baselineskip glue. \nobreak \endgroup \itemxneedsnegativevskipfalse \else % The item text fits into the space. Start a paragraph, so that the % following text (if any) will end up on the same line. Since that % text will be indented by \tableindent, we make the item text be in % a zero-width box. \noindent \rlap{\hskip -\tableindent\box0}\ignorespaces% \endgroup% \itemxneedsnegativevskiptrue% \fi } \def\item{\errmessage{@item while not in a table}} \def\itemx{\errmessage{@itemx while not in a table}} \def\kitem{\errmessage{@kitem while not in a table}} \def\kitemx{\errmessage{@kitemx while not in a table}} \def\xitem{\errmessage{@xitem while not in a table}} \def\xitemx{\errmessage{@xitemx while not in a table}} %% Contains a kludge to get @end[description] to work \def\description{\tablez{\dontindex}{1}{}{}{}{}} \def\table{\begingroup\inENV\obeylines\obeyspaces\tablex} {\obeylines\obeyspaces% \gdef\tablex #1^^M{% \tabley\dontindex#1 \endtabley}} \def\ftable{\begingroup\inENV\obeylines\obeyspaces\ftablex} {\obeylines\obeyspaces% \gdef\ftablex #1^^M{% \tabley\fnitemindex#1 \endtabley \def\Eftable{\endgraf\afterenvbreak\endgroup}% \let\Etable=\relax}} \def\vtable{\begingroup\inENV\obeylines\obeyspaces\vtablex} {\obeylines\obeyspaces% \gdef\vtablex #1^^M{% \tabley\vritemindex#1 \endtabley \def\Evtable{\endgraf\afterenvbreak\endgroup}% \let\Etable=\relax}} \def\dontindex #1{} \def\fnitemindex #1{\doind {fn}{\code{#1}}}% \def\vritemindex #1{\doind {vr}{\code{#1}}}% {\obeyspaces % \gdef\tabley#1#2 #3 #4 #5 #6 #7\endtabley{\endgroup% \tablez{#1}{#2}{#3}{#4}{#5}{#6}}} \def\tablez #1#2#3#4#5#6{% \aboveenvbreak % \begingroup % \def\Edescription{\Etable}% Neccessary kludge. \let\itemindex=#1% \ifnum 0#3>0 \advance \leftskip by #3\mil \fi % \ifnum 0#4>0 \tableindent=#4\mil \fi % \ifnum 0#5>0 \advance \rightskip by #5\mil \fi % \def\itemfont{#2}% \itemmax=\tableindent % \advance \itemmax by -\itemmargin % \advance \leftskip by \tableindent % \exdentamount=\tableindent \parindent = 0pt \parskip = \smallskipamount \ifdim \parskip=0pt \parskip=2pt \fi% \def\Etable{\endgraf\afterenvbreak\endgroup}% \let\item = \internalBitem % \let\itemx = \internalBitemx % \let\kitem = \internalBkitem % \let\kitemx = \internalBkitemx % \let\xitem = \internalBxitem % \let\xitemx = \internalBxitemx % } % This is the counter used by @enumerate, which is really @itemize \newcount \itemno \def\itemize{\parsearg\itemizezzz} \def\itemizezzz #1{% \begingroup % ended by the @end itemsize \itemizey {#1}{\Eitemize} } \def\itemizey #1#2{% \aboveenvbreak % \itemmax=\itemindent % \advance \itemmax by -\itemmargin % \advance \leftskip by \itemindent % \exdentamount=\itemindent \parindent = 0pt % \parskip = \smallskipamount % \ifdim \parskip=0pt \parskip=2pt \fi% \def#2{\endgraf\afterenvbreak\endgroup}% \def\itemcontents{#1}% \let\item=\itemizeitem} % Set sfcode to normal for the chars that usually have another value. % These are `.?!:;,' \def\frenchspacing{\sfcode46=1000 \sfcode63=1000 \sfcode33=1000 \sfcode58=1000 \sfcode59=1000 \sfcode44=1000 } % \splitoff TOKENS\endmark defines \first to be the first token in % TOKENS, and \rest to be the remainder. % \def\splitoff#1#2\endmark{\def\first{#1}\def\rest{#2}}% % Allow an optional argument of an uppercase letter, lowercase letter, % or number, to specify the first label in the enumerated list. No % argument is the same as `1'. % \def\enumerate{\parsearg\enumeratezzz} \def\enumeratezzz #1{\enumeratey #1 \endenumeratey} \def\enumeratey #1 #2\endenumeratey{% \begingroup % ended by the @end enumerate % % If we were given no argument, pretend we were given `1'. \def\thearg{#1}% \ifx\thearg\empty \def\thearg{1}\fi % % Detect if the argument is a single token. If so, it might be a % letter. Otherwise, the only valid thing it can be is a number. % (We will always have one token, because of the test we just made. % This is a good thing, since \splitoff doesn't work given nothing at % all -- the first parameter is undelimited.) \expandafter\splitoff\thearg\endmark \ifx\rest\empty % Only one token in the argument. It could still be anything. % A ``lowercase letter'' is one whose \lccode is nonzero. % An ``uppercase letter'' is one whose \lccode is both nonzero, and % not equal to itself. % Otherwise, we assume it's a number. % % We need the \relax at the end of the \ifnum lines to stop TeX from % continuing to look for a . % \ifnum\lccode\expandafter`\thearg=0\relax \numericenumerate % a number (we hope) \else % It's a letter. \ifnum\lccode\expandafter`\thearg=\expandafter`\thearg\relax \lowercaseenumerate % lowercase letter \else \uppercaseenumerate % uppercase letter \fi \fi \else % Multiple tokens in the argument. We hope it's a number. \numericenumerate \fi } % An @enumerate whose labels are integers. The starting integer is % given in \thearg. % \def\numericenumerate{% \itemno = \thearg \startenumeration{\the\itemno}% } % The starting (lowercase) letter is in \thearg. \def\lowercaseenumerate{% \itemno = \expandafter`\thearg \startenumeration{% % Be sure we're not beyond the end of the alphabet. \ifnum\itemno=0 \errmessage{No more lowercase letters in @enumerate; get a bigger alphabet}% \fi \char\lccode\itemno }% } % The starting (uppercase) letter is in \thearg. \def\uppercaseenumerate{% \itemno = \expandafter`\thearg \startenumeration{% % Be sure we're not beyond the end of the alphabet. \ifnum\itemno=0 \errmessage{No more uppercase letters in @enumerate; get a bigger alphabet} \fi \char\uccode\itemno }% } % Call itemizey, adding a period to the first argument and supplying the % common last two arguments. Also subtract one from the initial value in % \itemno, since @item increments \itemno. % \def\startenumeration#1{% \advance\itemno by -1 \itemizey{#1.}\Eenumerate\flushcr } % @alphaenumerate and @capsenumerate are abbreviations for giving an arg % to @enumerate. % \def\alphaenumerate{\enumerate{a}} \def\capsenumerate{\enumerate{A}} \def\Ealphaenumerate{\Eenumerate} \def\Ecapsenumerate{\Eenumerate} % Definition of @item while inside @itemize. \def\itemizeitem{% \advance\itemno by 1 {\let\par=\endgraf \smallbreak}% \ifhmode \errmessage{\in hmode at itemizeitem}\fi {\parskip=0in \hskip 0pt \hbox to 0pt{\hss \itemcontents\hskip \itemmargin}% \vadjust{\penalty 1200}}% \flushcr} \message{indexing,} % Index generation facilities % Define \newwrite to be identical to plain tex's \newwrite % except not \outer, so it can be used within \newindex. {\catcode`\@=11 \gdef\newwrite{\alloc@7\write\chardef\sixt@@n}} % \newindex {foo} defines an index named foo. % It automatically defines \fooindex such that % \fooindex ...rest of line... puts an entry in the index foo. % It also defines \fooindfile to be the number of the output channel for % the file that accumulates this index. The file's extension is foo. % The name of an index should be no more than 2 characters long % for the sake of vms. \def\newindex #1{ \expandafter\newwrite \csname#1indfile\endcsname% Define number for output file \openout \csname#1indfile\endcsname \jobname.#1 % Open the file \expandafter\xdef\csname#1index\endcsname{% % Define \xxxindex \noexpand\doindex {#1}} } % @defindex foo == \newindex{foo} \def\defindex{\parsearg\newindex} % Define @defcodeindex, like @defindex except put all entries in @code. \def\newcodeindex #1{ \expandafter\newwrite \csname#1indfile\endcsname% Define number for output file \openout \csname#1indfile\endcsname \jobname.#1 % Open the file \expandafter\xdef\csname#1index\endcsname{% % Define \xxxindex \noexpand\docodeindex {#1}} } \def\defcodeindex{\parsearg\newcodeindex} % @synindex foo bar makes index foo feed into index bar. % Do this instead of @defindex foo if you don't want it as a separate index. \def\synindex #1 #2 {% \expandafter\let\expandafter\synindexfoo\expandafter=\csname#2indfile\endcsname \expandafter\let\csname#1indfile\endcsname=\synindexfoo \expandafter\xdef\csname#1index\endcsname{% % Define \xxxindex \noexpand\doindex {#2}}% } % @syncodeindex foo bar similar, but put all entries made for index foo % inside @code. \def\syncodeindex #1 #2 {% \expandafter\let\expandafter\synindexfoo\expandafter=\csname#2indfile\endcsname \expandafter\let\csname#1indfile\endcsname=\synindexfoo \expandafter\xdef\csname#1index\endcsname{% % Define \xxxindex \noexpand\docodeindex {#2}}% } % Define \doindex, the driver for all \fooindex macros. % Argument #1 is generated by the calling \fooindex macro, % and it is "foo", the name of the index. % \doindex just uses \parsearg; it calls \doind for the actual work. % This is because \doind is more useful to call from other macros. % There is also \dosubind {index}{topic}{subtopic} % which makes an entry in a two-level index such as the operation index. \def\doindex#1{\edef\indexname{#1}\parsearg\singleindexer} \def\singleindexer #1{\doind{\indexname}{#1}} % like the previous two, but they put @code around the argument. \def\docodeindex#1{\edef\indexname{#1}\parsearg\singlecodeindexer} \def\singlecodeindexer #1{\doind{\indexname}{\code{#1}}} \def\indexdummies{% \def\_{{\realbackslash _}}% \def\w{\realbackslash w }% \def\bf{\realbackslash bf }% \def\rm{\realbackslash rm }% \def\sl{\realbackslash sl }% \def\sf{\realbackslash sf}% \def\tt{\realbackslash tt}% \def\gtr{\realbackslash gtr}% \def\less{\realbackslash less}% \def\hat{\realbackslash hat}% \def\char{\realbackslash char}% \def\TeX{\realbackslash TeX}% \def\dots{\realbackslash dots }% \def\copyright{\realbackslash copyright }% \def\tclose##1{\realbackslash tclose {##1}}% \def\code##1{\realbackslash code {##1}}% \def\samp##1{\realbackslash samp {##1}}% \def\t##1{\realbackslash r {##1}}% \def\r##1{\realbackslash r {##1}}% \def\i##1{\realbackslash i {##1}}% \def\b##1{\realbackslash b {##1}}% \def\cite##1{\realbackslash cite {##1}}% \def\key##1{\realbackslash key {##1}}% \def\file##1{\realbackslash file {##1}}% \def\var##1{\realbackslash var {##1}}% \def\kbd##1{\realbackslash kbd {##1}}% \def\dfn##1{\realbackslash dfn {##1}}% \def\emph##1{\realbackslash emph {##1}}% } % \indexnofonts no-ops all font-change commands. % This is used when outputting the strings to sort the index by. \def\indexdummyfont#1{#1} \def\indexdummytex{TeX} \def\indexdummydots{...} \def\indexnofonts{% \let\w=\indexdummyfont \let\t=\indexdummyfont \let\r=\indexdummyfont \let\i=\indexdummyfont \let\b=\indexdummyfont \let\emph=\indexdummyfont \let\strong=\indexdummyfont \let\cite=\indexdummyfont \let\sc=\indexdummyfont %Don't no-op \tt, since it isn't a user-level command % and is used in the definitions of the active chars like <, >, |... %\let\tt=\indexdummyfont \let\tclose=\indexdummyfont \let\code=\indexdummyfont \let\file=\indexdummyfont \let\samp=\indexdummyfont \let\kbd=\indexdummyfont \let\key=\indexdummyfont \let\var=\indexdummyfont \let\TeX=\indexdummytex \let\dots=\indexdummydots } % To define \realbackslash, we must make \ not be an escape. % We must first make another character (@) an escape % so we do not become unable to do a definition. {\catcode`\@=0 \catcode`\\=\other @gdef@realbackslash{\}} \let\indexbackslash=0 %overridden during \printindex. \def\doind #1#2{% {\count10=\lastpenalty % {\indexdummies % Must do this here, since \bf, etc expand at this stage \escapechar=`\\% {\let\folio=0% Expand all macros now EXCEPT \folio \def\rawbackslashxx{\indexbackslash}% \indexbackslash isn't defined now % so it will be output as is; and it will print as backslash in the indx. % % Now process the index-string once, with all font commands turned off, % to get the string to sort the index by. {\indexnofonts \xdef\temp1{#2}% }% % Now produce the complete index entry. We process the index-string again, % this time with font commands expanded, to get what to print in the index. \edef\temp{% \write \csname#1indfile\endcsname{% \realbackslash entry {\temp1}{\folio}{#2}}}% \temp }% }\penalty\count10}} \def\dosubind #1#2#3{% {\count10=\lastpenalty % {\indexdummies % Must do this here, since \bf, etc expand at this stage \escapechar=`\\% {\let\folio=0% \def\rawbackslashxx{\indexbackslash}% % % Now process the index-string once, with all font commands turned off, % to get the string to sort the index by. {\indexnofonts \xdef\temp1{#2 #3}% }% % Now produce the complete index entry. We process the index-string again, % this time with font commands expanded, to get what to print in the index. \edef\temp{% \write \csname#1indfile\endcsname{% \realbackslash entry {\temp1}{\folio}{#2}{#3}}}% \temp }% }\penalty\count10}} % The index entry written in the file actually looks like % \entry {sortstring}{page}{topic} % or % \entry {sortstring}{page}{topic}{subtopic} % The texindex program reads in these files and writes files % containing these kinds of lines: % \initial {c} % before the first topic whose initial is c % \entry {topic}{pagelist} % for a topic that is used without subtopics % \primary {topic} % for the beginning of a topic that is used with subtopics % \secondary {subtopic}{pagelist} % for each subtopic. % Define the user-accessible indexing commands % @findex, @vindex, @kindex, @cindex. \def\findex {\fnindex} \def\kindex {\kyindex} \def\cindex {\cpindex} \def\vindex {\vrindex} \def\tindex {\tpindex} \def\pindex {\pgindex} \def\cindexsub {\begingroup\obeylines\cindexsub} {\obeylines % \gdef\cindexsub "#1" #2^^M{\endgroup % \dosubind{cp}{#2}{#1}}} % Define the macros used in formatting output of the sorted index material. % This is what you call to cause a particular index to get printed. % Write % @unnumbered Function Index % @printindex fn \def\printindex{\parsearg\doprintindex} \def\doprintindex#1{% \tex \dobreak \chapheadingskip {10000} \catcode`\%=\other\catcode`\&=\other\catcode`\#=\other \catcode`\$=\other \catcode`\~=\other \indexbreaks % % The following don't help, since the chars were translated % when the raw index was written, and their fonts were discarded % due to \indexnofonts. %\catcode`\"=\active %\catcode`\^=\active %\catcode`\_=\active %\catcode`\|=\active %\catcode`\<=\active %\catcode`\>=\active % % \def\indexbackslash{\rawbackslashxx} \indexfonts\rm \tolerance=9500 \advance\baselineskip -1pt \begindoublecolumns % % See if the index file exists and is nonempty. \openin 1 \jobname.#1s \ifeof 1 % \enddoublecolumns gets confused if there is no text in the index, % and it loses the chapter title and the aux file entries for the % index. The easiest way to prevent this problem is to make sure % there is some text. (Index is nonexistent) \else % % If the index file exists but is empty, then \openin leaves \ifeof % false. We have to make TeX try to read something from the file, so % it can discover if there is anything in it. \read 1 to \temp \ifeof 1 (Index is empty) \else \input \jobname.#1s \fi \fi \closein 1 \enddoublecolumns \Etex } % These macros are used by the sorted index file itself. % Change them to control the appearance of the index. % Same as \bigskipamount except no shrink. % \balancecolumns gets confused if there is any shrink. \newskip\initialskipamount \initialskipamount 12pt plus4pt \def\initial #1{% {\let\tentt=\sectt \let\tt=\sectt \let\sf=\sectt \ifdim\lastskip<\initialskipamount \removelastskip \penalty-200 \vskip \initialskipamount\fi \line{\secbf#1\hfill}\kern 2pt\penalty10000}} % This typesets a paragraph consisting of #1, dot leaders, and then #2 % flush to the right margin. It is used for index and table of contents % entries. The paragraph is indented by \leftskip. % \def\entry #1#2{\begingroup % % Start a new paragraph if necessary, so our assignments below can't % affect previous text. \par % % Do not fill out the last line with white space. \parfillskip = 0in % % No extra space above this paragraph. \parskip = 0in % % Do not prefer a separate line ending with a hyphen to fewer lines. \finalhyphendemerits = 0 % % \hangindent is only relevant when the entry text and page number % don't both fit on one line. In that case, bob suggests starting the % dots pretty far over on the line. Unfortunately, a large % indentation looks wrong when the entry text itself is broken across % lines. So we use a small indentation and put up with long leaders. % % \hangafter is reset to 1 (which is the value we want) at the start % of each paragraph, so we need not do anything with that. \hangindent=2em % % When the entry text needs to be broken, just fill out the first line % with blank space. \rightskip = 0pt plus1fil % % Start a ``paragraph'' for the index entry so the line breaking % parameters we've set above will have an effect. \noindent % % Insert the text of the index entry. TeX will do line-breaking on it. #1% % The following is kluged to not output a line of dots in the index if % there are no page numbers. The next person who breaks this will be % cursed by a Unix daemon. \def\tempa{{\rm }}% \def\tempb{#2}% \edef\tempc{\tempa}% \edef\tempd{\tempb}% \ifx\tempc\tempd\ \else% % % If we must, put the page number on a line of its own, and fill out % this line with blank space. (The \hfil is overwhelmed with the % fill leaders glue in \indexdotfill if the page number does fit.) \hfil\penalty50 \null\nobreak\indexdotfill % Have leaders before the page number. % % The `\ ' here is removed by the implicit \unskip that TeX does as % part of (the primitive) \par. Without it, a spurious underfull % \hbox ensues. \ #2% The page number ends the paragraph. \fi% \par \endgroup} % Like \dotfill except takes at least 1 em. \def\indexdotfill{\cleaders \hbox{$\mathsurround=0pt \mkern1.5mu . \mkern1.5mu$}\hskip 1em plus 1fill} \def\primary #1{\line{#1\hfil}} \newskip\secondaryindent \secondaryindent=0.5cm \def\secondary #1#2{ {\parfillskip=0in \parskip=0in \hangindent =1in \hangafter=1 \noindent\hskip\secondaryindent\hbox{#1}\indexdotfill #2\par }} %% Define two-column mode, which is used in indexes. %% Adapted from the TeXbook, page 416. \catcode `\@=11 \newbox\partialpage \newdimen\doublecolumnhsize \def\begindoublecolumns{\begingroup % Grab any single-column material above us. \output = {\global\setbox\partialpage =\vbox{\unvbox255\kern -\topskip \kern \baselineskip}}% \eject % % Now switch to the double-column output routine. \output={\doublecolumnout}% % % Change the page size parameters. We could do this once outside this % routine, in each of @smallbook, @afourpaper, and the default 8.5x11 % format, but then we repeat the same computation. Repeating a couple % of assignments once per index is clearly meaningless for the % execution time, so we may as well do it once. % % First we halve the line length, less a little for the gutter between % the columns. We compute the gutter based on the line length, so it % changes automatically with the paper format. The magic constant % below is chosen so that the gutter has the same value (well, +- < % 1pt) as it did when we hard-coded it. % % We put the result in a separate register, \doublecolumhsize, so we % can restore it in \pagesofar, after \hsize itself has (potentially) % been clobbered. % \doublecolumnhsize = \hsize \advance\doublecolumnhsize by -.04154\hsize \divide\doublecolumnhsize by 2 \hsize = \doublecolumnhsize % % Double the \vsize as well. (We don't need a separate register here, % since nobody clobbers \vsize.) \vsize = 2\vsize \doublecolumnpagegoal } \def\enddoublecolumns{\eject \endgroup \pagegoal=\vsize \unvbox\partialpage} \def\doublecolumnsplit{\splittopskip=\topskip \splitmaxdepth=\maxdepth \global\dimen@=\pageheight \global\advance\dimen@ by-\ht\partialpage \global\setbox1=\vsplit255 to\dimen@ \global\setbox0=\vbox{\unvbox1} \global\setbox3=\vsplit255 to\dimen@ \global\setbox2=\vbox{\unvbox3} \ifdim\ht0>\dimen@ \setbox255=\vbox{\unvbox0\unvbox2} \global\setbox255=\copy5 \fi \ifdim\ht2>\dimen@ \setbox255=\vbox{\unvbox0\unvbox2} \global\setbox255=\copy5 \fi } \def\doublecolumnpagegoal{% \dimen@=\vsize \advance\dimen@ by-2\ht\partialpage \global\pagegoal=\dimen@ } \def\pagesofar{\unvbox\partialpage % \hsize=\doublecolumnhsize % have to restore this since output routine \wd0=\hsize \wd2=\hsize \hbox to\pagewidth{\box0\hfil\box2}} \def\doublecolumnout{% \setbox5=\copy255 {\vbadness=10000 \doublecolumnsplit} \ifvbox255 \setbox0=\vtop to\dimen@{\unvbox0} \setbox2=\vtop to\dimen@{\unvbox2} \onepageout\pagesofar \unvbox255 \penalty\outputpenalty \else \setbox0=\vbox{\unvbox5} \ifvbox0 \dimen@=\ht0 \advance\dimen@ by\topskip \advance\dimen@ by-\baselineskip \divide\dimen@ by2 \splittopskip=\topskip \splitmaxdepth=\maxdepth {\vbadness=10000 \loop \global\setbox5=\copy0 \setbox1=\vsplit5 to\dimen@ \setbox3=\vsplit5 to\dimen@ \ifvbox5 \global\advance\dimen@ by1pt \repeat \setbox0=\vbox to\dimen@{\unvbox1} \setbox2=\vbox to\dimen@{\unvbox3} \global\setbox\partialpage=\vbox{\pagesofar} \doublecolumnpagegoal } \fi \fi } \catcode `\@=\other \message{sectioning,} % Define chapters, sections, etc. \newcount \chapno \newcount \secno \secno=0 \newcount \subsecno \subsecno=0 \newcount \subsubsecno \subsubsecno=0 % This counter is funny since it counts through charcodes of letters A, B, ... \newcount \appendixno \appendixno = `\@ \def\appendixletter{\char\the\appendixno} \newwrite \contentsfile % This is called from \setfilename. \def\opencontents{\openout \contentsfile = \jobname.toc} % Each @chapter defines this as the name of the chapter. % page headings and footings can use it. @section does likewise \def\thischapter{} \def\thissection{} \def\seccheck#1{\if \pageno<0 % \errmessage{@#1 not allowed after generating table of contents}\fi % } \def\chapternofonts{% \let\rawbackslash=\relax% \let\frenchspacing=\relax% \def\result{\realbackslash result} \def\equiv{\realbackslash equiv} \def\expansion{\realbackslash expansion} \def\print{\realbackslash print} \def\TeX{\realbackslash TeX} \def\dots{\realbackslash dots} \def\copyright{\realbackslash copyright} \def\tt{\realbackslash tt} \def\bf{\realbackslash bf } \def\w{\realbackslash w} \def\less{\realbackslash less} \def\gtr{\realbackslash gtr} \def\hat{\realbackslash hat} \def\char{\realbackslash char} \def\tclose##1{\realbackslash tclose {##1}} \def\code##1{\realbackslash code {##1}} \def\samp##1{\realbackslash samp {##1}} \def\r##1{\realbackslash r {##1}} \def\b##1{\realbackslash b {##1}} \def\key##1{\realbackslash key {##1}} \def\file##1{\realbackslash file {##1}} \def\kbd##1{\realbackslash kbd {##1}} % These are redefined because @smartitalic wouldn't work inside xdef. \def\i##1{\realbackslash i {##1}} \def\cite##1{\realbackslash cite {##1}} \def\var##1{\realbackslash var {##1}} \def\emph##1{\realbackslash emph {##1}} \def\dfn##1{\realbackslash dfn {##1}} } \newcount\absseclevel % used to calculate proper heading level \newcount\secbase\secbase=0 % @raise/lowersections modify this count % @raisesections: treat @section as chapter, @subsection as section, etc. \def\raisesections{\global\advance\secbase by -1} \let\up=\raisesections % original BFox name % @lowersections: treat @chapter as section, @section as subsection, etc. \def\lowersections{\global\advance\secbase by 1} \let\down=\lowersections % original BFox name % Choose a numbered-heading macro % #1 is heading level if unmodified by @raisesections or @lowersections % #2 is text for heading \def\numhead#1#2{\absseclevel=\secbase\advance\absseclevel by #1 \ifcase\absseclevel \chapterzzz{#2} \or \seczzz{#2} \or \numberedsubseczzz{#2} \or \numberedsubsubseczzz{#2} \else \ifnum \absseclevel<0 \chapterzzz{#2} \else \numberedsubsubseczzz{#2} \fi \fi } % like \numhead, but chooses appendix heading levels \def\apphead#1#2{\absseclevel=\secbase\advance\absseclevel by #1 \ifcase\absseclevel \appendixzzz{#2} \or \appendixsectionzzz{#2} \or \appendixsubseczzz{#2} \or \appendixsubsubseczzz{#2} \else \ifnum \absseclevel<0 \appendixzzz{#2} \else \appendixsubsubseczzz{#2} \fi \fi } % like \numhead, but chooses numberless heading levels \def\unnmhead#1#2{\absseclevel=\secbase\advance\absseclevel by #1 \ifcase\absseclevel \unnumberedzzz{#2} \or \unnumberedseczzz{#2} \or \unnumberedsubseczzz{#2} \or \unnumberedsubsubseczzz{#2} \else \ifnum \absseclevel<0 \unnumberedzzz{#2} \else \unnumberedsubsubseczzz{#2} \fi \fi } \def\thischaptername{No Chapter Title} \outer\def\chapter{\parsearg\chapteryyy} \def\chapteryyy #1{\numhead0{#1}} % normally numhead0 calls chapterzzz \def\chapterzzz #1{\seccheck{chapter}% \secno=0 \subsecno=0 \subsubsecno=0 \global\advance \chapno by 1 \message{Chapter \the\chapno}% \chapmacro {#1}{\the\chapno}% \gdef\thissection{#1}% \gdef\thischaptername{#1}% % We don't substitute the actual chapter name into \thischapter % because we don't want its macros evaluated now. \xdef\thischapter{Chapter \the\chapno: \noexpand\thischaptername}% {\chapternofonts% \edef\temp{{\realbackslash chapentry {#1}{\the\chapno}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \donoderef % \global\let\section = \numberedsec \global\let\subsection = \numberedsubsec \global\let\subsubsection = \numberedsubsubsec }} \outer\def\appendix{\parsearg\appendixyyy} \def\appendixyyy #1{\apphead0{#1}} % normally apphead0 calls appendixzzz \def\appendixzzz #1{\seccheck{appendix}% \secno=0 \subsecno=0 \subsubsecno=0 \global\advance \appendixno by 1 \message{Appendix \appendixletter}% \chapmacro {#1}{Appendix \appendixletter}% \gdef\thissection{#1}% \gdef\thischaptername{#1}% \xdef\thischapter{Appendix \appendixletter: \noexpand\thischaptername}% {\chapternofonts% \edef\temp{{\realbackslash chapentry {#1}{Appendix \appendixletter}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \appendixnoderef % \global\let\section = \appendixsec \global\let\subsection = \appendixsubsec \global\let\subsubsection = \appendixsubsubsec }} \outer\def\top{\parsearg\unnumberedyyy} \outer\def\unnumbered{\parsearg\unnumberedyyy} \def\unnumberedyyy #1{\unnmhead0{#1}} % normally unnmhead0 calls unnumberedzzz \def\unnumberedzzz #1{\seccheck{unnumbered}% \secno=0 \subsecno=0 \subsubsecno=0 % % This used to be simply \message{#1}, but TeX fully expands the % argument to \message. Therefore, if #1 contained @-commands, TeX % expanded them. For example, in `@unnumbered The @cite{Book}', TeX % expanded @cite (which turns out to cause errors because \cite is meant % to be executed, not expanded). % % Anyway, we don't want the fully-expanded definition of @cite to appear % as a result of the \message, we just want `@cite' itself. We use % \the to achieve this: TeX expands \the only once, % simply yielding the contents of the . \toks0 = {#1}\message{(\the\toks0)}% % \unnumbchapmacro {#1}% \gdef\thischapter{#1}\gdef\thissection{#1}% {\chapternofonts% \edef\temp{{\realbackslash unnumbchapentry {#1}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \unnumbnoderef % \global\let\section = \unnumberedsec \global\let\subsection = \unnumberedsubsec \global\let\subsubsection = \unnumberedsubsubsec }} \outer\def\numberedsec{\parsearg\secyyy} \def\secyyy #1{\numhead1{#1}} % normally calls seczzz \def\seczzz #1{\seccheck{section}% \subsecno=0 \subsubsecno=0 \global\advance \secno by 1 % \gdef\thissection{#1}\secheading {#1}{\the\chapno}{\the\secno}% {\chapternofonts% \edef\temp{{\realbackslash secentry % {#1}{\the\chapno}{\the\secno}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \donoderef % \penalty 10000 % }} \outer\def\appenixsection{\parsearg\appendixsecyyy} \outer\def\appendixsec{\parsearg\appendixsecyyy} \def\appendixsecyyy #1{\apphead1{#1}} % normally calls appendixsectionzzz \def\appendixsectionzzz #1{\seccheck{appendixsection}% \subsecno=0 \subsubsecno=0 \global\advance \secno by 1 % \gdef\thissection{#1}\secheading {#1}{\appendixletter}{\the\secno}% {\chapternofonts% \edef\temp{{\realbackslash secentry % {#1}{\appendixletter}{\the\secno}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \appendixnoderef % \penalty 10000 % }} \outer\def\unnumberedsec{\parsearg\unnumberedsecyyy} \def\unnumberedsecyyy #1{\unnmhead1{#1}} % normally calls unnumberedseczzz \def\unnumberedseczzz #1{\seccheck{unnumberedsec}% \plainsecheading {#1}\gdef\thissection{#1}% {\chapternofonts% \edef\temp{{\realbackslash unnumbsecentry{#1}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \unnumbnoderef % \penalty 10000 % }} \outer\def\numberedsubsec{\parsearg\numberedsubsecyyy} \def\numberedsubsecyyy #1{\numhead2{#1}} % normally calls numberedsubseczzz \def\numberedsubseczzz #1{\seccheck{subsection}% \gdef\thissection{#1}\subsubsecno=0 \global\advance \subsecno by 1 % \subsecheading {#1}{\the\chapno}{\the\secno}{\the\subsecno}% {\chapternofonts% \edef\temp{{\realbackslash subsecentry % {#1}{\the\chapno}{\the\secno}{\the\subsecno}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \donoderef % \penalty 10000 % }} \outer\def\appendixsubsec{\parsearg\appendixsubsecyyy} \def\appendixsubsecyyy #1{\apphead2{#1}} % normally calls appendixsubseczzz \def\appendixsubseczzz #1{\seccheck{appendixsubsec}% \gdef\thissection{#1}\subsubsecno=0 \global\advance \subsecno by 1 % \subsecheading {#1}{\appendixletter}{\the\secno}{\the\subsecno}% {\chapternofonts% \edef\temp{{\realbackslash subsecentry % {#1}{\appendixletter}{\the\secno}{\the\subsecno}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \appendixnoderef % \penalty 10000 % }} \outer\def\unnumberedsubsec{\parsearg\unnumberedsubsecyyy} \def\unnumberedsubsecyyy #1{\unnmhead2{#1}} %normally calls unnumberedsubseczzz \def\unnumberedsubseczzz #1{\seccheck{unnumberedsubsec}% \plainsecheading {#1}\gdef\thissection{#1}% {\chapternofonts% \edef\temp{{\realbackslash unnumbsubsecentry{#1}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \unnumbnoderef % \penalty 10000 % }} \outer\def\numberedsubsubsec{\parsearg\numberedsubsubsecyyy} \def\numberedsubsubsecyyy #1{\numhead3{#1}} % normally numberedsubsubseczzz \def\numberedsubsubseczzz #1{\seccheck{subsubsection}% \gdef\thissection{#1}\global\advance \subsubsecno by 1 % \subsubsecheading {#1} {\the\chapno}{\the\secno}{\the\subsecno}{\the\subsubsecno}% {\chapternofonts% \edef\temp{{\realbackslash subsubsecentry % {#1} {\the\chapno}{\the\secno}{\the\subsecno}{\the\subsubsecno} {\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \donoderef % \penalty 10000 % }} \outer\def\appendixsubsubsec{\parsearg\appendixsubsubsecyyy} \def\appendixsubsubsecyyy #1{\apphead3{#1}} % normally appendixsubsubseczzz \def\appendixsubsubseczzz #1{\seccheck{appendixsubsubsec}% \gdef\thissection{#1}\global\advance \subsubsecno by 1 % \subsubsecheading {#1} {\appendixletter}{\the\secno}{\the\subsecno}{\the\subsubsecno}% {\chapternofonts% \edef\temp{{\realbackslash subsubsecentry{#1}% {\appendixletter} {\the\secno}{\the\subsecno}{\the\subsubsecno}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \appendixnoderef % \penalty 10000 % }} \outer\def\unnumberedsubsubsec{\parsearg\unnumberedsubsubsecyyy} \def\unnumberedsubsubsecyyy #1{\unnmhead3{#1}} %normally unnumberedsubsubseczzz \def\unnumberedsubsubseczzz #1{\seccheck{unnumberedsubsubsec}% \plainsecheading {#1}\gdef\thissection{#1}% {\chapternofonts% \edef\temp{{\realbackslash unnumbsubsubsecentry{#1}{\noexpand\folio}}}% \escapechar=`\\% \write \contentsfile \temp % \unnumbnoderef % \penalty 10000 % }} % These are variants which are not "outer", so they can appear in @ifinfo. % Actually, they should now be obsolete; ordinary section commands should work. \def\infotop{\parsearg\unnumberedzzz} \def\infounnumbered{\parsearg\unnumberedzzz} \def\infounnumberedsec{\parsearg\unnumberedseczzz} \def\infounnumberedsubsec{\parsearg\unnumberedsubseczzz} \def\infounnumberedsubsubsec{\parsearg\unnumberedsubsubseczzz} \def\infoappendix{\parsearg\appendixzzz} \def\infoappendixsec{\parsearg\appendixseczzz} \def\infoappendixsubsec{\parsearg\appendixsubseczzz} \def\infoappendixsubsubsec{\parsearg\appendixsubsubseczzz} \def\infochapter{\parsearg\chapterzzz} \def\infosection{\parsearg\sectionzzz} \def\infosubsection{\parsearg\subsectionzzz} \def\infosubsubsection{\parsearg\subsubsectionzzz} % These macros control what the section commands do, according % to what kind of chapter we are in (ordinary, appendix, or unnumbered). % Define them by default for a numbered chapter. \global\let\section = \numberedsec \global\let\subsection = \numberedsubsec \global\let\subsubsection = \numberedsubsubsec % Define @majorheading, @heading and @subheading % NOTE on use of \vbox for chapter headings, section headings, and % such: % 1) We use \vbox rather than the earlier \line to permit % overlong headings to fold. % 2) \hyphenpenalty is set to 10000 because hyphenation in a % heading is obnoxious; this forbids it. % 3) Likewise, headings look best if no \parindent is used, and % if justification is not attempted. Hence \raggedright. \def\majorheading{\parsearg\majorheadingzzz} \def\majorheadingzzz #1{% {\advance\chapheadingskip by 10pt \chapbreak }% {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}\bigskip \par\penalty 200} \def\chapheading{\parsearg\chapheadingzzz} \def\chapheadingzzz #1{\chapbreak % {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}\bigskip \par\penalty 200} \def\heading{\parsearg\secheadingi} \def\subheading{\parsearg\subsecheadingi} \def\subsubheading{\parsearg\subsubsecheadingi} % These macros generate a chapter, section, etc. heading only % (including whitespace, linebreaking, etc. around it), % given all the information in convenient, parsed form. %%% Args are the skip and penalty (usually negative) \def\dobreak#1#2{\par\ifdim\lastskip<#1\removelastskip\penalty#2\vskip#1\fi} \def\setchapterstyle #1 {\csname CHAPF#1\endcsname} %%% Define plain chapter starts, and page on/off switching for it % Parameter controlling skip before chapter headings (if needed) \newskip \chapheadingskip \chapheadingskip = 30pt plus 8pt minus 4pt \def\chapbreak{\dobreak \chapheadingskip {-4000}} \def\chappager{\par\vfill\supereject} \def\chapoddpage{\chappager \ifodd\pageno \else \hbox to 0pt{} \chappager\fi} \def\setchapternewpage #1 {\csname CHAPPAG#1\endcsname} \def\CHAPPAGoff{ \global\let\pchapsepmacro=\chapbreak \global\let\pagealignmacro=\chappager} \def\CHAPPAGon{ \global\let\pchapsepmacro=\chappager \global\let\pagealignmacro=\chappager \global\def\HEADINGSon{\HEADINGSsingle}} \def\CHAPPAGodd{ \global\let\pchapsepmacro=\chapoddpage \global\let\pagealignmacro=\chapoddpage \global\def\HEADINGSon{\HEADINGSdouble}} \CHAPPAGon \def\CHAPFplain{ \global\let\chapmacro=\chfplain \global\let\unnumbchapmacro=\unnchfplain} \def\chfplain #1#2{% \pchapsepmacro {% \chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #2\enspace #1}% }% \bigskip \penalty5000 } \def\unnchfplain #1{% \pchapsepmacro % {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}\bigskip \par\penalty 10000 % } \CHAPFplain % The default \def\unnchfopen #1{% \chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}\bigskip \par\penalty 10000 % } \def\chfopen #1#2{\chapoddpage {\chapfonts \vbox to 3in{\vfil \hbox to\hsize{\hfil #2} \hbox to\hsize{\hfil #1} \vfil}}% \par\penalty 5000 % } \def\CHAPFopen{ \global\let\chapmacro=\chfopen \global\let\unnumbchapmacro=\unnchfopen} % Parameter controlling skip before section headings. \newskip \subsecheadingskip \subsecheadingskip = 17pt plus 8pt minus 4pt \def\subsecheadingbreak{\dobreak \subsecheadingskip {-500}} \newskip \secheadingskip \secheadingskip = 21pt plus 8pt minus 4pt \def\secheadingbreak{\dobreak \secheadingskip {-1000}} % @paragraphindent is defined for the Info formatting commands only. \let\paragraphindent=\comment % Section fonts are the base font at magstep2, which produces % a size a bit more than 14 points in the default situation. \def\secheading #1#2#3{\secheadingi {#2.#3\enspace #1}} \def\plainsecheading #1{\secheadingi {#1}} \def\secheadingi #1{{\advance \secheadingskip by \parskip % \secheadingbreak}% {\secfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}% \ifdim \parskip<10pt \kern 10pt\kern -\parskip\fi \penalty 10000 } % Subsection fonts are the base font at magstep1, % which produces a size of 12 points. \def\subsecheading #1#2#3#4{\subsecheadingi {#2.#3.#4\enspace #1}} \def\subsecheadingi #1{{\advance \subsecheadingskip by \parskip % \subsecheadingbreak}% {\subsecfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}% \ifdim \parskip<10pt \kern 10pt\kern -\parskip\fi \penalty 10000 } \def\subsubsecfonts{\subsecfonts} % Maybe this should change: % Perhaps make sssec fonts scaled % magstep half \def\subsubsecheading #1#2#3#4#5{\subsubsecheadingi {#2.#3.#4.#5\enspace #1}} \def\subsubsecheadingi #1{{\advance \subsecheadingskip by \parskip % \subsecheadingbreak}% {\subsubsecfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\raggedright \rm #1\hfill}}% \ifdim \parskip<10pt \kern 10pt\kern -\parskip\fi \penalty 10000} \message{toc printing,} % Finish up the main text and prepare to read what we've written % to \contentsfile. \newskip\contentsrightmargin \contentsrightmargin=1in \def\startcontents#1{% \pagealignmacro \immediate\closeout \contentsfile \ifnum \pageno>0 \pageno = -1 % Request roman numbered pages. \fi % Don't need to put `Contents' or `Short Contents' in the headline. % It is abundantly clear what they are. \unnumbchapmacro{#1}\def\thischapter{}% \begingroup % Set up to handle contents files properly. \catcode`\\=0 \catcode`\{=1 \catcode`\}=2 \catcode`\@=11 \raggedbottom % Worry more about breakpoints than the bottom. \advance\hsize by -\contentsrightmargin % Don't use the full line length. } % Normal (long) toc. \outer\def\contents{% \startcontents{Table of Contents}% \input \jobname.toc \endgroup \vfill \eject } % And just the chapters. \outer\def\summarycontents{% \startcontents{Short Contents}% % \let\chapentry = \shortchapentry \let\unnumbchapentry = \shortunnumberedentry % We want a true roman here for the page numbers. \secfonts \let\rm=\shortcontrm \let\bf=\shortcontbf \let\sl=\shortcontsl \rm \advance\baselineskip by 1pt % Open it up a little. \def\secentry ##1##2##3##4{} \def\unnumbsecentry ##1##2{} \def\subsecentry ##1##2##3##4##5{} \def\unnumbsubsecentry ##1##2{} \def\subsubsecentry ##1##2##3##4##5##6{} \def\unnumbsubsubsecentry ##1##2{} \input \jobname.toc \endgroup \vfill \eject } \let\shortcontents = \summarycontents % These macros generate individual entries in the table of contents. % The first argument is the chapter or section name. % The last argument is the page number. % The arguments in between are the chapter number, section number, ... % Chapter-level things, for both the long and short contents. \def\chapentry#1#2#3{\dochapentry{#2\labelspace#1}{#3}} % See comments in \dochapentry re vbox and related settings \def\shortchapentry#1#2#3{% \tocentry{\shortchaplabel{#2}\labelspace #1}{\doshortpageno{#3}}% } % Typeset the label for a chapter or appendix for the short contents. % The arg is, e.g. `Appendix A' for an appendix, or `3' for a chapter. % We could simplify the code here by writing out an \appendixentry % command in the toc file for appendices, instead of using \chapentry % for both, but it doesn't seem worth it. \setbox0 = \hbox{\shortcontrm Appendix } \newdimen\shortappendixwidth \shortappendixwidth = \wd0 \def\shortchaplabel#1{% % We typeset #1 in a box of constant width, regardless of the text of % #1, so the chapter titles will come out aligned. \setbox0 = \hbox{#1}% \dimen0 = \ifdim\wd0 > \shortappendixwidth \shortappendixwidth \else 0pt \fi % % This space should be plenty, since a single number is .5em, and the % widest letter (M) is 1em, at least in the Computer Modern fonts. % (This space doesn't include the extra space that gets added after % the label; that gets put in in \shortchapentry above.) \advance\dimen0 by 1.1em \hbox to \dimen0{#1\hfil}% } \def\unnumbchapentry#1#2{\dochapentry{#1}{#2}} \def\shortunnumberedentry#1#2{\tocentry{#1}{\doshortpageno{#2}}} % Sections. \def\secentry#1#2#3#4{\dosecentry{#2.#3\labelspace#1}{#4}} \def\unnumbsecentry#1#2{\dosecentry{#1}{#2}} % Subsections. \def\subsecentry#1#2#3#4#5{\dosubsecentry{#2.#3.#4\labelspace#1}{#5}} \def\unnumbsubsecentry#1#2{\dosubsecentry{#1}{#2}} % And subsubsections. \def\subsubsecentry#1#2#3#4#5#6{% \dosubsubsecentry{#2.#3.#4.#5\labelspace#1}{#6}} \def\unnumbsubsubsecentry#1#2{\dosubsubsecentry{#1}{#2}} % This parameter controls the indentation of the various levels. \newdimen\tocindent \tocindent = 3pc % Now for the actual typesetting. In all these, #1 is the text and #2 is the % page number. % % If the toc has to be broken over pages, we would want to be at chapters % if at all possible; hence the \penalty. \def\dochapentry#1#2{% \penalty-300 \vskip\baselineskip \begingroup \chapentryfonts \tocentry{#1}{\dopageno{#2}}% \endgroup \nobreak\vskip .25\baselineskip } \def\dosecentry#1#2{\begingroup \secentryfonts \leftskip=\tocindent \tocentry{#1}{\dopageno{#2}}% \endgroup} \def\dosubsecentry#1#2{\begingroup \subsecentryfonts \leftskip=2\tocindent \tocentry{#1}{\dopageno{#2}}% \endgroup} \def\dosubsubsecentry#1#2{\begingroup \subsubsecentryfonts \leftskip=3\tocindent \tocentry{#1}{\dopageno{#2}}% \endgroup} % Final typesetting of a toc entry; we use the same \entry macro as for % the index entries, but we want to suppress hyphenation here. (We % can't do that in the \entry macro, since index entries might consist % of hyphenated-identifiers-that-do-not-fit-on-a-line-and-nothing-else.) % \def\tocentry#1#2{\begingroup \hyphenpenalty = 10000 \entry{#1}{#2}% \endgroup} % Space between chapter (or whatever) number and the title. \def\labelspace{\hskip1em \relax} \def\dopageno#1{{\rm #1}} \def\doshortpageno#1{{\rm #1}} \def\chapentryfonts{\secfonts \rm} \def\secentryfonts{\textfonts} \let\subsecentryfonts = \textfonts \let\subsubsecentryfonts = \textfonts \message{environments,} % Since these characters are used in examples, it should be an even number of % \tt widths. Each \tt character is 1en, so two makes it 1em. % Furthermore, these definitions must come after we define our fonts. \newbox\dblarrowbox \newbox\longdblarrowbox \newbox\pushcharbox \newbox\bullbox \newbox\equivbox \newbox\errorbox \let\ptexequiv = \equiv %{\tentt %\global\setbox\dblarrowbox = \hbox to 1em{\hfil$\Rightarrow$\hfil} %\global\setbox\longdblarrowbox = \hbox to 1em{\hfil$\mapsto$\hfil} %\global\setbox\pushcharbox = \hbox to 1em{\hfil$\dashv$\hfil} %\global\setbox\equivbox = \hbox to 1em{\hfil$\ptexequiv$\hfil} % Adapted from the manmac format (p.420 of TeXbook) %\global\setbox\bullbox = \hbox to 1em{\kern.15em\vrule height .75ex width .85ex % depth .1ex\hfil} %} \def\point{$\star$} \def\result{\leavevmode\raise.15ex\hbox to 1em{\hfil$\Rightarrow$\hfil}} \def\expansion{\leavevmode\raise.1ex\hbox to 1em{\hfil$\mapsto$\hfil}} \def\print{\leavevmode\lower.1ex\hbox to 1em{\hfil$\dashv$\hfil}} \def\equiv{\leavevmode\lower.1ex\hbox to 1em{\hfil$\ptexequiv$\hfil}} % Adapted from the TeXbook's \boxit. {\tentt \global\dimen0 = 3em}% Width of the box. \dimen2 = .55pt % Thickness of rules % The text. (`r' is open on the right, `e' somewhat less so on the left.) \setbox0 = \hbox{\kern-.75pt \tensf error\kern-1.5pt} \global\setbox\errorbox=\hbox to \dimen0{\hfil \hsize = \dimen0 \advance\hsize by -5.8pt % Space to left+right. \advance\hsize by -2\dimen2 % Rules. \vbox{ \hrule height\dimen2 \hbox{\vrule width\dimen2 \kern3pt % Space to left of text. \vtop{\kern2.4pt \box0 \kern2.4pt}% Space above/below. \kern3pt\vrule width\dimen2}% Space to right. \hrule height\dimen2} \hfil} % The @error{} command. \def\error{\leavevmode\lower.7ex\copy\errorbox} % @tex ... @end tex escapes into raw Tex temporarily. % One exception: @ is still an escape character, so that @end tex works. % But \@ or @@ will get a plain tex @ character. \def\tex{\begingroup \catcode `\\=0 \catcode `\{=1 \catcode `\}=2 \catcode `\$=3 \catcode `\&=4 \catcode `\#=6 \catcode `\^=7 \catcode `\_=8 \catcode `\~=13 \let~=\tie \catcode `\%=14 \catcode 43=12 \catcode`\"=12 \catcode`\==12 \catcode`\|=12 \catcode`\<=12 \catcode`\>=12 \escapechar=`\\ % \let\{=\ptexlbrace \let\}=\ptexrbrace \let\.=\ptexdot \let\*=\ptexstar \let\dots=\ptexdots \def\@{@}% \let\bullet=\ptexbullet \let\b=\ptexb \let\c=\ptexc \let\i=\ptexi \let\t=\ptext \let\l=\ptexl \let\L=\ptexL % \let\Etex=\endgroup} % Define @lisp ... @endlisp. % @lisp does a \begingroup so it can rebind things, % including the definition of @endlisp (which normally is erroneous). % Amount to narrow the margins by for @lisp. \newskip\lispnarrowing \lispnarrowing=0.4in % This is the definition that ^^M gets inside @lisp, @example, and other % such environments. \null is better than a space, since it doesn't % have any width. \def\lisppar{\null\endgraf} % Make each space character in the input produce a normal interword % space in the output. Don't allow a line break at this space, as this % is used only in environments like @example, where each line of input % should produce a line of output anyway. % {\obeyspaces % \gdef\sepspaces{\obeyspaces\let =\tie}} % Define \obeyedspace to be our active space, whatever it is. This is % for use in \parsearg. {\sepspaces% \global\let\obeyedspace= } % This space is always present above and below environments. \newskip\envskipamount \envskipamount = 0pt % Make spacing and below environment symmetrical. We use \parskip here % to help in doing that, since in @example-like environments \parskip % is reset to zero; thus the \afterenvbreak inserts no space -- but the % start of the next paragraph will insert \parskip % \def\aboveenvbreak{{\advance\envskipamount by \parskip \endgraf \ifdim\lastskip<\envskipamount \removelastskip \penalty-50 \vskip\envskipamount \fi}} \let\afterenvbreak = \aboveenvbreak % \nonarrowing is a flag. If "set", @lisp etc don't narrow margins. \let\nonarrowing=\relax %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % \cartouche: draw rectangle w/rounded corners around argument \font\circle=lcircle10 \newdimen\circthick \newdimen\cartouter\newdimen\cartinner \newskip\normbskip\newskip\normpskip\newskip\normlskip \circthick=\fontdimen8\circle % \def\ctl{{\circle\char'013\hskip -6pt}}% 6pt from pl file: 1/2charwidth \def\ctr{{\hskip 6pt\circle\char'010}} \def\cbl{{\circle\char'012\hskip -6pt}} \def\cbr{{\hskip 6pt\circle\char'011}} \def\carttop{\hbox to \cartouter{\hskip\lskip \ctl\leaders\hrule height\circthick\hfil\ctr \hskip\rskip}} \def\cartbot{\hbox to \cartouter{\hskip\lskip \cbl\leaders\hrule height\circthick\hfil\cbr \hskip\rskip}} % \newskip\lskip\newskip\rskip \long\def\cartouche{% \begingroup \lskip=\leftskip \rskip=\rightskip \leftskip=0pt\rightskip=0pt %we want these *outside*. \cartinner=\hsize \advance\cartinner by-\lskip \advance\cartinner by-\rskip \cartouter=\hsize \advance\cartouter by 18pt % allow for 3pt kerns on either % side, and for 6pt waste from % each corner char \normbskip=\baselineskip \normpskip=\parskip \normlskip=\lineskip % Flag to tell @lisp, etc., not to narrow margin. \let\nonarrowing=\comment \vbox\bgroup \baselineskip=0pt\parskip=0pt\lineskip=0pt \carttop \hbox\bgroup \hskip\lskip \vrule\kern3pt \vbox\bgroup \hsize=\cartinner \kern3pt \begingroup \baselineskip=\normbskip \lineskip=\normlskip \parskip=\normpskip \vskip -\parskip \def\Ecartouche{% \endgroup \kern3pt \egroup \kern3pt\vrule \hskip\rskip \egroup \cartbot \egroup \endgroup }} % This macro is called at the beginning of all the @example variants, % inside a group. \def\nonfillstart{% \aboveenvbreak \inENV % This group ends at the end of the body \hfuzz = 12pt % Don't be fussy \sepspaces % Make spaces be word-separators rather than space tokens. \singlespace \let\par = \lisppar % don't ignore blank lines \obeylines % each line of input is a line of output \parskip = 0pt \parindent = 0pt \emergencystretch = 0pt % don't try to avoid overfull boxes % @cartouche defines \nonarrowing to inhibit narrowing % at next level down. \ifx\nonarrowing\relax \advance \leftskip by \lispnarrowing \exdentamount=\lispnarrowing \let\exdent=\nofillexdent \let\nonarrowing=\relax \fi } % To ending an @example-like environment, we first end the paragraph % (via \afterenvbreak's vertical glue), and then the group. That way we % keep the zero \parskip that the environments set -- \parskip glue % will be inserted at the beginning of the next paragraph in the % document, after the environment. % \def\nonfillfinish{\afterenvbreak\endgroup}% % This macro is \def\lisp{\begingroup \nonfillstart \let\Elisp = \nonfillfinish \tt \rawbackslash % have \ input char produce \ char from current font \gobble } % Define the \E... control sequence only if we are inside the % environment, so the error checking in \end will work. % % We must call \lisp last in the definition, since it reads the % return following the @example (or whatever) command. % \def\example{\begingroup \def\Eexample{\nonfillfinish\endgroup}\lisp} \def\smallexample{\begingroup \def\Esmallexample{\nonfillfinish\endgroup}\lisp} \def\smalllisp{\begingroup \def\Esmalllisp{\nonfillfinish\endgroup}\lisp} % @smallexample and @smalllisp. This is not used unless the @smallbook % command is given. Originally contributed by Pavel@xerox. % \def\smalllispx{\begingroup \nonfillstart \let\Esmalllisp = \nonfillfinish \let\Esmallexample = \nonfillfinish % % Smaller interline space and fonts for small examples. \baselineskip 10pt \indexfonts \tt \rawbackslash % output the \ character from the current font \gobble } % This is @display; same as @lisp except use roman font. % \def\display{\begingroup \nonfillstart \let\Edisplay = \nonfillfinish \gobble } % This is @format; same as @display except don't narrow margins. % \def\format{\begingroup \let\nonarrowing = t \nonfillstart \let\Eformat = \nonfillfinish \gobble } % @flushleft (same as @format) and @flushright. % \def\flushleft{\begingroup \let\nonarrowing = t \nonfillstart \let\Eflushleft = \nonfillfinish \gobble } \def\flushright{\begingroup \let\nonarrowing = t \nonfillstart \let\Eflushright = \nonfillfinish \advance\leftskip by 0pt plus 1fill \gobble} % @quotation does normal linebreaking and narrows the margins. % \def\quotation{% \begingroup\inENV %This group ends at the end of the @quotation body {\parskip=0pt % because we will skip by \parskip too, later \aboveenvbreak}% \singlespace \parindent=0pt \let\Equotation = \nonfillfinish % @cartouche defines \nonarrowing to inhibit narrowing % at next level down. \ifx\nonarrowing\relax \advance \leftskip by \lispnarrowing \advance \rightskip by \lispnarrowing \exdentamount=\lispnarrowing \let\nonarrowing=\relax \fi} \message{defuns,} % Define formatter for defuns % First, allow user to change definition object font (\df) internally \def\setdeffont #1 {\csname DEF#1\endcsname} \newskip\defbodyindent \defbodyindent=.4in \newskip\defargsindent \defargsindent=50pt \newskip\deftypemargin \deftypemargin=12pt \newskip\deflastargmargin \deflastargmargin=18pt \newcount\parencount % define \functionparens, which makes ( and ) and & do special things. % \functionparens affects the group it is contained in. \def\activeparens{% \catcode`\(=\active \catcode`\)=\active \catcode`\&=\active \catcode`\[=\active \catcode`\]=\active} % Make control sequences which act like normal parenthesis chars. \let\lparen = ( \let\rparen = ) {\activeparens % Now, smart parens don't turn on until &foo (see \amprm) % Be sure that we always have a definition for `(', etc. For example, % if the fn name has parens in it, \boldbrax will not be in effect yet, % so TeX would otherwise complain about undefined control sequence. \global\let(=\lparen \global\let)=\rparen \global\let[=\lbrack \global\let]=\rbrack \gdef\functionparens{\boldbrax\let&=\amprm\parencount=0 } \gdef\boldbrax{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb} % Definitions of (, ) and & used in args for functions. % This is the definition of ( outside of all parentheses. \gdef\oprm#1 {{\rm\char`\(}#1 \bf \let(=\opnested % \global\advance\parencount by 1 } % % This is the definition of ( when already inside a level of parens. \gdef\opnested{\char`\(\global\advance\parencount by 1 } % \gdef\clrm{% Print a paren in roman if it is taking us back to depth of 0. % also in that case restore the outer-level definition of (. \ifnum \parencount=1 {\rm \char `\)}\sl \let(=\oprm \else \char `\) \fi \global\advance \parencount by -1 } % If we encounter &foo, then turn on ()-hacking afterwards \gdef\amprm#1 {{\rm\}\let(=\oprm \let)=\clrm\ } % \gdef\normalparens{\boldbrax\let&=\ampnr} } % End of definition inside \activeparens %% These parens (in \boldbrax) actually are a little bolder than the %% contained text. This is especially needed for [ and ] \def\opnr{{\sf\char`\(}} \def\clnr{{\sf\char`\)}} \def\ampnr{\&} \def\lbrb{{\bf\char`\[}} \def\rbrb{{\bf\char`\]}} % First, defname, which formats the header line itself. % #1 should be the function name. % #2 should be the type of definition, such as "Function". \def\defname #1#2{% % Get the values of \leftskip and \rightskip as they were % outside the @def... \dimen2=\leftskip \advance\dimen2 by -\defbodyindent \dimen3=\rightskip \advance\dimen3 by -\defbodyindent \noindent % \setbox0=\hbox{\hskip \deflastargmargin{\rm #2}\hskip \deftypemargin}% \dimen0=\hsize \advance \dimen0 by -\wd0 % compute size for first line \dimen1=\hsize \advance \dimen1 by -\defargsindent %size for continuations \parshape 2 0in \dimen0 \defargsindent \dimen1 % % Now output arg 2 ("Function" or some such) % ending at \deftypemargin from the right margin, % but stuck inside a box of width 0 so it does not interfere with linebreaking {% Adjust \hsize to exclude the ambient margins, % so that \rightline will obey them. \advance \hsize by -\dimen2 \advance \hsize by -\dimen3 \rlap{\rightline{{\rm #2}\hskip \deftypemargin}}}% % Make all lines underfull and no complaints: \tolerance=10000 \hbadness=10000 \advance\leftskip by -\defbodyindent \exdentamount=\defbodyindent {\df #1}\enskip % Generate function name } % Actually process the body of a definition % #1 should be the terminating control sequence, such as \Edefun. % #2 should be the "another name" control sequence, such as \defunx. % #3 should be the control sequence that actually processes the header, % such as \defunheader. \def\defparsebody #1#2#3{\begingroup\inENV% Environment for definitionbody \medbreak % % Define the end token that this defining construct specifies % so that it will exit this group. \def#1{\endgraf\endgroup\medbreak}% \def#2{\begingroup\obeylines\activeparens\spacesplit#3}% \parindent=0in \advance\leftskip by \defbodyindent \advance \rightskip by \defbodyindent \exdentamount=\defbodyindent \begingroup % \catcode 61=\active % 61 is `=' \obeylines\activeparens\spacesplit#3} \def\defmethparsebody #1#2#3#4 {\begingroup\inENV % \medbreak % % Define the end token that this defining construct specifies % so that it will exit this group. \def#1{\endgraf\endgroup\medbreak}% \def#2##1 {\begingroup\obeylines\activeparens\spacesplit{#3{##1}}}% \parindent=0in \advance\leftskip by \defbodyindent \advance \rightskip by \defbodyindent \exdentamount=\defbodyindent \begingroup\obeylines\activeparens\spacesplit{#3{#4}}} \def\defopparsebody #1#2#3#4#5 {\begingroup\inENV % \medbreak % % Define the end token that this defining construct specifies % so that it will exit this group. \def#1{\endgraf\endgroup\medbreak}% \def#2##1 ##2 {\def#4{##1}% \begingroup\obeylines\activeparens\spacesplit{#3{##2}}}% \parindent=0in \advance\leftskip by \defbodyindent \advance \rightskip by \defbodyindent \exdentamount=\defbodyindent \begingroup\obeylines\activeparens\spacesplit{#3{#5}}} % These parsing functions are similar to the preceding ones % except that they do not make parens into active characters. % These are used for "variables" since they have no arguments. \def\defvarparsebody #1#2#3{\begingroup\inENV% Environment for definitionbody \medbreak % % Define the end token that this defining construct specifies % so that it will exit this group. \def#1{\endgraf\endgroup\medbreak}% \def#2{\begingroup\obeylines\spacesplit#3}% \parindent=0in \advance\leftskip by \defbodyindent \advance \rightskip by \defbodyindent \exdentamount=\defbodyindent \begingroup % \catcode 61=\active % \obeylines\spacesplit#3} % This is used for \def{tp,vr}parsebody. It could probably be used for % some of the others, too, with some judicious conditionals. % \def\parsebodycommon#1#2#3{% \begingroup\inENV % \medbreak % % Define the end token that this defining construct specifies % so that it will exit this group. \def#1{\endgraf\endgroup\medbreak}% \def#2##1 {\begingroup\obeylines\spacesplit{#3{##1}}}% \parindent=0in \advance\leftskip by \defbodyindent \advance \rightskip by \defbodyindent \exdentamount=\defbodyindent \begingroup\obeylines } \def\defvrparsebody#1#2#3#4 {% \parsebodycommon{#1}{#2}{#3}% \spacesplit{#3{#4}}% } % This loses on `@deftp {Data Type} {struct termios}' -- it thinks the % type is just `struct', because we lose the braces in `{struct % termios}' when \spacesplit reads its undelimited argument. Sigh. % \let\deftpparsebody=\defvrparsebody % % So, to get around this, we put \empty in with the type name. That % way, TeX won't find exactly `{...}' as an undelimited argument, and % won't strip off the braces. % \def\deftpparsebody #1#2#3#4 {% \parsebodycommon{#1}{#2}{#3}% \spacesplit{\parsetpheaderline{#3{#4}}}\empty } % Fine, but then we have to eventually remove the \empty *and* the % braces (if any). That's what this does, putting the result in \tptemp. % \def\removeemptybraces\empty#1\relax{\def\tptemp{#1}}% % After \spacesplit has done its work, this is called -- #1 is the final % thing to call, #2 the type name (which starts with \empty), and #3 % (which might be empty) the arguments. % \def\parsetpheaderline#1#2#3{% \removeemptybraces#2\relax #1{\tptemp}{#3}% }% \def\defopvarparsebody #1#2#3#4#5 {\begingroup\inENV % \medbreak % % Define the end token that this defining construct specifies % so that it will exit this group. \def#1{\endgraf\endgroup\medbreak}% \def#2##1 ##2 {\def#4{##1}% \begingroup\obeylines\spacesplit{#3{##2}}}% \parindent=0in \advance\leftskip by \defbodyindent \advance \rightskip by \defbodyindent \exdentamount=\defbodyindent \begingroup\obeylines\spacesplit{#3{#5}}} % Split up #2 at the first space token. % call #1 with two arguments: % the first is all of #2 before the space token, % the second is all of #2 after that space token. % If #2 contains no space token, all of it is passed as the first arg % and the second is passed as empty. {\obeylines \gdef\spacesplit#1#2^^M{\endgroup\spacesplitfoo{#1}#2 \relax\spacesplitfoo}% \long\gdef\spacesplitfoo#1#2 #3#4\spacesplitfoo{% \ifx\relax #3% #1{#2}{}\else #1{#2}{#3#4}\fi}} % So much for the things common to all kinds of definitions. % Define @defun. % First, define the processing that is wanted for arguments of \defun % Use this to expand the args and terminate the paragraph they make up \def\defunargs #1{\functionparens \sl % Expand, preventing hyphenation at `-' chars. % Note that groups don't affect changes in \hyphenchar. \hyphenchar\tensl=0 #1% \hyphenchar\tensl=45 \ifnum\parencount=0 \else \errmessage{unbalanced parens in @def arguments}\fi% \interlinepenalty=10000 \advance\rightskip by 0pt plus 1fil \endgraf\penalty 10000\vskip -\parskip\penalty 10000% } \def\deftypefunargs #1{% % Expand, preventing hyphenation at `-' chars. % Note that groups don't affect changes in \hyphenchar. \functionparens \code{#1}% \interlinepenalty=10000 \advance\rightskip by 0pt plus 1fil \endgraf\penalty 10000\vskip -\parskip\penalty 10000% } % Do complete processing of one @defun or @defunx line already parsed. % @deffn Command forward-char nchars \def\deffn{\defmethparsebody\Edeffn\deffnx\deffnheader} \def\deffnheader #1#2#3{\doind {fn}{\code{#2}}% \begingroup\defname {#2}{#1}\defunargs{#3}\endgroup % \catcode 61=\other % Turn off change made in \defparsebody } % @defun == @deffn Function \def\defun{\defparsebody\Edefun\defunx\defunheader} \def\defunheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index \begingroup\defname {#1}{Function}% \defunargs {#2}\endgroup % \catcode 61=\other % Turn off change made in \defparsebody } % @deftypefun int foobar (int @var{foo}, float @var{bar}) \def\deftypefun{\defparsebody\Edeftypefun\deftypefunx\deftypefunheader} % #1 is the data type. #2 is the name and args. \def\deftypefunheader #1#2{\deftypefunheaderx{#1}#2 \relax} % #1 is the data type, #2 the name, #3 the args. \def\deftypefunheaderx #1#2 #3\relax{% \doind {fn}{\code{#2}}% Make entry in function index \begingroup\defname {\code{#1} #2}{Function}% \deftypefunargs {#3}\endgroup % \catcode 61=\other % Turn off change made in \defparsebody } % @deftypefn {Library Function} int foobar (int @var{foo}, float @var{bar}) \def\deftypefn{\defmethparsebody\Edeftypefn\deftypefnx\deftypefnheader} % #1 is the classification. #2 is the data type. #3 is the name and args. \def\deftypefnheader #1#2#3{\deftypefnheaderx{#1}{#2}#3 \relax} % #1 is the classification, #2 the data type, #3 the name, #4 the args. \def\deftypefnheaderx #1#2#3 #4\relax{% \doind {fn}{\code{#3}}% Make entry in function index \begingroup \normalparens % notably, turn off `&' magic, which prevents % at least some C++ text from working \defname {\code{#2} #3}{#1}% \deftypefunargs {#4}\endgroup % \catcode 61=\other % Turn off change made in \defparsebody } % @defmac == @deffn Macro \def\defmac{\defparsebody\Edefmac\defmacx\defmacheader} \def\defmacheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index \begingroup\defname {#1}{Macro}% \defunargs {#2}\endgroup % \catcode 61=\other % Turn off change made in \defparsebody } % @defspec == @deffn Special Form \def\defspec{\defparsebody\Edefspec\defspecx\defspecheader} \def\defspecheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index \begingroup\defname {#1}{Special Form}% \defunargs {#2}\endgroup % \catcode 61=\other % Turn off change made in \defparsebody } % This definition is run if you use @defunx % anywhere other than immediately after a @defun or @defunx. \def\deffnx #1 {\errmessage{@deffnx in invalid context}} \def\defunx #1 {\errmessage{@defunx in invalid context}} \def\defmacx #1 {\errmessage{@defmacx in invalid context}} \def\defspecx #1 {\errmessage{@defspecx in invalid context}} \def\deftypefnx #1 {\errmessage{@deftypefnx in invalid context}} \def\deftypeunx #1 {\errmessage{@deftypeunx in invalid context}} % @defmethod, and so on % @defop {Funny Method} foo-class frobnicate argument \def\defop #1 {\def\defoptype{#1}% \defopparsebody\Edefop\defopx\defopheader\defoptype} \def\defopheader #1#2#3{% \dosubind {fn}{\code{#2}}{on #1}% Make entry in function index \begingroup\defname {#2}{\defoptype{} on #1}% \defunargs {#3}\endgroup % } % @defmethod == @defop Method \def\defmethod{\defmethparsebody\Edefmethod\defmethodx\defmethodheader} \def\defmethodheader #1#2#3{% \dosubind {fn}{\code{#2}}{on #1}% entry in function index \begingroup\defname {#2}{Method on #1}% \defunargs {#3}\endgroup % } % @defcv {Class Option} foo-class foo-flag \def\defcv #1 {\def\defcvtype{#1}% \defopvarparsebody\Edefcv\defcvx\defcvarheader\defcvtype} \def\defcvarheader #1#2#3{% \dosubind {vr}{\code{#2}}{of #1}% Make entry in var index \begingroup\defname {#2}{\defcvtype{} of #1}% \defvarargs {#3}\endgroup % } % @defivar == @defcv {Instance Variable} \def\defivar{\defvrparsebody\Edefivar\defivarx\defivarheader} \def\defivarheader #1#2#3{% \dosubind {vr}{\code{#2}}{of #1}% Make entry in var index \begingroup\defname {#2}{Instance Variable of #1}% \defvarargs {#3}\endgroup % } % These definitions are run if you use @defmethodx, etc., % anywhere other than immediately after a @defmethod, etc. \def\defopx #1 {\errmessage{@defopx in invalid context}} \def\defmethodx #1 {\errmessage{@defmethodx in invalid context}} \def\defcvx #1 {\errmessage{@defcvx in invalid context}} \def\defivarx #1 {\errmessage{@defivarx in invalid context}} % Now @defvar % First, define the processing that is wanted for arguments of @defvar. % This is actually simple: just print them in roman. % This must expand the args and terminate the paragraph they make up \def\defvarargs #1{\normalparens #1% \interlinepenalty=10000 \endgraf\penalty 10000\vskip -\parskip\penalty 10000} % @defvr Counter foo-count \def\defvr{\defvrparsebody\Edefvr\defvrx\defvrheader} \def\defvrheader #1#2#3{\doind {vr}{\code{#2}}% \begingroup\defname {#2}{#1}\defvarargs{#3}\endgroup} % @defvar == @defvr Variable \def\defvar{\defvarparsebody\Edefvar\defvarx\defvarheader} \def\defvarheader #1#2{\doind {vr}{\code{#1}}% Make entry in var index \begingroup\defname {#1}{Variable}% \defvarargs {#2}\endgroup % } % @defopt == @defvr {User Option} \def\defopt{\defvarparsebody\Edefopt\defoptx\defoptheader} \def\defoptheader #1#2{\doind {vr}{\code{#1}}% Make entry in var index \begingroup\defname {#1}{User Option}% \defvarargs {#2}\endgroup % } % @deftypevar int foobar \def\deftypevar{\defvarparsebody\Edeftypevar\deftypevarx\deftypevarheader} % #1 is the data type. #2 is the name. \def\deftypevarheader #1#2{% \doind {vr}{\code{#2}}% Make entry in variables index \begingroup\defname {\code{#1} #2}{Variable}% \interlinepenalty=10000 \endgraf\penalty 10000\vskip -\parskip\penalty 10000 \endgroup} % @deftypevr {Global Flag} int enable \def\deftypevr{\defvrparsebody\Edeftypevr\deftypevrx\deftypevrheader} \def\deftypevrheader #1#2#3{\doind {vr}{\code{#3}}% \begingroup\defname {\code{#2} #3}{#1} \interlinepenalty=10000 \endgraf\penalty 10000\vskip -\parskip\penalty 10000 \endgroup} % This definition is run if you use @defvarx % anywhere other than immediately after a @defvar or @defvarx. \def\defvrx #1 {\errmessage{@defvrx in invalid context}} \def\defvarx #1 {\errmessage{@defvarx in invalid context}} \def\defoptx #1 {\errmessage{@defoptx in invalid context}} \def\deftypevarx #1 {\errmessage{@deftypevarx in invalid context}} \def\deftypevrx #1 {\errmessage{@deftypevrx in invalid context}} % Now define @deftp % Args are printed in bold, a slight difference from @defvar. \def\deftpargs #1{\bf \defvarargs{#1}} % @deftp Class window height width ... \def\deftp{\deftpparsebody\Edeftp\deftpx\deftpheader} \def\deftpheader #1#2#3{\doind {tp}{\code{#2}}% \begingroup\defname {#2}{#1}\deftpargs{#3}\endgroup} % This definition is run if you use @deftpx, etc % anywhere other than immediately after a @deftp, etc. \def\deftpx #1 {\errmessage{@deftpx in invalid context}} \message{cross reference,} % Define cross-reference macros \newwrite \auxfile \newif\ifhavexrefs % True if xref values are known. \newif\ifwarnedxrefs % True if we warned once that they aren't known. % \setref{foo} defines a cross-reference point named foo. \def\setref#1{% \dosetq{#1-title}{Ytitle}% \dosetq{#1-pg}{Ypagenumber}% \dosetq{#1-snt}{Ysectionnumberandtype}} \def\unnumbsetref#1{% \dosetq{#1-title}{Ytitle}% \dosetq{#1-pg}{Ypagenumber}% \dosetq{#1-snt}{Ynothing}} \def\appendixsetref#1{% \dosetq{#1-title}{Ytitle}% \dosetq{#1-pg}{Ypagenumber}% \dosetq{#1-snt}{Yappendixletterandtype}} % \xref, \pxref, and \ref generate cross-references to specified points. % For \xrefX, #1 is the node name, #2 the name of the Info % cross-reference, #3 the printed node name, #4 the name of the Info % file, #5 the name of the printed manual. All but the node name can be % omitted. % \def\pxref#1{see \xrefX[#1,,,,,,,]} \def\xref#1{See \xrefX[#1,,,,,,,]} \def\ref#1{\xrefX[#1,,,,,,,]} \def\xrefX[#1,#2,#3,#4,#5,#6]{\begingroup% \def\printedmanual{\ignorespaces #5}% \def\printednodename{\ignorespaces #3}% % \setbox1=\hbox{\printedmanual}% \setbox0=\hbox{\printednodename}% \ifdim \wd0=0pt% % No printed node name was explicitly given. \ifx SETxref-automatic-section-title % % This line should make the actual chapter or section title appear inside % the square brackets. Use the real section title if we have it. \ifdim \wd1>0pt% % It is in another manual, so we don't have it. \def\printednodename{\ignorespaces #1} \else% % We know the real title if we have the xref values. \ifhavexrefs \def\printednodename{\refx{#1-title}}% % Otherwise just copy the Info node name. \else \def\printednodename{\ignorespaces #1} \fi% \fi\def\printednodename{#1-title}% \else% This line just uses the node name. \def\printednodename{\ignorespaces #1}% \fi% ends \ifx SETxref-automatic-section-title \fi% ends \ifdim \wd0 % % % If we use \unhbox0 and \unhbox1 to print the node names, TeX does % not insert empty discretionaries after hyphens, which means that it % will not find a line break at a hyphen in a node names. Since some % manuals are best written with fairly long node names, containing % hyphens, this is a loss. Therefore, we simply give the text of % the node name again, so it is as if TeX is seeing it for the first % time. \ifdim \wd1>0pt section ``\printednodename'' in \cite{\printedmanual}% \else% \turnoffactive% \refx{#1-snt}{} [\printednodename], page\tie\refx{#1-pg}{}% \fi \endgroup} % \dosetq is the interface for calls from other macros % Use \turnoffactive so that punctuation chars such as underscore % work in node names. \def\dosetq #1#2{{\let\folio=0 \turnoffactive% \edef\next{\write\auxfile{\internalsetq {#1}{#2}}}% \next}} % \internalsetq {foo}{page} expands into % CHARACTERS 'xrdef {foo}{...expansion of \Ypage...} % When the aux file is read, ' is the escape character \def\internalsetq #1#2{'xrdef {#1}{\csname #2\endcsname}} % Things to be expanded by \internalsetq \def\Ypagenumber{\folio} \def\Ytitle{\thissection} \def\Ynothing{} \def\Ysectionnumberandtype{% \ifnum\secno=0 Chapter\xreftie\the\chapno % \else \ifnum \subsecno=0 Section\xreftie\the\chapno.\the\secno % \else \ifnum \subsubsecno=0 % Section\xreftie\the\chapno.\the\secno.\the\subsecno % \else % Section\xreftie\the\chapno.\the\secno.\the\subsecno.\the\subsubsecno % \fi \fi \fi } \def\Yappendixletterandtype{% \ifnum\secno=0 Appendix\xreftie'char\the\appendixno{}% \else \ifnum \subsecno=0 Section\xreftie'char\the\appendixno.\the\secno % \else \ifnum \subsubsecno=0 % Section\xreftie'char\the\appendixno.\the\secno.\the\subsecno % \else % Section\xreftie'char\the\appendixno.\the\secno.\the\subsecno.\the\subsubsecno % \fi \fi \fi } \gdef\xreftie{'tie} % Use TeX 3.0's \inputlineno to get the line number, for better error % messages, but if we're using an old version of TeX, don't do anything. % \ifx\inputlineno\thisisundefined \let\linenumber = \empty % Non-3.0. \else \def\linenumber{\the\inputlineno:\space} \fi % Define \refx{NAME}{SUFFIX} to reference a cross-reference string named NAME. % If its value is nonempty, SUFFIX is output afterward. \def\refx#1#2{% \expandafter\ifx\csname X#1\endcsname\relax % If not defined, say something at least. $\langle$un\-de\-fined$\rangle$% \ifhavexrefs \message{\linenumber Undefined cross reference `#1'.}% \else \ifwarnedxrefs\else \global\warnedxrefstrue \message{Cross reference values unknown; you must run TeX again.}% \fi \fi \else % It's defined, so just use it. \csname X#1\endcsname \fi #2% Output the suffix in any case. } % Read the last existing aux file, if any. No error if none exists. % This is the macro invoked by entries in the aux file. \def\xrdef #1#2{ {\catcode`\'=\other\expandafter \gdef \csname X#1\endcsname {#2}}} \def\readauxfile{% \begingroup \catcode `\^^@=\other \catcode `\=\other \catcode `\=\other \catcode `\^^C=\other \catcode `\^^D=\other \catcode `\^^E=\other \catcode `\^^F=\other \catcode `\^^G=\other \catcode `\^^H=\other \catcode `\ =\other \catcode `\^^L=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode `\=\other \catcode 26=\other \catcode `\^^[=\other \catcode `\^^\=\other \catcode `\^^]=\other \catcode `\^^^=\other \catcode `\^^_=\other \catcode `\@=\other \catcode `\^=\other \catcode `\~=\other \catcode `\[=\other \catcode `\]=\other \catcode`\"=\other \catcode`\_=\other \catcode`\|=\other \catcode`\<=\other \catcode`\>=\other \catcode `\$=\other \catcode `\#=\other \catcode `\&=\other % `\+ does not work, so use 43. \catcode 43=\other % the aux file uses ' as the escape. % Turn off \ as an escape so we do not lose on % entries which were dumped with control sequences in their names. % For example, 'xrdef {$\leq $-fun}{page ...} made by @defun ^^ % Reference to such entries still does not work the way one would wish, % but at least they do not bomb out when the aux file is read in. \catcode `\{=1 \catcode `\}=2 \catcode `\%=\other \catcode `\'=0 \catcode `\\=\other \openin 1 \jobname.aux \ifeof 1 \else \closein 1 \input \jobname.aux \global\havexrefstrue \global\warnedobstrue \fi % Open the new aux file. Tex will close it automatically at exit. \openout \auxfile=\jobname.aux \endgroup} % Footnotes. \newcount \footnoteno % The trailing space in the following definition for supereject is % vital for proper filling; pages come out unaligned when you do a % pagealignmacro call if that space before the closing brace is % removed. \def\supereject{\par\penalty -20000\footnoteno =0 } % @footnotestyle is meaningful for info output only.. \let\footnotestyle=\comment \let\ptexfootnote=\footnote {\catcode `\@=11 % % Auto-number footnotes. Otherwise like plain. \gdef\footnote{% \global\advance\footnoteno by \@ne \edef\thisfootno{$^{\the\footnoteno}$}% % % In case the footnote comes at the end of a sentence, preserve the % extra spacing after we do the footnote number. \let\@sf\empty \ifhmode\edef\@sf{\spacefactor\the\spacefactor}\/\fi % % Remove inadvertent blank space before typesetting the footnote number. \unskip \thisfootno\@sf \footnotezzz }% % Don't bother with the trickery in plain.tex to not require the % footnote text as a parameter. Our footnotes don't need to be so general. % \long\gdef\footnotezzz#1{\insert\footins{% % We want to typeset this text as a normal paragraph, even if the % footnote reference occurs in (for example) a display environment. % So reset some parameters. \interlinepenalty\interfootnotelinepenalty \splittopskip\ht\strutbox % top baseline for broken footnotes \splitmaxdepth\dp\strutbox \floatingpenalty\@MM \leftskip\z@skip \rightskip\z@skip \spaceskip\z@skip \xspaceskip\z@skip \parindent\defaultparindent % % Hang the footnote text off the number. \hang \textindent{\thisfootno}% % % Don't crash into the line above the footnote text. Since this % expands into a box, it must come within the paragraph, lest it % provide a place where TeX can split the footnote. \footstrut #1\strut}% } }%end \catcode `\@=11 % Set the baselineskip to #1, and the lineskip and strut size % correspondingly. There is no deep meaning behind these magic numbers % used as factors; they just match (closely enough) what Knuth defined. % \def\lineskipfactor{.08333} \def\strutheightpercent{.70833} \def\strutdepthpercent {.29167} % \def\setleading#1{% \normalbaselineskip = #1\relax \normallineskip = \lineskipfactor\normalbaselineskip \normalbaselines \setbox\strutbox =\hbox{% \vrule width0pt height\strutheightpercent\baselineskip depth \strutdepthpercent \baselineskip }% } % @| inserts a changebar to the left of the current line. It should % surround any changed text. This approach does *not* work if the % change spans more than two lines of output. To handle that, we would % have adopt a much more difficult approach (putting marks into the main % vertical list for the beginning and end of each change). % \def\|{% % \vadjust can only be used in horizontal mode. \leavevmode % % Append this vertical mode material after the current line in the output. \vadjust{% % We want to insert a rule with the height and depth of the current % leading; that is exactly what \strutbox is supposed to record. \vskip-\baselineskip % % \vadjust-items are inserted at the left edge of the type. So % the \llap here moves out into the left-hand margin. \llap{% % % For a thicker or thinner bar, change the `1pt'. \vrule height\baselineskip width1pt % % This is the space between the bar and the text. \hskip 12pt }% }% } % For a final copy, take out the rectangles % that mark overfull boxes (in case you have decided % that the text looks ok even though it passes the margin). % \def\finalout{\overfullrule=0pt} % End of control word definitions. \message{and turning on texinfo input format.} \def\openindices{% \newindex{cp}% \newcodeindex{fn}% \newcodeindex{vr}% \newcodeindex{tp}% \newcodeindex{ky}% \newcodeindex{pg}% } % Set some numeric style parameters, for 8.5 x 11 format. %\hsize = 6.5in \newdimen\defaultparindent \defaultparindent = 15pt \parindent = \defaultparindent \parskip 18pt plus 1pt \setleading{15pt} \advance\topskip by 1.2cm % Prevent underfull vbox error messages. \vbadness=10000 % Following George Bush, just get rid of widows and orphans. \widowpenalty=10000 \clubpenalty=10000 % Use TeX 3.0's \emergencystretch to help line breaking, but if we're % using an old version of TeX, don't do anything. We want the amount of % stretch added to depend on the line length, hence the dependence on % \hsize. This makes it come to about 9pt for the 8.5x11 format. % \ifx\emergencystretch\thisisundefined % Allow us to assign to \emergencystretch anyway. \def\emergencystretch{\dimen0}% \else \emergencystretch = \hsize \divide\emergencystretch by 45 \fi % Use @smallbook to reset parameters for 7x9.5 format (or else 7x9.25) \def\smallbook{ % These values for secheadingskip and subsecheadingskip are % experiments. RJC 7 Aug 1992 \global\secheadingskip = 17pt plus 6pt minus 3pt \global\subsecheadingskip = 14pt plus 6pt minus 3pt \global\lispnarrowing = 0.3in \setleading{12pt} \advance\topskip by -1cm \global\parskip 3pt plus 1pt \global\hsize = 5in \global\vsize=7.5in \global\tolerance=700 \global\hfuzz=1pt \global\contentsrightmargin=0pt \global\pagewidth=\hsize \global\pageheight=\vsize \global\let\smalllisp=\smalllispx \global\let\smallexample=\smalllispx \global\def\Esmallexample{\Esmalllisp} } % Use @afourpaper to print on European A4 paper. \def\afourpaper{ \global\tolerance=700 \global\hfuzz=1pt \setleading{12pt} \global\parskip 15pt plus 1pt \global\vsize= 53\baselineskip \advance\vsize by \topskip %\global\hsize= 5.85in % A4 wide 10pt \global\hsize= 6.5in \global\outerhsize=\hsize \global\advance\outerhsize by 0.5in \global\outervsize=\vsize \global\advance\outervsize by 0.6in \global\pagewidth=\hsize \global\pageheight=\vsize } % Define macros to output various characters with catcode for normal text. \catcode`\"=\other \catcode`\~=\other \catcode`\^=\other \catcode`\_=\other \catcode`\|=\other \catcode`\<=\other \catcode`\>=\other \catcode`\+=\other \def\normaldoublequote{"} \def\normaltilde{~} \def\normalcaret{^} \def\normalunderscore{_} \def\normalverticalbar{|} \def\normalless{<} \def\normalgreater{>} \def\normalplus{+} % This macro is used to make a character print one way in ttfont % where it can probably just be output, and another way in other fonts, % where something hairier probably needs to be done. % % #1 is what to print if we are indeed using \tt; #2 is what to print % otherwise. Since all the Computer Modern typewriter fonts have zero % interword stretch (and shrink), and it is reasonable to expect all % typewriter fonts to have this, we can check that font parameter. % \def\ifusingtt#1#2{\ifdim \fontdimen3\the\font=0pt #1\else #2\fi} % Turn off all special characters except @ % (and those which the user can use as if they were ordinary). % Most of these we simply print from the \tt font, but for some, we can % use math or other variants that look better in normal text. \catcode`\"=\active \def\activedoublequote{{\tt \char '042}} \let"=\activedoublequote \catcode`\~=\active \def~{{\tt \char '176}} \chardef\hat=`\^ \catcode`\^=\active \def^{{\tt \hat}} \catcode`\_=\active \def_{\ifusingtt\normalunderscore\_} % Subroutine for the previous macro. \def\_{\lvvmode \kern.06em \vbox{\hrule width.3em height.1ex}} % \lvvmode is equivalent in function to \leavevmode. % Using \leavevmode runs into trouble when written out to % an index file due to the expansion of \leavevmode into ``\unhbox % \voidb@x'' ---which looks to TeX like ``\unhbox \voidb\x'' due to our % magic tricks with @. \def\lvvmode{\vbox to 0pt{}} \catcode`\|=\active \def|{{\tt \char '174}} \chardef \less=`\< \catcode`\<=\active \def<{{\tt \less}} \chardef \gtr=`\> \catcode`\>=\active \def>{{\tt \gtr}} \catcode`\+=\active \def+{{\tt \char 43}} %\catcode 27=\active %\def^^[{$\diamondsuit$} % Used sometimes to turn off (effectively) the active characters % even after parsing them. \def\turnoffactive{\let"=\normaldoublequote \let~=\normaltilde \let^=\normalcaret \let_=\normalunderscore \let|=\normalverticalbar \let<=\normalless \let>=\normalgreater \let+=\normalplus} % Set up an active definition for =, but don't enable it most of the time. {\catcode`\==\active \global\def={{\tt \char 61}}} \catcode`\@=0 % \rawbackslashxx output one backslash character in current font \global\chardef\rawbackslashxx=`\\ %{\catcode`\\=\other %@gdef@rawbackslashxx{\}} % \rawbackslash redefines \ as input to do \rawbackslashxx. {\catcode`\\=\active @gdef@rawbackslash{@let\=@rawbackslashxx }} % \normalbackslash outputs one backslash in fixed width font. \def\normalbackslash{{\tt\rawbackslashxx}} % Say @foo, not \foo, in error messages. \escapechar=`\@ % \catcode 17=0 % Define control-q \catcode`\\=\active % If a .fmt file is being used, we don't want the `\input texinfo' to show up. % That is what \eatinput is for; after that, the `\' should revert to printing % a backslash. % @gdef@eatinput input texinfo{@fixbackslash} @global@let\ = @eatinput % On the other hand, perhaps the file did not have a `\input texinfo'. Then % the first `\{ in the file would cause an error. This macro tries to fix % that, assuming it is called before the first `\' could plausibly occur. % @gdef@fixbackslash{@ifx\@eatinput @let\ = @normalbackslash @fi} %% These look ok in all fonts, so just make them not special. The @rm below %% makes sure that the current font starts out as the newly loaded cmr10 @catcode`@$=@other @catcode`@%=@other @catcode`@&=@other @catcode`@#=@other @textfonts @rm @c Local variables: @c page-delimiter: "^\\\\message" @c End: autoconf-1.7/acconfig.h100666 21570 13 15742 5455560366 13200 0ustar djmuser/* acconfig.h This file is in the public domain. Descriptive text for the C preprocessor macros that the distributed Autoconf macros can define. No software package will use all of them; autoheader copies the ones your configure.in uses into your configuration header file templates. The entries are in sort -df order: alphabetical, case insensitive, ignoring punctuation (such as underscores). Although this order can split up related entries, it makes it easier to check whether a given entry is in the file. Leave the following blank line there!! Autoheader needs it. */ /* Define if on AIX 3. System headers sometimes define this. We just want to avoid a redefinition error message. */ #ifndef _ALL_SOURCE #undef _ALL_SOURCE #endif /* Define if using alloca.c. */ #undef C_ALLOCA /* Define if type char is unsigned and you are not using gcc. */ #undef __CHAR_UNSIGNED__ /* Define to empty if the keyword does not work. */ #undef const /* Define to one of _getb67, GETB67, getb67 for Cray-2 and Cray-YMP systems. This function is required for alloca.c support on those systems. */ #undef CRAY_STACKSEG_END /* Define as the proper declaration for yytext. */ #undef DECLARE_YYTEXT /* Define for DGUX with . */ #undef DGUX /* Define if you have dirent.h. */ #undef DIRENT /* Define to the type of elements in the array set by `getgroups'. Usually this is either `int' or `gid_t'. */ #undef GETGROUPS_T /* Define if the `getloadavg' function needs to be run setuid or setgid. */ #undef GETLOADAVG_PRIVILEGED /* Define to `int' if doesn't define. */ #undef gid_t /* Define if you have alloca.h and it should be used (not Ultrix). */ #undef HAVE_ALLOCA_H /* Define if you don't have vprintf but do have _doprnt. */ #undef HAVE_DOPRNT /* Define if your system has its own `getloadavg' function. */ #undef HAVE_GETLOADAVG /* Define if the `long double' type works. */ #undef HAVE_LONG_DOUBLE /* Define if you support file names longer than 14 characters. */ #undef HAVE_LONG_FILE_NAMES /* Define if you have netdb.h (and thus the rexec function). */ #undef HAVE_NETDB_H /* Define if system calls automatically restart after interruption by a signal. */ #undef HAVE_RESTARTABLE_SYSCALLS /* Define if your struct stat has st_blksize. */ #undef HAVE_ST_BLKSIZE /* Define if your struct stat has st_blocks. */ #undef HAVE_ST_BLOCKS /* Define if you have the strcoll function and it is properly defined. */ #undef HAVE_STRCOLL /* Define if your struct stat has st_rdev. */ #undef HAVE_ST_RDEV /* Define if you have sys/mtio.h. */ #undef HAVE_SYS_MTIO_H /* Define if you have a with `union wait'. */ #undef HAVE_SYS_WAIT /* Define if your struct tm has tm_zone. */ #undef HAVE_TM_ZONE /* Define if you don't have tm_zone but do have the external array tzname. */ #undef HAVE_TZNAME /* Define if `union wait' is the type of the first arg to wait functions. */ #undef HAVE_UNION_WAIT /* Define if you have unistd.h. */ #undef HAVE_UNISTD_H /* Define if utime(file, NULL) sets file's timestamp to the present. */ #undef HAVE_UTIME_NULL /* Define if you have vfork.h. */ #undef HAVE_VFORK_H /* Define if you have the vprintf function. */ #undef HAVE_VPRINTF /* Define if you have the wait3 system call. */ #undef HAVE_WAIT3 /* Define if you have the POSIX.1 `waitpid' function. */ #undef HAVE_WAITPID /* Define as __inline if that's what the C compiler calls it. */ #undef inline /* Define if int is 16 bits instead of 32. */ #undef INT_16_BITS /* Define if long int is 64 bits. */ #undef LONG_64_BITS /* Define if major, minor, and makedev are declared in mkdev.h. */ #undef MAJOR_IN_MKDEV /* Define if major, minor, and makedev are declared in sysmacros.h. */ #undef MAJOR_IN_SYSMACROS /* Define if on MINIX. */ #undef _MINIX /* Define to `int' if doesn't define. */ #undef mode_t /* Define if you don't have dirent.h, but have ndir.h. */ #undef NDIR /* Define if you have memory.h, and string.h doesn't declare the mem* functions. */ #undef NEED_MEMORY_H /* Define if your struct nlist has an n_un member. */ #undef NLIST_NAME_UNION /* Define if you have nlist.h. */ #undef NLIST_STRUCT /* Define if you can't use the address of an argument to a function as the start of an array. */ #undef NO_ARG_ARRAY /* Define if your C compiler doesn't accept -c and -o together. */ #undef NO_MINUS_C_MINUS_O /* Define if you have neither a remote shell nor the rexec function. */ #undef NO_REMOTE /* Define to `long' if doesn't define. */ #undef off_t /* Define to `int' if doesn't define. */ #undef pid_t /* Define if the system does not provide POSIX.1 features except with this defined. */ #undef _POSIX_1_SOURCE /* Define if you need to in order for stat and other things to work. */ #undef _POSIX_SOURCE /* Define as the return type of signal handlers (int or void). */ #undef RETSIGTYPE /* Define if the setvbuf function takes the buffering type as its second argument and the buffer pointer as the third, as on System V before release 3. */ #undef SETVBUF_REVERSED /* Define to `unsigned' if doesn't define. */ #undef size_t /* If using the C implementation of alloca, define if you know the direction of stack growth for your system; otherwise it will be automatically deduced at run-time. STACK_DIRECTION > 0 => grows toward higher addresses STACK_DIRECTION < 0 => grows toward lower addresses STACK_DIRECTION = 0 => direction of growth unknown */ #undef STACK_DIRECTION /* Define if the `S_IS*' macros in do not work properly. */ #undef STAT_MACROS_BROKEN /* Define if you have the ANSI C header files. */ #undef STDC_HEADERS /* Define on System V Release 4. */ #undef SVR4 /* Define if you don't have dirent.h, but have sys/dir.h. */ #undef SYSDIR /* Define if you don't have dirent.h, but have sys/ndir.h. */ #undef SYSNDIR /* Define if `sys_siglist' is declared by . */ #undef SYS_SIGLIST_DECLARED /* Define if you can safely include both and . */ #undef TIME_WITH_SYS_TIME /* Define if your declares struct tm. */ #undef TM_IN_SYS_TIME /* Define to `int' if doesn't define. */ #undef uid_t /* Define for Encore UMAX. */ #undef UMAX /* Define for Encore UMAX 4.3 that has instead of . */ #undef UMAX4_3 /* Define if you do not have strings.h, index, bzero, etc.. */ #undef USG /* Define vfork as fork if vfork does not work. */ #undef vfork /* Define if the closedir function returns void instead of int. */ #undef VOID_CLOSEDIR /* Define if your processor stores words with the most significant byte first (like Motorola and SPARC, unlike Intel and VAX). */ #undef WORDS_BIGENDIAN /* Leave that blank line there!! Autoheader needs it. If you're adding to this file, keep in mind: The entries are in sort -df order: alphabetical, case insensitive, ignoring punctuation (such as underscores). */ autoconf-1.7/autoconf.sh100666 74504 13 6675 5456046735 14467 0ustar meyeringuser#!/bin/sh # autoconf -- create `configure' using m4 macros # Copyright (C) 1992, 1993 Free Software Foundation, Inc. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. # If given no args, create `configure' from template file `configure.in'. # With one arg, create a configure script on standard output from # the given template file. usage="Usage: autoconf [-h] [--help] [-m dir] [--macrodir=dir] [-v] [--version] [template-file]" # NLS nuisances. # These must not be set unconditionally because not all systems understand # e.g. LANG=C (notably SCO). if test "${LC_ALL+set}" = 'set' ; then LC_ALL=C; export LC_ALL; fi if test "${LANG+set}" = 'set' ; then LANG=C; export LANG; fi test -z "${AC_MACRODIR}" && AC_MACRODIR=@datadir@ test -z "${M4}" && M4=@M4@ tmpout=/tmp/acout.$$ print_version= while test $# -gt 0 ; do case "z${1}" in z-h | z--help | z--h* ) echo "${usage}" 1>&2; exit 0 ;; z--macrodir=* | z--m*=* ) AC_MACRODIR="`echo \"${1}\" | sed -e 's/^[^=]*=//'`" shift ;; z-m | z--macrodir | z--m* ) shift test $# -eq 0 && { echo "${usage}" 1>&2; exit 1; } AC_MACRODIR="${1}" shift ;; z-v | z--version | z--v* ) print_version="-DAC_PRINT_VERSION" infile=/dev/null tmpout=/dev/null shift ;; z-- ) # Stop option processing shift; break ;; z- ) # Use stdin as input. break ;; z-* ) echo "${usage}" 1>&2; exit 1 ;; * ) break ;; esac done if test -z "$print_version"; then case $# in 0) infile=configure.in ;; 1) infile="$1" ;; *) echo "$usage" >&2; exit 1 ;; esac if test z$infile = z-; then infile=/tmp/acin.$$ trap 'rm -f $infile' 1 3 15 cat > $infile elif test ! -s "${infile}"; then echo "autoconf: ${infile}: No such file or directory" >&2 exit 1 fi trap 'rm -f $tmpout; exit 1' 1 3 15 fi MACROFILES="${AC_MACRODIR}/acgeneral.m4 ${AC_MACRODIR}/acspecific.m4" test -r ${AC_MACRODIR}/aclocal.m4 \ && MACROFILES="${MACROFILES} ${AC_MACRODIR}/aclocal.m4" test -r aclocal.m4 && MACROFILES="${MACROFILES} aclocal.m4" MACROFILES="${print_version} ${MACROFILES}" $M4 $MACROFILES $infile > $tmpout test -n "$print_version" && exit 0 # You could add your own prefixes to pattern if you wanted to check for # them too, e.g. pattern="AC_\|ILT_", except that UNIX sed doesn't do # alternation, and GNU sed is dreadfully slow. Sigh. pattern="AC_" status=0 if grep "${pattern}" $tmpout > /dev/null 2>&1; then echo "autoconf: Undefined macros:" >&2 grep "${pattern}" $tmpout | sed "s/.*\(${pattern}[_A-Z0-9]*\).*/\1/" | while read name; do grep -n $name $infile /dev/null done | sort -u >&2 status=1 fi case $# in 0) cat $tmpout > configure; chmod +x configure ;; 1) cat $tmpout ;; esac rm -f $tmpout exit $status autoconf-1.7/acgeneral.m4100666 21570 1 61710 5463274656 13702 0ustar djmdaemondnl Parameterized macros that do not check for something specific. dnl This file is part of Autoconf. dnl Copyright (C) 1992, 1993 Free Software Foundation, Inc. dnl dnl This program is free software; you can redistribute it and/or modify dnl it under the terms of the GNU General Public License as published by dnl the Free Software Foundation; either version 2, or (at your option) dnl any later version. dnl dnl This program is distributed in the hope that it will be useful, dnl but WITHOUT ANY WARRANTY; without even the implied warranty of dnl MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the dnl GNU General Public License for more details. dnl dnl You should have received a copy of the GNU General Public License dnl along with this program; if not, write to the Free Software dnl Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. dnl dnl Written by David MacKenzie, with help from dnl Franc,ois Pinard, Karl Berry, Richard Pixley, Ian Lance Taylor, dnl Roland McGrath, and Noah Friedman. dnl changequote([,])dnl undefine([eval])dnl undefine([include])dnl undefine([shift])dnl undefine([format])dnl dnl ifdef([__gnu__], , [errprint(Autoconf requires GNU m4 )exit(1)])dnl dnl dnl dnl Utility functions for stamping the configure script. dnl dnl define(AC_ACVERSION, 1.7)dnl dnl This is defined by the --version option of the autoconf script. ifdef([AC_PRINT_VERSION], [errprint(Autoconf version AC_ACVERSION )])dnl dnl dnl These are currently not used, for the sake of people who diff dnl configure scripts and don't want spurious differences. dnl But they are too clever to just delete. dnl define(AC_USER, [esyscmd( changequote({,})dnl # Extract the user name from the first pair of parentheses. ({sedcmd='s/[^(]*(\([^)]*\)).*/\1/';} changequote([,])dnl whoami || id|sed "$sedcmd") 2>/dev/null|tr -d '\012')])dnl dnl define(AC_HOST, [esyscmd((hostname || uname -n) 2>/dev/null|tr -d '\012')])dnl dnl define(AC_DATE, [esyscmd(date|tr -d '\012')])dnl dnl dnl dnl Controlling Autoconf operation dnl dnl dnl This is separate from AC_INIT to prevent GNU m4 1.0 from coredumping dnl when AC_CONFIG_HEADER is used. define(AC_NOTICE, [# Guess values for system-dependent variables and create Makefiles. # Generated automatically using autoconf. dnl [#] Generated automatically using autoconf version AC_ACVERSION dnl [#] by AC_USER@AC_HOST on AC_DATE # Copyright (C) 1991, 1992, 1993 Free Software Foundation, Inc. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. # Usage: configure [--srcdir=DIR] [--host=HOST] [--gas] [--nfp] # [--prefix=PREFIX] [--exec-prefix=PREFIX] [--with-PACKAGE[=VALUE]] # Ignores all args except --srcdir, --prefix, --exec-prefix, and # --with-PACKAGE[=VALUE] unless this script has special code to handle it. ])dnl dnl define(AC_PARSEARGS, [for arg do # Handle --exec-prefix with a space before the argument. if test x$next_exec_prefix = xyes; then exec_prefix=$arg; next_exec_prefix= # Handle --host with a space before the argument. elif test x$next_host = xyes; then next_host= # Handle --prefix with a space before the argument. elif test x$next_prefix = xyes; then prefix=$arg; next_prefix= # Handle --srcdir with a space before the argument. elif test x$next_srcdir = xyes; then srcdir=$arg; next_srcdir= else case $arg in # For backward compatibility, recognize -exec-prefix and --exec_prefix. -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* | --exec=* | --exe=* | --ex=* | --e=*) changequote(,)dnl exec_prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; changequote([,])dnl -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- | --exec | --exe | --ex | --e) next_exec_prefix=yes ;; -gas | --gas | --ga | --g) ;; -host=* | --host=* | --hos=* | --ho=* | --h=*) ;; -host | --host | --hos | --ho | --h) next_host=yes ;; -nfp | --nfp | --nf) ;; -prefix=* | --prefix=* | --prefi=* | --pref=* | --pre=* | --pr=* | --p=*) changequote(,)dnl prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; changequote([,])dnl -prefix | --prefix | --prefi | --pref | --pre | --pr | --p) next_prefix=yes ;; -srcdir=* | --srcdir=* | --srcdi=* | --srcd=* | --src=* | --sr=* | --s=*) changequote(,)dnl srcdir=`echo $arg | sed 's/[-a-z_]*=//'` ;; changequote([,])dnl -srcdir | --srcdir | --srcdi | --srcd | --src | --sr | --s) next_srcdir=yes ;; -with-* | --with-*) package=`echo $arg|sed -e 's/-*with-//' -e 's/=.*//'` # Reject names that aren't valid shell variable names. changequote(,)dnl if test -n "`echo $package| sed 's/[-a-zA-Z0-9_]//g'`"; then changequote([,])dnl echo "configure: $package: invalid package name" >&2; exit 1 fi package=`echo $package| sed 's/-/_/g'` case "$arg" in changequote(,)dnl *=*) val="`echo $arg|sed 's/[^=]*=//'`" ;; changequote([,])dnl *) val=1 ;; esac eval "with_$package='$val'" ;; -v | -verbose | --verbose | --verbos | --verbo | --verb | --ver | --ve | --v) verbose=yes ;; *) ;; esac fi done ])dnl dnl define(AC_INIT, [#!/bin/sh AC_NOTICE AC_PARSEARGS AC_PREPARE($1)])dnl dnl define(AC_PREPARE, [trap 'rm -fr conftest* confdefs* core; exit 1' 1 3 15 trap 'rm -f confdefs*' 0 # NLS nuisances. # These must not be set unconditionally because not all systems understand # e.g. LANG=C (notably SCO). if test "${LC_ALL+set}" = 'set' ; then LC_ALL=C; export LC_ALL; fi if test "${LANG+set}" = 'set' ; then LANG=C; export LANG; fi rm -f conftest* confdefs.h # AIX cpp loses on an empty file, so make sure it contains at least a newline. echo > confdefs.h compile='${CC-cc} $CFLAGS conftest.c -o conftest $LIBS >/dev/null 2>&1' # A filename unique to this package, relative to the directory that # configure is in, which we can look for to find out if srcdir is correct. unique_file=$1 # Find the source files, if location was not specified. if test -z "$srcdir"; then srcdirdefaulted=yes # Try the directory containing this script, then `..'. prog=[$]0 changequote(,)dnl confdir=`echo $prog|sed 's%/[^/][^/]*$%%'` changequote([,])dnl test "X$confdir" = "X$prog" && confdir=. srcdir=$confdir if test ! -r $srcdir/$unique_file; then srcdir=.. fi fi if test ! -r $srcdir/$unique_file; then if test x$srcdirdefaulted = xyes; then echo "configure: Can not find sources in \`${confdir}' or \`..'." 1>&2 else echo "configure: Can not find sources in \`${srcdir}'." 1>&2 fi exit 1 fi # Preserve a srcdir of `.' to avoid automounter screwups with pwd. # But we can't avoid them for `..', to make subdirectories work. case $srcdir in .|/*|~*) ;; *) srcdir=`cd $srcdir; pwd` ;; # Make relative path absolute. esac dnl This is effectively disabled because it is a nuisance if not using dnl a configuration header file. ifelse([AC_CONFIG_NAMES], [], [ # Check for an already configured srcdir. if test -r $srcdir/config.status && \ test z"`sh -c pwd`" != z"`(cd $srcdir; sh -c pwd)`"; then cat <&2 configure: The directory tree \`${srcdir}' is being used as a build directory right now; it has been configured in its own right. You cannot currently configure and build in both the source directory and another build directory. To configure in this directory, you must first do \`make distclean' in ${srcdir}, and then run configure again. EOF exit 1 ifelse([not ready yet], [], [ XXX this should do the following: * Edit "VPATH=..." into GNU vpath directives for %.c and %.h (%.y?). * Add -DCONFIG_BROKETS to DEFS (config.h is not good enough, because it is used to FIND config.h). cat <&2 configure: WARNING: The directory tree \`${srcdir}' is being used as a build directory right now; it has been configured in its own right. To configure in another directory as well, you MUST use GNU make. If you do not have GNU make, then you must now do \`make distclean' in ${srcdir}, and then run ${progname} again. EOF])dnl fi])dnl # Save the original args to write them into config.status later. configure_args="[$]*" ])dnl dnl dnl Protects the argument from being diverted twice dnl if this macro is called twice for it. dnl Diversion 0 is the normal output. dnl Diversion 1 is sed substitutions for output files. dnl Diversion 2 is variable assignments for config.status. define(AC_SUBST, [ifdef([AC_SUBST_$1], , [define([AC_SUBST_$1], )dnl divert(1)dnl s%@$1@%[$]$1%g divert(2)dnl $1='[$]$1' divert(0)dnl ])])dnl dnl define(AC_WITH, [[#] check whether --with-$1 was given withval="[$with_]patsubst($1,-,_)" if test -n "$withval"; then ifelse([$2], , :, [$2]) ifelse([$3], , , [else $3 ])dnl fi ])dnl dnl dnl Guess the value for the `prefix' variable by looking for dnl the argument program along PATH and taking its parent. dnl Example: if the argument is `gcc' and we find /usr/local/gnu/bin/gcc, dnl set `prefix' to /usr/local/gnu. define(AC_PREFIX, [if test -z "$prefix" then echo checking for $1 to derive installation directory prefix IFS="${IFS= }"; saveifs="$IFS"; IFS="$IFS:" for dir in $PATH; do test -z "$dir" && dir=. if test $dir != . && test -f $dir/$1; then changequote(,)dnl # Not all systems have dirname. prefix=`echo $dir|sed 's%/[^/][^/]*$%%'` changequote([,])dnl break fi done IFS="$saveifs" echo " chose installation directory prefix ${prefix}" fi ])dnl dnl define(AC_CONFIG_HEADER, [define(AC_CONFIG_NAMES, $1)])dnl dnl define(AC_DOREV, [#!/bin/sh # From configure.in $1 ])dnl define(AC_REVISION, [AC_DOREV(translit($1,$"))])dnl dnl dnl dnl Several simple subroutines to do various flavors of quoting. dnl dnl Quote $1 against shell "s. define(AC_QUOTE_DQUOTE, [dnl We use \1 instead of \& to avoid an m4 1.0.3 bug. patsubst($1, changequote(,)\([$"`\\]\)changequote([,]), \\\1)])dnl dnl dnl Quote $1 against shell 's. define(AC_QUOTE_SQUOTE, [patsubst($1, ', '\\'')])dnl dnl dnl Quote $1 against shell here documents (<> confdefs.h dnl Define DEFS even if AC_CONFIG_NAMES for use in user case statements. DEFS="$DEFS -D$1=AC_DEFINE_QUOTE(AC_VAL)" ifdef([AC_CONFIG_NAMES], SEDDEFS="dnl ${SEDDEFS}\${SEDdA}$1\${SEDdB}$1\${SEDdC}AC_DEFINE_SEDQUOTE(AC_VAL)\${SEDdD} \${SEDuA}$1\${SEDuB}$1\${SEDuC}AC_DEFINE_SEDQUOTE(AC_VAL)\${SEDuD} \${SEDeA}$1\${SEDeB}$1\${SEDeC}AC_DEFINE_SEDQUOTE(AC_VAL)\${SEDeD} " )dnl } ])dnl dnl dnl Unsafe version of AC_DEFINE. dnl Users are responsible for the quoting nightmare. dnl Well, not all of it. We need to pull the identify function out to dnl the top level, because m4 doesn't really support nested functions; dnl it doesn't distinguish between the arguments to the outer dnl function, which should be expanded, and the arguments to the inner dnl function, which shouldn't yet. define(AC_IDENTITY,$1)dnl define(AC_DEFINE_UNQUOTED,[dnl pushdef([AC_DEFINE_QUOTE],defn([AC_IDENTITY]))dnl pushdef([AC_DEFINE_SEDQUOTE],defn([AC_IDENTITY]))dnl AC_DEFINE($1,$2)dnl popdef([AC_DEFINE_SEDQUOTE])dnl popdef([AC_DEFINE_QUOTE])dnl ])dnl dnl define(AC_BEFORE, [ifdef([AC_PROVIDE_$2], [errprint(__file__:__line__: [$2 was called before $1 ])])])dnl dnl define(AC_REQUIRE, [ifdef([AC_PROVIDE_$1],,[indir([$1]) ])])dnl dnl define(AC_PROVIDE, [define([AC_PROVIDE_$1],)])dnl dnl define(AC_OBSOLETE, [errprint(__file__:__line__: warning: [$1] is obsolete[$2] )])dnl dnl dnl dnl Checks for kinds of features dnl dnl define(AC_PROGRAM_CHECK, [if test -z "[$]$1"; then # Extract the first word of `$2', so it can be a program name with args. set dummy $2; word=[$]2 echo checking for $word IFS="${IFS= }"; saveifs="$IFS"; IFS="${IFS}:" for dir in $PATH; do test -z "$dir" && dir=. if test -f $dir/$word; then $1="$3" break fi done IFS="$saveifs" fi ifelse([$4],,, [test -z "[$]$1" && $1="$4"]) test -n "[$]$1" && test -n "$verbose" && echo " setting $1 to [$]$1" AC_SUBST($1)dnl ])dnl dnl define(AC_PROGRAMS_CHECK, [for p in $2 do AC_PROGRAM_CHECK($1, [$]p, [$]p, ) test -n "[$]$1" && break done ifelse([$3],,, [test -n "[$]$1" || $1="$3" ])])dnl dnl define(AC_HEADER_EGREP, [AC_REQUIRE([AC_PROG_CPP])AC_PROVIDE([$0])echo '#include "confdefs.h" #include <$2>' > conftest.c eval "$CPP conftest.c > conftest.out 2>&1" if egrep "$1" conftest.out >/dev/null 2>&1; then ifelse([$3], , :, [rm -rf conftest* $3 ]) ifelse([$4], , , [else rm -rf conftest* $4 ])dnl fi rm -f conftest* ])dnl dnl dnl Because this macro is used by AC_GCC_TRADITIONAL, which must come early, dnl it is not included in AC_BEFORE checks. define(AC_PROGRAM_EGREP, [AC_REQUIRE([AC_PROG_CPP])AC_PROVIDE([$0])cat > conftest.c < conftest.out 2>&1" if egrep "$1" conftest.out >/dev/null 2>&1; then ifelse([$3], , :, [rm -rf conftest* $3 ]) ifelse([$4], , , [else rm -rf conftest* $4 ])dnl fi rm -f conftest* ])dnl dnl define(AC_HEADER_CHECK, [echo checking for $1 ifelse([$3], , [AC_TEST_CPP([#include <$1>], [$2])], [AC_TEST_CPP([#include <$1>], [$2], [$3])]) ])dnl dnl define(AC_COMPILE_CHECK, [AC_PROVIDE([$0])dnl ifelse([$1], , , [echo checking for $1] )dnl cat > conftest.c < conftest.c </dev/null; then ifelse([$2], , :, [$2 ]) ifelse([$3], , , [else $3 ])dnl fi ifelse([$4], , , fi )dnl rm -fr conftest*])dnl dnl define(AC_TEST_CPP, [AC_REQUIRE([AC_PROG_CPP])cat > conftest.c </dev/null) 2>&1"` if test -z "$err"; then ifelse([$2], , :, [rm -rf conftest* $2 ]) ifelse([$3], , , [else rm -rf conftest* $3 ])dnl fi rm -f conftest*])dnl dnl define(AC_REPLACE_FUNCS, [for func in $1 do AC_COMPILE_CHECK([${func}], [#include ], [ /* The GNU C library defines this for functions which it implements to always fail with ENOSYS. Some functions are actually named something starting with __ and the normal name is an alias. */ #if defined (__stub_${func}) || defined (__stub___${func}) choke me #else /* Override any gcc2 internal prototype to avoid an error. */ extern char ${func}(); ${func}(); #endif ], , [LIBOBJS="$LIBOBJS ${func}.o" test -n "$verbose" && echo " using ${func}.o instead"]) done AC_SUBST(LIBOBJS)dnl ])dnl dnl define(AC_FUNC_CHECK, [ifelse([$3], , [AC_COMPILE_CHECK($1, [#include ], [ /* The GNU C library defines this for functions which it implements to always fail with ENOSYS. Some functions are actually named something starting with __ and the normal name is an alias. */ #if defined (__stub_$1) || defined (__stub___$1) choke me #else /* Override any gcc2 internal prototype to avoid an error. */ extern char $1(); $1(); #endif ], $2)], [AC_COMPILE_CHECK($1, [#include ], [ /* The GNU C library defines this for functions which it implements to always fail with ENOSYS. Some functions are actually named something starting with __ and the normal name is an alias. */ #if defined (__stub_$1) || defined (__stub___$1) choke me #else /* Override any gcc2 internal prototype to avoid an error. */ extern char $1(); $1(); #endif ], $2, $3)])dnl ])dnl dnl define(AC_HAVE_FUNCS, [for func in $1 do changequote(,)dnl trfunc=HAVE_`echo $func | tr '[a-z]' '[A-Z]'` changequote([,])dnl AC_FUNC_CHECK(${func}, AC_DEFINE(${trfunc}))dnl done ])dnl dnl define(AC_HAVE_HEADERS, [for hdr in $1 do changequote(,)dnl trhdr=HAVE_`echo $hdr | tr '[a-z]./' '[A-Z]__'` changequote([,])dnl AC_HEADER_CHECK(${hdr}, AC_DEFINE(${trhdr}))dnl done ])dnl dnl define(AC_HAVE_LIBRARY, [dnl changequote(/,/)dnl define(/libname/, dnl patsubst(patsubst($1, /lib\([^\.]*\)\.a/, /\1/), /-l/, //))dnl changequote([,])dnl LIBS_save="${LIBS}" LIBS="${LIBS} -l[]libname[]" have_lib="" AC_COMPILE_CHECK([-l[]libname[]], , [main();], [have_lib="1"])dnl LIBS="${LIBS_save}" ifelse($#, 1, [dnl if test -n "${have_lib}"; then AC_DEFINE([HAVE_LIB]translit(libname, [a-z], [A-Z])) LIBS="${LIBS} -l[]libname[]" fi undefine(libname)dnl ], [dnl if test -n "${have_lib}"; then :; $2 else :; $3 fi ])])dnl dnl dnl dnl The big finish dnl dnl define(AC_OUTPUT, [changequote(,)dnl # Set default prefixes. if test -n "$prefix"; then test -z "$exec_prefix" && exec_prefix='${prefix}' prsub="s%^prefix\\([ ]*\\)=\\([ ]*\\).*$%prefix\\1=\\2$prefix%" fi if test -n "$exec_prefix"; then prsub="$prsub s%^exec_prefix\\([ ]*\\)=\\([ ]*\\).*$%exec_prefix\\1=\\2$exec_prefix%" fi # Quote sed substitution magic chars in DEFS. cat >conftest.def < config.status </dev/null | sed 1q`: # [#] [$]0 [$]configure_args for arg do case "[\$]arg" in -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r) echo running [\$]{CONFIG_SHELL-/bin/sh} [$]0 [$]configure_args exec [\$]{CONFIG_SHELL-/bin/sh} [$]0 [$]configure_args ;; *) echo "Usage: config.status [--recheck]" 2>&1; exit 1 ;; esac done ifdef([AC_CONFIG_NAMES], [trap 'rm -fr $1 AC_CONFIG_NAMES conftest*; exit 1' 1 3 15], [trap 'rm -f $1; exit 1' 1 3 15]) dnl Insert the variable assignments. undivert(2)dnl EOF cat >> config.status <<\EOF top_srcdir=$srcdir CONFIG_FILES=${CONFIG_FILES-"$1"} for file in .. ${CONFIG_FILES}; do if test "x$file" != x..; then srcdir=$top_srcdir # Remove last slash and all that follows it. Not all systems have dirname. changequote(,)dnl dir=`echo $file|sed 's%/[^/][^/]*$%%'` changequote([,])dnl if test "$dir" != "$file"; then test "$top_srcdir" != . && srcdir=$top_srcdir/$dir test ! -d $dir && mkdir $dir fi echo creating $file rm -f $file echo "# Generated automatically from `echo $file|sed 's|.*/||'`.in by configure." > $file sed -e " $prsub dnl Shell code in configure.in might set extrasub. $extrasub dnl Insert the sed substitutions. undivert(1)dnl " $top_srcdir/${file}.in >> $file fi; done AC_OUTPUT_HEADER $2 exit 0 EOF chmod +x config.status ${CONFIG_SHELL-/bin/sh} config.status ])dnl dnl This is a subroutine of AC_OUTPUT, broken out primarily to avoid bugs dnl with long definitions in GNU m4 1.0. This is called inside a quoted dnl here document whose contents are going into config.status. define(AC_OUTPUT_HEADER,[dnl ifdef([AC_CONFIG_NAMES],[dnl changequote(<<,>>)dnl CONFIG_HEADERS=${CONFIG_HEADERS-"AC_CONFIG_NAMES"} for file in .. ${CONFIG_HEADERS}; do if test "x$file" != x..; then echo creating $file # These sed commands are put into SEDDEFS when defining a macro. # They are broken into pieces to make the sed script easier to manage. # They are passed to sed as "A NAME B NAME C VALUE D", where NAME # is the cpp macro being defined and VALUE is the value it is being given. # Each defining turns into a single global substitution command. # # SEDd sets the value in "#define NAME VALUE" lines. SEDdA='s@^\([ ]*\)#\([ ]*define[ ][ ]*\)' SEDdB='\([ ][ ]*\)[^ ]*@\1#\2' SEDdC='\3' SEDdD='@g' # SEDu turns "#undef NAME" with trailing blanks into "#define NAME VALUE". SEDuA='s@^\([ ]*\)#\([ ]*\)undef\([ ][ ]*\)' SEDuB='\([ ]\)@\1#\2define\3' SEDuC=' ' SEDuD='\4@g' # SEDe turns "#undef NAME" without trailing blanks into "#define NAME VALUE". SEDeA='s@^\([ ]*\)#\([ ]*\)undef\([ ][ ]*\)' SEDeB='<<$>>@\1#\2define\3' SEDeC=' ' SEDeD='@g' changequote([,])dnl rm -f conftest.sed EOF # Turn off quoting long enough to insert the sed commands. rm -f conftest.sh cat > conftest.sh < conftest.s1 # Like head -20. sed 1,${maxshlines}d conftest.sh > conftest.s2 # Like tail +21. # Write a limited-size here document to append to conftest.sed. echo 'cat >> conftest.sed <> config.status cat conftest.s1 >> config.status echo 'CONFEOF' >> config.status rm -f conftest.s1 conftest.sh mv conftest.s2 conftest.sh done rm -f conftest.sh # Now back to your regularly scheduled config.status. cat >> config.status <<\EOF # This sed command replaces #undef's with comments. This is necessary, for # example, in the case of _POSIX_SOURCE, which is predefined and required # on some systems where configure will not decide to define it in [#] AC_CONFIG_NAMES. cat >> conftest.sed <<\CONFEOF changequote(,)dnl s,^[ ]*#[ ]*undef[ ][ ]*[a-zA-Z_][a-zA-Z_0-9]*,/* & */, changequote([,])dnl CONFEOF rm -f conftest.h # Break up the sed commands because old seds have small limits. maxsedlines=20 cp $top_srcdir/$file.in conftest.h1 while : do lines=`grep -c . conftest.sed` if test -z "$lines" || test "$lines" -eq 0; then break; fi rm -f conftest.s1 conftest.s2 conftest.h2 sed ${maxsedlines}q conftest.sed > conftest.s1 # Like head -20. sed 1,${maxsedlines}d conftest.sed > conftest.s2 # Like tail +21. sed -f conftest.s1 < conftest.h1 > conftest.h2 rm -f conftest.s1 conftest.h1 conftest.sed mv conftest.h2 conftest.h1 mv conftest.s2 conftest.sed done rm -f conftest.sed conftest.h echo "/* $file. Generated automatically by configure. */" > conftest.h cat conftest.h1 >> conftest.h rm -f conftest.h1 if cmp -s $file conftest.h 2>/dev/null; then # The file exists and we would not be changing it. echo "$file is unchanged" rm -f conftest.h else rm -f $file mv conftest.h $file fi fi; done ])])dnl autoconf-1.7/acspecific.m4100666 21570 13 77335 5461347257 13616 0ustar djmuserdnl Macros that test for specific features. dnl This file is part of Autoconf. dnl Copyright (C) 1992, 1993 Free Software Foundation, Inc. dnl dnl This program is free software; you can redistribute it and/or modify dnl it under the terms of the GNU General Public License as published by dnl the Free Software Foundation; either version 2, or (at your option) dnl any later version. dnl dnl This program is distributed in the hope that it will be useful, dnl but WITHOUT ANY WARRANTY; without even the implied warranty of dnl MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the dnl GNU General Public License for more details. dnl dnl You should have received a copy of the GNU General Public License dnl along with this program; if not, write to the Free Software dnl Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. dnl dnl Written by David MacKenzie, with help from dnl Franc,ois Pinard, Karl Berry, Richard Pixley, Ian Lance Taylor, dnl Roland McGrath, and Noah Friedman. dnl dnl dnl checks for programs dnl dnl define(AC_PROG_CC, [AC_BEFORE([$0], [AC_PROG_CPP])AC_PROVIDE([$0])AC_PROGRAM_CHECK(CC, gcc, gcc, cc) # Find out if we are using GNU C, under whatever name. cat > conftest.c < conftest.out 2>&1 if egrep yes conftest.out >/dev/null 2>&1; then GCC=1 # For later tests. fi rm -f conftest* ])dnl dnl define(AC_GCC_TRADITIONAL, [AC_REQUIRE([AC_PROG_CC])AC_REQUIRE([AC_PROG_CPP])if test -n "$GCC"; then echo checking whether -traditional is needed changequote(,)dnl pattern="Autoconf.*'x'" changequote([,])dnl prog='#include Autoconf TIOCGETP' AC_PROGRAM_EGREP($pattern, $prog, need_trad=1) if test -z "$need_trad"; then prog='#include Autoconf TCGETA' AC_PROGRAM_EGREP($pattern, $prog, need_trad=1) fi test -n "$need_trad" && CC="$CC -traditional" fi ])dnl dnl define(AC_MINUS_C_MINUS_O, [echo checking whether $CC and cc understand -c and -o together echo 'foo(){}' > conftest.c # Make sure it works both with $CC and with simple cc. # We do the test twice because some compilers refuse to overwrite an # existing .o file with -o, though they will create one. if ${CC-cc} -c conftest.c -o conftest.o >/dev/null 2>&1 \ && test -f conftest.o && ${CC-cc} -c conftest.c -o conftest.o >/dev/null 2>&1 then # Test first that cc exists at all. if cc -c conftest.c >/dev/null 2>&1 then if cc -c conftest.c -o conftest2.o >/dev/null 2>&1 && \ test -f conftest2.o && cc -c conftest.c -o conftest2.o >/dev/null 2>&1 then : else AC_DEFINE(NO_MINUS_C_MINUS_O) fi fi else AC_DEFINE(NO_MINUS_C_MINUS_O) fi rm -f conftest* ])dnl dnl define(AC_PROG_RANLIB, [AC_PROGRAM_CHECK(RANLIB, ranlib, ranlib, :)])dnl dnl define(AC_PROG_AWK, [AC_PROGRAMS_CHECK(AWK, mawk gawk nawk awk,)])dnl dnl define(AC_PROG_YACC,[AC_PROGRAMS_CHECK(YACC, 'bison -y' byacc, yacc)])dnl dnl define(AC_PROG_CPP, [AC_PROVIDE([$0])echo checking how to run the C preprocessor if test -z "$CPP"; then # This must be in double quotes, not single quotes, because CPP may get # substituted into the Makefile and ``${CC-cc}'' will simply confuse # make. It must be expanded now. CPP="${CC-cc} -E" dnl On the NeXT, cc -E appears to run the code through the compiler's parser, dnl not just through cpp. AC_TEST_CPP([#include Syntax Error], , CPP=/lib/cpp) fi test ".${verbose}" != "." && echo " setting CPP to $CPP" AC_SUBST(CPP)dnl ])dnl dnl define(AC_PROG_LEX, [AC_PROVIDE([$0])AC_PROGRAM_CHECK(LEX, flex, flex, lex) if test -z "$LEXLIB" then case "$LEX" in flex*) AC_HAVE_LIBRARY(fl, LEXLIB="-lfl") ;; *) LEXLIB="-ll" ;; esac fi AC_SUBST(LEXLIB)])dnl dnl define(AC_DECLARE_YYTEXT, [AC_REQUIRE([AC_PROG_CPP])AC_REQUIRE([AC_PROG_LEX])dnl echo checking how to declare yytext # Figure out what yytext is by creating a minimal parser and # examining the (preprocessed, in case macros are used) output. if test "z${DECLARE_YYTEXT}" = "z"; then changequote(,)dnl echo '%% %%' | ${LEX} if test -f lex.yy.c; then LEX_OUTPUT_ROOT=lex.yy elif test -f lexyy.c; then LEX_OUTPUT_ROOT=lexyy else # Don't know what to do here. echo "Can't find output from $LEX; assuming lex.yy.c." 1>&2 LEX_OUTPUT_ROOT=lex.yy fi AC_SUBST(LEX_OUTPUT_ROOT)dnl DECLARE_YYTEXT=`eval ${CPP} "${LEX_OUTPUT_ROOT}.c" | sed -n '/extern.*yytext[^a-zA-Z0-9_]/s/^.*extern/extern/p'` rm -f "${LEX_OUTPUT_ROOT}.c" changequote([,])dnl fi AC_DEFINE_UNQUOTED(DECLARE_YYTEXT, \"$DECLARE_YYTEXT\")dnl ])dnl dnl define(AC_PROG_INSTALL, [# Make sure to not get the incompatible SysV /etc/install and # /usr/sbin/install, which might be in PATH before a BSD-like install, # or the SunOS /usr/etc/install directory, or the AIX /bin/install, # or the AFS install, which mishandles nonexistent args, or # /usr/ucb/install on SVR4, which tries to use the nonexistent group # `staff'. On most BSDish systems install is in /usr/bin, not /usr/ucb # anyway. Sigh. if test "z${INSTALL}" = "z" ; then echo checking for install IFS="${IFS= }"; saveifs="$IFS"; IFS="${IFS}:" for dir in $PATH; do test -z "$dir" && dir=. case $dir in /etc|/usr/sbin|/usr/etc|/usr/afsws/bin|/usr/ucb) ;; *) if test -f $dir/installbsd; then INSTALL="$dir/installbsd -c" # OSF1 INSTALL_PROGRAM='$(INSTALL)' INSTALL_DATA='$(INSTALL) -m 644' break fi if test -f $dir/install; then if grep dspmsg $dir/install >/dev/null 2>&1; then : # AIX else INSTALL="$dir/install -c" INSTALL_PROGRAM='$(INSTALL)' INSTALL_DATA='$(INSTALL) -m 644' break fi fi ;; esac done IFS="$saveifs" fi INSTALL=${INSTALL-cp} AC_SUBST(INSTALL)dnl test -n "$verbose" && echo " setting INSTALL to $INSTALL" INSTALL_PROGRAM=${INSTALL_PROGRAM-'$(INSTALL)'} AC_SUBST(INSTALL_PROGRAM)dnl test -n "$verbose" && echo " setting INSTALL_PROGRAM to $INSTALL_PROGRAM" INSTALL_DATA=${INSTALL_DATA-'$(INSTALL)'} AC_SUBST(INSTALL_DATA)dnl test -n "$verbose" && echo " setting INSTALL_DATA to $INSTALL_DATA" ])dnl dnl define(AC_LN_S, [echo checking for ln -s rm -f conftestdata if ln -s X conftestdata 2>/dev/null then rm -f conftestdata LN_S="ln -s" else LN_S=ln fi AC_SUBST(LN_S) ])dnl dnl define(AC_RSH, [echo checking for remote shell if test -f /usr/ucb/rsh || test -f /usr/bin/remsh || test -f /usr/bin/rsh || test -f /usr/bsd/rsh || test -f /usr/bin/nsh; then RTAPELIB=rtapelib.o else AC_HEADER_CHECK(netdb.h, AC_DEFINE(HAVE_NETDB_H) RTAPELIB=rtapelib.o, AC_DEFINE(NO_REMOTE)) fi AC_SUBST(RTAPELIB)dnl ])dnl dnl dnl dnl checks for header files dnl dnl define(AC_STDC_HEADERS, [AC_REQUIRE([AC_PROG_CPP])dnl echo checking for ANSI C header files AC_TEST_CPP([#include #include #include #include ], [# SunOS 4.x string.h does not declare mem*, contrary to ANSI. AC_HEADER_EGREP(memchr, string.h, # SGI's /bin/cc from Irix-4.0.5 gets non-ANSI ctype macros unless using -ansi. AC_TEST_PROGRAM([#include #define ISLOWER(c) ('a' <= (c) && (c) <= 'z') #define TOUPPER(c) (ISLOWER(c) ? 'A' + ((c) - 'a') : (c)) #define XOR(e,f) (((e) && !(f)) || (!(e) && (f))) int main () { int i; for (i = 0; i < 256; i++) if (XOR (islower (i), ISLOWER (i)) || toupper (i) != TOUPPER (i)) exit(2); exit (0); } ],AC_DEFINE(STDC_HEADERS)))]) ])dnl dnl define(AC_UNISTD_H, [AC_OBSOLETE([$0], [; instead use AC_HAVE_HEADERS(unistd.h)])AC_HEADER_CHECK(unistd.h, AC_DEFINE(HAVE_UNISTD_H))])dnl dnl define(AC_USG, [AC_OBSOLETE([$0], [; instead use AC_HAVE_HEADERS(string.h) and HAVE_STRING_H])AC_COMPILE_CHECK([BSD string and memory functions], [#include ], [rindex(0, 0); bzero(0, 0);], , AC_DEFINE(USG))])dnl dnl dnl dnl If memchr and the like aren't declared in , include . dnl To avoid problems, don't check for gcc2 built-ins. define(AC_MEMORY_H, [AC_OBSOLETE([$0], [; instead use AC_HAVE_HEADERS(memory.h) and HAVE_MEMORY_H])echo checking whether string.h declares mem functions AC_HEADER_EGREP(memchr, string.h, , AC_HEADER_CHECK(memory.h, AC_DEFINE(NEED_MEMORY_H)))] )dnl dnl define(AC_MAJOR_HEADER, [AC_COMPILE_CHECK([major, minor and makedev header], [#include ], [return makedev(0, 0);], makedev=1) if test -z "$makedev"; then AC_HEADER_CHECK(sys/mkdev.h, AC_DEFINE(MAJOR_IN_MKDEV) makedev=1) fi if test -z "$makedev"; then AC_HEADER_CHECK(sys/sysmacros.h, AC_DEFINE(MAJOR_IN_SYSMACROS)) fi] )dnl dnl define(AC_DIR_HEADER, [AC_PROVIDE([$0])echo checking for directory library header dirheader= AC_DIR_HEADER_CHECK(dirent.h, DIRENT) AC_DIR_HEADER_CHECK(sys/ndir.h, SYSNDIR) AC_DIR_HEADER_CHECK(sys/dir.h, SYSDIR) AC_DIR_HEADER_CHECK(ndir.h, NDIR) echo checking for closedir return value AC_TEST_PROGRAM([#include #include <$dirheader> int closedir(); main() { exit(closedir(opendir(".")) != 0); }], , AC_DEFINE(VOID_CLOSEDIR)) ])dnl dnl Subroutine of AC_DIR_HEADER. dnl ??? I tried to put this define inside AC_DIR_HEADER, but when I did dnl that, $1 and $2 did not get expanded. --roland dnl Check if $1 is the winning directory library header file. dnl It must not only exist, but also correctly define the `DIR' type. dnl If it is really winning, define $2 and set shell var `dirheader' to $1. define(AC_DIR_HEADER_CHECK, [dnl if test -z "$dirheader"; then AC_COMPILE_CHECK($1, [#include #include <]$1[>], [DIR *dirp = 0;], AC_DEFINE($2) dirheader=$1)dnl fi])dnl dnl dnl dnl checks for typedefs dnl dnl define(AC_GETGROUPS_T, [AC_REQUIRE([AC_UID_T])dnl echo checking for type of array argument to getgroups changequote(,)dnl dnl Do not put single quotes in the C program text!! prog='/* Thanks to Mike Rendell for this test. */ #include #define NGID 256 #undef MAX #define MAX(x,y) ((x) > (y) ? (x) : (y)) main() { gid_t gidset[NGID]; int i, n; union { gid_t gval; long lval; } val; val.lval = -1; for (i = 0; i < NGID; i++) gidset[i] = val.gval; n = getgroups (sizeof (gidset) / MAX (sizeof (int), sizeof (gid_t)) - 1, gidset); /* Exit non-zero if getgroups seems to require an array of ints. This happens when gid_t is short but getgroups modifies an array of ints. */ exit ((n > 0 && gidset[n] != val.gval) ? 1 : 0); }' changequote([,])dnl AC_TEST_PROGRAM([$prog], AC_DEFINE(GETGROUPS_T, gid_t), AC_DEFINE(GETGROUPS_T, int)) ])dnl dnl define(AC_UID_T, [AC_PROVIDE([$0])echo checking for uid_t in sys/types.h AC_HEADER_EGREP(uid_t, sys/types.h, , AC_DEFINE(uid_t, int) AC_DEFINE(gid_t, int))])dnl dnl define(AC_SIZE_T, [echo checking for size_t in sys/types.h AC_HEADER_EGREP(size_t, sys/types.h, , AC_DEFINE(size_t, unsigned))])dnl dnl define(AC_PID_T, [AC_PROVIDE([$0])echo checking for pid_t in sys/types.h AC_HEADER_EGREP(pid_t, sys/types.h, , AC_DEFINE(pid_t, int))])dnl dnl define(AC_OFF_T, [AC_PROVIDE([$0])echo checking for off_t in sys/types.h AC_HEADER_EGREP(off_t, sys/types.h, , AC_DEFINE(off_t, long))])dnl dnl define(AC_MODE_T, [echo checking for mode_t in sys/types.h AC_HEADER_EGREP(mode_t, sys/types.h, , AC_DEFINE(mode_t, int))])dnl dnl define(AC_RETSIGTYPE, [AC_PROVIDE([$0])AC_COMPILE_CHECK([return type of signal handlers], [#include #include #ifdef signal #undef signal #endif extern void (*signal ()) ();], [int i;], [AC_DEFINE(RETSIGTYPE, void)], [AC_DEFINE(RETSIGTYPE, int)])] )dnl dnl dnl dnl checks for functions dnl dnl define(AC_VPRINTF, [AC_COMPILE_CHECK([vprintf], , [vprintf();], AC_DEFINE(HAVE_VPRINTF), vprintf_missing=1) if test -n "$vprintf_missing"; then AC_COMPILE_CHECK([_doprnt], , [_doprnt();], AC_DEFINE(HAVE_DOPRNT)) fi ])dnl dnl define(AC_VFORK, [AC_REQUIRE([AC_PID_T])AC_HEADER_CHECK(vfork.h, AC_DEFINE(HAVE_VFORK_H)) echo checking for working vfork AC_REQUIRE([AC_RETSIGTYPE]) AC_TEST_PROGRAM([/* Thanks to Paul Eggert for this test. */ #include #include #include #include #ifdef HAVE_UNISTD_H #include #endif #ifdef HAVE_VFORK_H #include #endif static int signalled; static RETSIGTYPE catch (s) int s; { signalled = 1; } main() { pid_t parent = getpid (); pid_t child; signal (SIGINT, catch); child = vfork (); if (child == 0) { /* On sparc systems, changes by the child to local and incoming argument registers are propagated back to the parent. The compiler is told about this with #include , but some compilers (e.g. gcc -O) don't grok . Test for this by using lots of local variables, at least as many local variables as main has allocated so far including compiler temporaries. 4 locals are enough for gcc 1.40.3 on a sparc, but we use 8 to be safe. A buggy compiler should reuse the register of parent for one of the local variables, since it will think that parent can't possibly be used any more in this routine. Assigning to the local variable will thus munge parent in the parent process. */ pid_t p = getpid(), p1 = getpid(), p2 = getpid(), p3 = getpid(), p4 = getpid(), p5 = getpid(), p6 = getpid(), p7 = getpid(); /* Convince the compiler that p..p7 are live; otherwise, it might use the same hardware register for all 8 local variables. */ if (p != p1 || p != p2 || p != p3 || p != p4 || p != p5 || p != p6 || p != p7) _exit(1); /* On some systems (e.g. SunOS 5.2), if the parent is catching a signal, the child ignores the signal before execing, and the parent later receives that signal, the parent dumps core. Test for this by ignoring SIGINT in the child. */ signal (SIGINT, SIG_IGN); /* On some systems (e.g. IRIX 3.3), vfork doesn't separate parent from child file descriptors. If the child closes a descriptor before it execs or exits, this munges the parent's descriptor as well. Test for this by closing stdout in the child. */ _exit(close(fileno(stdout)) != 0); } else { int status; struct stat st; while (wait(&status) != child) ; exit( /* Was there some problem with vforking? */ child < 0 /* Did the child fail? (This shouldn't happen.) */ || status /* Did the vfork/compiler bug occur? */ || parent != getpid() /* Did the signal handling bug occur? */ || kill(parent, SIGINT) != 0 || signalled != 1 /* Did the file descriptor bug occur? */ || fstat(fileno(stdout), &st) != 0 ); } }], , AC_DEFINE(vfork, fork)) ])dnl dnl define(AC_WAIT3, [echo checking for wait3 that fills in rusage AC_TEST_PROGRAM([#include #include #include #include /* HP-UX has wait3 but does not fill in rusage at all. */ main() { struct rusage r; int i; /* Use a field that we can force nonzero -- voluntary context switches. For systems like NeXT and OSF/1 that don't set it, also use the system CPU time. */ r.ru_nvcsw = 0; r.ru_stime.tv_sec = 0; r.ru_stime.tv_usec = 0; switch (fork()) { case 0: /* Child. */ sleep(1); /* Give up the CPU. */ _exit(0); case -1: _exit(0); /* What can we do? */ default: /* Parent. */ wait3(&i, 0, &r); exit(r.ru_nvcsw == 0 && r.ru_stime.tv_sec == 0 && r.ru_stime.tv_usec == 0); } }], AC_DEFINE(HAVE_WAIT3)) ])dnl dnl define(AC_ALLOCA, [# The Ultrix 4.2 mips builtin alloca declared by alloca.h only works # for constant arguments. Useless! AC_COMPILE_CHECK(working alloca.h, [#include ], [char *p = alloca(2 * sizeof(int));], AC_DEFINE(HAVE_ALLOCA_H)) decl="#ifdef __GNUC__ #define alloca __builtin_alloca #else #if HAVE_ALLOCA_H #include #else #ifdef _AIX #pragma alloca #else char *alloca (); #endif #endif #endif " AC_COMPILE_CHECK([alloca], $decl, [char *p = (char *) alloca(1);], , [dnl alloca_missing=1 AC_PROGRAM_EGREP(winnitude, [ #if defined(CRAY) && ! defined(CRAY2) winnitude #else lossage #endif ], AC_FUNC_CHECK([_getb67],AC_DEFINE([CRAY_STACKSEG_END],[_getb67]), AC_FUNC_CHECK([GETB67],AC_DEFINE([CRAY_STACKSEG_END],[GETB67]), AC_FUNC_CHECK([getb67],AC_DEFINE([CRAY_STACKSEG_END],[getb67]))))) ]) if test -n "$alloca_missing"; then # The SVR3 libPW and SVR4 libucb both contain incompatible functions # that cause trouble. Some versions do not even contain alloca or # contain a buggy version. If you still want to use their alloca, # use ar to extract alloca.o from them instead of compiling alloca.c. ALLOCA=alloca.o AC_DEFINE(C_ALLOCA) echo 'checking stack direction for C alloca' AC_TEST_PROGRAM([find_stack_direction () { static char *addr = 0; auto char dummy; if (addr == 0) { addr = &dummy; return find_stack_direction (); } else return (&dummy > addr) ? 1 : -1; } main () { exit (find_stack_direction() < 0); }], dnl AC_DEFINE(STACK_DIRECTION,1), AC_DEFINE(STACK_DIRECTION,-1), dnl AC_DEFINE(STACK_DIRECTION,0)) fi AC_SUBST(ALLOCA)dnl ])dnl dnl define(AC_GETLOADAVG, [# Some definitions of getloadavg require that the program be installed setgid. AC_SUBST(NEED_SETGID)NEED_SETGID=false need_func=true # Check for the 4.4BSD definition of getloadavg. AC_HAVE_LIBRARY(util, AC_HAVE_LIBRARY(kvm, LIBS="$LIBS -lutil -lkvm" need_func=false)) if $need_func; then # There is a commonly available library for RS/6000 AIX. # Since it is not a standard part of AIX, it might be installed locally. LIBS_old="$LIBS" LIBS="-L/usr/local/lib $LIBS" AC_HAVE_LIBRARY(getloadavg, LIBS="$LIBS -lgetloadavg" need_func=false, LIBS="$LIBS_old") fi # Make sure it is really in the library, if we think we found it at all. AC_REPLACE_FUNCS(getloadavg) case "$LIBOBJS" in *getloadavg*) need_func=true AC_HEADER_CHECK(sys/dg_sys_info.h, [dnl AC_DEFINE(DGUX) need_func=false # Some versions of DGUX need -ldgc for dg_sys_info. AC_HAVE_LIBRARY(dgc)]) if $need_func; then # We cannot check for , because Solaris 2 does not use dwarf (it # uses stabs), but it's still SVR4. We cannot check for because # Irix 4.0.5F has the header but not the library. AC_HAVE_LIBRARY(elf, AC_DEFINE(SVR4) LIBS="$LIBS -lelf" need_func=false AC_HAVE_LIBRARY(kvm, LIBS="$LIBS -lkvm")) fi if $need_func; then AC_HEADER_CHECK(inq_stats/cpustats.h, AC_DEFINE(UMAX4_3) AC_DEFINE(UMAX) need_func=false) fi if $need_func; then AC_HEADER_CHECK(sys/cpustats.h, AC_DEFINE(UMAX) need_func=false) fi if $need_func; then AC_HAVE_HEADERS(mach/mach.h) fi AC_HEADER_CHECK(nlist.h, [AC_DEFINE(NLIST_STRUCT) AC_COMPILE_CHECK(n_un in struct nlist, [#include ], [struct nlist n; n.n_un.n_name = 0;], AC_DEFINE(NLIST_NAME_UNION))])dnl # Figure out whether we will need to install setgid. AC_PROGRAM_EGREP([Yowza Am I SETGID yet], [dnl #include "${srcdir}/getloadavg.c" #ifdef LDAV_PRIVILEGED Yowza Am I SETGID yet #endif], [AC_DEFINE(GETLOADAVG_PRIVILEGED) NEED_SETGID=true])dnl ;; *) AC_DEFINE(HAVE_GETLOADAVG) ;; esac if $NEED_SETGID; then AC_SUBST(KMEM_GROUP)# Figure out what group owns /dev/kmem. # The installed program will need to be setgid and owned by that group. changequote(,)dnl # On Solaris, /dev/kmem is a symlink. Get info on the real file. ls_output=`ls -lgL /dev/kmem 2>/dev/null` # If we got an error (system does not support symlinks), try without -L. test -z "$ls_output" && ls_output=`ls -lg /dev/kmem` KMEM_GROUP=`echo $ls_output \ | sed -ne 's/[ ][ ]*/ /g; s/^.[sSrwx-]* *[0-9]* *\([^0-9]*\) *.*/\1/; / /s/.* //;p;'` changequote([,])dnl fi ])dnl dnl define(AC_UTIME_NULL, [echo checking utime with null argument rm -f conftestdata; > conftestdata # Sequent interprets utime(file, 0) to mean use start of epoch. Wrong. AC_TEST_PROGRAM([#include #include main() { struct stat s, t; exit(!(stat ("conftestdata", &s) == 0 && utime("conftestdata", (long *)0) == 0 && stat("conftestdata", &t) == 0 && t.st_mtime >= s.st_mtime && t.st_mtime - s.st_mtime < 120)); }], AC_DEFINE(HAVE_UTIME_NULL)) rm -f core ])dnl dnl define(AC_STRCOLL, [echo checking for strcoll AC_TEST_PROGRAM([#include main () { exit (strcoll ("abc", "def") >= 0 || strcoll ("ABC", "DEF") >= 0 || strcoll ("123", "456") >= 0); }], AC_DEFINE(HAVE_STRCOLL))])dnl dnl define(AC_SETVBUF_REVERSED, [AC_TEST_PROGRAM([#include /* If setvbuf has the reversed format, exit 0. */ main () { /* This call has the arguments reversed. A reversed system may check and see that the address of main is not _IOLBF, _IONBF, or _IOFBF, and return nonzero. */ if (setvbuf(stdout, _IOLBF, (char *) main, BUFSIZ) != 0) exit(1); putc('\r', stdout); exit(0); /* Non-reversed systems segv here. */ }], AC_DEFINE(SETVBUF_REVERSED)) rm -f core ])dnl dnl dnl dnl checks for structure members dnl dnl define(AC_STRUCT_TM, [AC_PROVIDE([$0])AC_COMPILE_CHECK([struct tm in time.h], [#include #include ], [struct tm *tp; tp->tm_sec;], , AC_DEFINE(TM_IN_SYS_TIME))])dnl dnl define(AC_TIME_WITH_SYS_TIME, [AC_COMPILE_CHECK([whether time.h and sys/time.h may both be included], [#include #include #include ], [struct tm *tp;], AC_DEFINE(TIME_WITH_SYS_TIME))])dnl dnl define(AC_TIMEZONE, [AC_REQUIRE([AC_STRUCT_TM])decl='#include ' case "$DEFS" in *TM_IN_SYS_TIME*) decl="$decl #include " ;; *) decl="$decl #include " ;; esac AC_COMPILE_CHECK([tm_zone in struct tm], $decl, [struct tm tm; tm.tm_zone;], AC_DEFINE(HAVE_TM_ZONE), no_tm_zone=1) if test -n "$no_tm_zone"; then AC_COMPILE_CHECK(tzname, changequote(<<,>>)dnl <<#include #ifndef tzname /* For SGI. */ extern char *tzname[]; /* RS6000 and others want it this way. */ #endif>>, changequote([,])dnl [atoi(*tzname);], AC_DEFINE(HAVE_TZNAME)) fi ])dnl dnl define(AC_ST_BLOCKS, [AC_COMPILE_CHECK([st_blocks in struct stat], [#include #include ], [struct stat s; s.st_blocks;], AC_DEFINE(HAVE_ST_BLOCKS), LIBOBJS="$LIBOBJS fileblocks.o")dnl AC_SUBST(LIBOBJS)dnl ])dnl dnl define(AC_ST_BLKSIZE, [AC_COMPILE_CHECK([st_blksize in struct stat], [#include #include ], [struct stat s; s.st_blksize;], AC_DEFINE(HAVE_ST_BLKSIZE))])dnl dnl define(AC_ST_RDEV, [AC_COMPILE_CHECK([st_rdev in struct stat], [#include #include ], [struct stat s; s.st_rdev;], AC_DEFINE(HAVE_ST_RDEV))])dnl dnl dnl dnl checks for compiler characteristics dnl dnl define(AC_CROSS_CHECK, [AC_PROVIDE([$0])echo checking whether cross-compiling # If we cannot run a trivial program, we must be cross compiling. AC_TEST_PROGRAM([main(){exit(0);}], , cross_compiling=1) ])dnl dnl define(AC_CHAR_UNSIGNED, [echo checking for unsigned characters AC_TEST_PROGRAM( [/* volatile prevents gcc2 from optimizing the test away on sparcs. */ #if !__STDC__ #define volatile #endif main() { #ifdef __CHAR_UNSIGNED__ exit(1); /* No need to redefine it. */ #else volatile char c = 255; exit(c < 0); #endif }], AC_DEFINE(__CHAR_UNSIGNED__))] )dnl dnl define(AC_LONG_DOUBLE, [AC_REQUIRE([AC_PROG_CC])dnl echo checking for long double if test -n "$GCC"; then AC_DEFINE(HAVE_LONG_DOUBLE) else AC_TEST_PROGRAM([int main() { /* On Ultrix 4.3 cc, long double is 4 and double is 8. */ exit(sizeof(long double) < sizeof(double)); }], AC_DEFINE(HAVE_LONG_DOUBLE)) fi ])dnl dnl define(AC_INT_16_BITS, [echo checking integer size AC_TEST_PROGRAM([main() { exit(sizeof(int) != 2); }], AC_DEFINE(INT_16_BITS)) ])dnl dnl define(AC_LONG_64_BITS, [echo checking for 64-bit long ints AC_TEST_PROGRAM([main() { exit(sizeof(long int) != 8); }], AC_DEFINE(LONG_64_BITS)) ])dnl dnl define(AC_WORDS_BIGENDIAN, [echo checking byte ordering AC_TEST_PROGRAM([main () { /* Are we little or big endian? From Harbison&Steele. */ union { long l; char c[sizeof (long)]; } u; u.l = 1; exit (u.c[sizeof (long) - 1] == 1); }], , AC_DEFINE(WORDS_BIGENDIAN)) ])dnl dnl define(AC_ARG_ARRAY, [echo checking whether the address of an argument can be used as an array AC_TEST_PROGRAM([main() { /* Return 0 iff arg arrays are ok. */ exit(!x(1, 2, 3, 4)); } x(a, b, c, d) { return y(a, &b); } /* Return 1 iff arg arrays are ok. */ y(a, b) int *b; { return a == 1 && b[0] == 2 && b[1] == 3 && b[2] == 4; }], , AC_DEFINE(NO_ARG_ARRAY)) rm -f core ])dnl dnl define(AC_INLINE, [AC_REQUIRE([AC_PROG_CC])if test -n "$GCC"; then AC_COMPILE_CHECK([inline], , [} inline foo() {], , AC_DEFINE(inline, __inline)) fi ])dnl define(AC_CONST, [changequote(,)dnl dnl Do not put single quotes in the C program text!! prog='/* Ultrix mips cc rejects this. */ typedef int charset[2]; const charset x; /* SunOS 4.1.1 cc rejects this. */ char const *const *ccp; char **p; /* AIX XL C 1.02.0.0 rejects this. It does not let you subtract one const X* pointer from another in an arm of an if-expression whose if-part is not a constant expression */ const char *g = "string"; ccp = &g + (g ? g-g : 0); /* HPUX 7.0 cc rejects these. */ ++ccp; p = (char**) ccp; ccp = (char const *const *) p; { /* SCO 3.2v4 cc rejects this. */ char *t; char const *s = 0 ? (char *) 0 : (char const *) 0; *t++ = 0; } { /* Someone thinks the Sun supposedly-ANSI compiler will reject this. */ int x[] = {25,17}; const int *foo = &x[0]; ++foo; } { /* Sun SC1.0 ANSI compiler rejects this -- but not the above. */ typedef const int *iptr; iptr p = 0; ++p; } { /* AIX XL C 1.02.0.0 rejects this saying "k.c", line 2.27: 1506-025 (S) Operand must be a modifiable lvalue. */ struct s { int j; const int *ap[3]; }; struct s *b; b->j = 5; } { /* ULTRIX-32 V3.1 (Rev 9) vcc rejects this */ const int foo = 10; }' changequote([,])dnl AC_COMPILE_CHECK([dnl Do not "break" this again. lack of working const], , [$prog], , AC_DEFINE(const,))])dnl dnl dnl dnl checks for operating system services dnl dnl define(AC_HAVE_POUNDBANG, [dnl echo "checking if \`#!' works in shell scripts" echo '#!/bin/cat exit 69 ' > conftest chmod u+x conftest (SHELL=/bin/sh; export SHELL; ./conftest > /dev/null) if test $? -ne 69; then :; $1 else :; $2 fi rm -f conftest ])dnl define(AC_REMOTE_TAPE, [echo checking for remote tape and socket header files AC_HEADER_CHECK(sys/mtio.h, AC_DEFINE(HAVE_SYS_MTIO_H) have_mtio=1) if test -n "$have_mtio"; then AC_TEST_CPP([#include #include ], PROGS="$PROGS rmt") fi ])dnl dnl define(AC_LONG_FILE_NAMES, [echo checking for long file names some_dir_failed=false # Test for long file names in all the places we know might matter: # . the current directory, where building will happen # /tmp where it might want to write temporary files # /var/tmp likewise # /usr/tmp likewise # $prefix/lib where we will be installing things # $exec_prefix/lib likewise # eval it to expand exec_prefix. for dir in `eval echo . /tmp /var/tmp /usr/tmp $prefix/lib $exec_prefix/lib` ; do test -d $dir || continue test -w $dir || continue # It's less confusing to not echo anything here. (echo 1 > $dir/conftest9012345) 2>/dev/null (echo 2 > $dir/conftest9012346) 2>/dev/null val=`cat $dir/conftest9012345 2>/dev/null` test -f $dir/conftest9012345 && test "$val" = 1 || some_dir_failed=true rm -f $dir/conftest9012345 $dir/conftest9012346 2> /dev/null done $some_dir_failed || AC_DEFINE(HAVE_LONG_FILE_NAMES) ])dnl dnl define(AC_RESTARTABLE_SYSCALLS, [echo checking for restartable system calls AC_TEST_PROGRAM( [/* Exit 0 (true) if wait returns something other than -1, i.e. the pid of the child, which means that wait was restarted after getting the signal. */ #include #include ucatch (isig) { } main () { int i = fork (), status; if (i == 0) { sleep (3); kill (getppid (), SIGINT); sleep (3); exit (0); } signal (SIGINT, ucatch); status = wait(&i); if (status == -1) wait(&i); exit (status == -1); } ], AC_DEFINE(HAVE_RESTARTABLE_SYSCALLS)) ])dnl dnl define(AC_FIND_X, [# If we find X, set shell vars x_includes and x_libraries to the paths. no_x=true AC_FIND_X_XMKMF if test -z "$im_usrlibdir"; then AC_FIND_X_DIRECT fi if test -n "$verbose"; then test -n "$x_includes" && echo " found X11 headers in $x_includes" test -n "$x_libraries" && echo " found X11 libraries in $x_libraries" fi])dnl dnl dnl Internal subroutine of AC_FIND_X. define(AC_FIND_X_XMKMF, [echo checking for X include and library files with xmkmf rm -fr conftestdir if mkdir conftestdir; then cd conftestdir cat > Imakefile <<\EOF acfindx: @echo "im_incroot=$(INCROOT); im_usrlibdir=$(USRLIBDIR); im_libdir=$(LIBDIR)" EOF if (xmkmf) >/dev/null 2>/dev/null && test -f Makefile; then no_x= # GNU make sometimes prints "make[1]: Entering...", which would confuse us. eval `make acfindx | grep -v make` # Open Windows xmkmf reportedly sets LIBDIR instead of USRLIBDIR. if test ! -f $im_usrlibdir/libX11.a && test -f $im_libdir/libX11.a; then im_usrlibdir=$im_libdir fi case "$im_incroot" in /usr/include) ;; *) x_includes="$im_incroot" ;; esac case "$im_usrlibdir" in /usr/lib | /lib) ;; *) x_libraries="$im_usrlibdir" ;; esac fi cd .. rm -fr conftestdir fi ])dnl dnl dnl Internal subroutine of AC_FIND_X. define(AC_FIND_X_DIRECT, [echo checking for X include and library files directly AC_TEST_CPP([#include ], no_x=, for dir in \ /usr/local/include \ /usr/unsupported/include \ /usr/x386/include \ /usr/local/x11r5/include \ /usr/include/X11R5 \ /usr/include/X11R4 \ /usr/X11R5/include \ /usr/X11/include \ /usr/openwin/include \ /usr/openwin/share/include \ /usr/lpp/Xamples/include \ ; \ do if test -r $dir/X11/Intrinsic.h; then x_includes=$dir; no_x= break fi done) # Check for the libraries. First see if replacing the `include' by # `lib' works. AC_HAVE_LIBRARY(Xt, no_x=, for dir in `echo "$x_includes" | sed s/include/lib/` \ /usr/local/lib \ /usr/unsupported/lib \ /usr/x386/lib \ /usr/local/x11r5/lib \ /usr/lib/X11 \ /usr/lib/X11R4 \ /usr/X11R5/lib \ /usr/X11/lib \ /usr/openwin/lib \ /usr/lpp/Xamples/lib \ ; \ do for extension in a so sl; do if test -r $dir/libXt.$extension; then x_libraries=$dir; no_x= break 2 fi done done)])dnl dnl dnl dnl checks for UNIX variants dnl dnl define(AC_AIX, [echo checking for AIX AC_BEFORE([$0], [AC_COMPILE_CHECK])AC_BEFORE([$0], [AC_TEST_PROGRAM])AC_BEFORE([$0], [AC_HEADER_EGREP])AC_BEFORE([$0], [AC_TEST_CPP])AC_PROGRAM_EGREP(yes, [#ifdef _AIX yes #endif ], AC_DEFINE(_ALL_SOURCE)) ])dnl dnl define(AC_MINIX, [AC_BEFORE([$0], [AC_COMPILE_CHECK])AC_BEFORE([$0], [AC_TEST_PROGRAM])AC_BEFORE([$0], [AC_HEADER_EGREP])AC_BEFORE([$0], [AC_TEST_CPP])AC_HEADER_CHECK(minix/config.h, MINIX=1) # The Minix shell can't assign to the same variable on the same line! if test -n "$MINIX"; then AC_DEFINE(_POSIX_SOURCE) AC_DEFINE(_POSIX_1_SOURCE, 2) AC_DEFINE(_MINIX) fi ])dnl dnl define(AC_ISC_POSIX, [AC_BEFORE([$0], [AC_COMPILE_CHECK])AC_BEFORE([$0], [AC_TEST_PROGRAM])AC_BEFORE([$0], [AC_HEADER_EGREP])AC_BEFORE([$0], [AC_TEST_CPP])echo checking for POSIXized ISC if test -d /etc/conf/kconfig.d && grep _POSIX_VERSION [/usr/include/sys/unistd.h] >/dev/null 2>&1 then ISC=1 # If later tests want to check for ISC. AC_DEFINE(_POSIX_SOURCE) if test -n "$GCC"; then CC="$CC -posix" else CC="$CC -Xp" fi fi ])dnl dnl define(AC_XENIX_DIR, [AC_REQUIRE([AC_DIR_HEADER])echo checking for Xenix AC_PROGRAM_EGREP(yes, [#if defined(M_XENIX) && !defined(M_UNIX) yes #endif ], XENIX=1) if test -n "$XENIX"; then LIBS="$LIBS -lx" case "$DEFS" in *SYSNDIR*) ;; *) LIBS="-ldir $LIBS" ;; # Make sure -ldir precedes any -lx. esac fi ])dnl dnl define(AC_SCO_INTL, [AC_HAVE_LIBRARY(intl, LIBS="$LIBS -lintl") ])dnl dnl define(AC_IRIX_SUN, [AC_HAVE_LIBRARY(sun, LIBS="$LIBS -lsun") ])dnl dnl define(AC_DYNIX_SEQ, [AC_HAVE_LIBRARY(seq, LIBS="$LIBS -lseq") ])dnl dnl define(AC_STAT_MACROS_BROKEN,[echo checking for broken stat file mode macros AC_PROGRAM_EGREP([You lose], [#include #include #ifdef S_ISBLK #if S_ISBLK (S_IFDIR) You lose. #endif #ifdef S_IFCHR #if S_ISBLK (S_IFCHR) You lose. #endif #endif /* S_IFCHR */ #endif /* S_ISBLK */ #ifdef S_ISLNK #if S_ISLNK (S_IFREG) You lose. #endif #endif /* S_ISLNK */ #ifdef S_ISSOCK #if S_ISSOCK (S_IFREG) You lose. #endif #endif /* S_ISSOCK */ ], AC_DEFINE(STAT_MACROS_BROKEN))])dnl dnl autoconf-1.7/configure100777 74720 14 21677 5463274675 14057 0ustar eggertguest#!/bin/sh # Guess values for system-dependent variables and create Makefiles. # Generated automatically using autoconf. # Copyright (C) 1991, 1992, 1993 Free Software Foundation, Inc. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. # Usage: configure [--srcdir=DIR] [--host=HOST] [--gas] [--nfp] # [--prefix=PREFIX] [--exec-prefix=PREFIX] [--with-PACKAGE[=VALUE]] # Ignores all args except --srcdir, --prefix, --exec-prefix, and # --with-PACKAGE[=VALUE] unless this script has special code to handle it. for arg do # Handle --exec-prefix with a space before the argument. if test x$next_exec_prefix = xyes; then exec_prefix=$arg; next_exec_prefix= # Handle --host with a space before the argument. elif test x$next_host = xyes; then next_host= # Handle --prefix with a space before the argument. elif test x$next_prefix = xyes; then prefix=$arg; next_prefix= # Handle --srcdir with a space before the argument. elif test x$next_srcdir = xyes; then srcdir=$arg; next_srcdir= else case $arg in # For backward compatibility, recognize -exec-prefix and --exec_prefix. -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* | --exec=* | --exe=* | --ex=* | --e=*) exec_prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- | --exec | --exe | --ex | --e) next_exec_prefix=yes ;; -gas | --gas | --ga | --g) ;; -host=* | --host=* | --hos=* | --ho=* | --h=*) ;; -host | --host | --hos | --ho | --h) next_host=yes ;; -nfp | --nfp | --nf) ;; -prefix=* | --prefix=* | --prefi=* | --pref=* | --pre=* | --pr=* | --p=*) prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; -prefix | --prefix | --prefi | --pref | --pre | --pr | --p) next_prefix=yes ;; -srcdir=* | --srcdir=* | --srcdi=* | --srcd=* | --src=* | --sr=* | --s=*) srcdir=`echo $arg | sed 's/[-a-z_]*=//'` ;; -srcdir | --srcdir | --srcdi | --srcd | --src | --sr | --s) next_srcdir=yes ;; -with-* | --with-*) package=`echo $arg|sed -e 's/-*with-//' -e 's/=.*//'` # Reject names that aren't valid shell variable names. if test -n "`echo $package| sed 's/[-a-zA-Z0-9_]//g'`"; then echo "configure: $package: invalid package name" >&2; exit 1 fi package=`echo $package| sed 's/-/_/g'` case "$arg" in *=*) val="`echo $arg|sed 's/[^=]*=//'`" ;; *) val=1 ;; esac eval "with_$package='$val'" ;; -v | -verbose | --verbose | --verbos | --verbo | --verb | --ver | --ve | --v) verbose=yes ;; *) ;; esac fi done trap 'rm -fr conftest* confdefs* core; exit 1' 1 3 15 trap 'rm -f confdefs*' 0 # NLS nuisances. # These must not be set unconditionally because not all systems understand # e.g. LANG=C (notably SCO). if test "${LC_ALL+set}" = 'set' ; then LC_ALL=C; export LC_ALL; fi if test "${LANG+set}" = 'set' ; then LANG=C; export LANG; fi rm -f conftest* confdefs.h # AIX cpp loses on an empty file, so make sure it contains at least a newline. echo > confdefs.h compile='${CC-cc} $CFLAGS conftest.c -o conftest $LIBS >/dev/null 2>&1' # A filename unique to this package, relative to the directory that # configure is in, which we can look for to find out if srcdir is correct. unique_file=acgeneral.m4 # Find the source files, if location was not specified. if test -z "$srcdir"; then srcdirdefaulted=yes # Try the directory containing this script, then `..'. prog=$0 confdir=`echo $prog|sed 's%/[^/][^/]*$%%'` test "X$confdir" = "X$prog" && confdir=. srcdir=$confdir if test ! -r $srcdir/$unique_file; then srcdir=.. fi fi if test ! -r $srcdir/$unique_file; then if test x$srcdirdefaulted = xyes; then echo "configure: Can not find sources in \`${confdir}' or \`..'." 1>&2 else echo "configure: Can not find sources in \`${srcdir}'." 1>&2 fi exit 1 fi # Preserve a srcdir of `.' to avoid automounter screwups with pwd. # But we can't avoid them for `..', to make subdirectories work. case $srcdir in .|/*|~*) ;; *) srcdir=`cd $srcdir; pwd` ;; # Make relative path absolute. esac # Save the original args to write them into config.status later. configure_args="$*" for p in gnum4 gm4 do if test -z "$M4"; then # Extract the first word of `$p', so it can be a program name with args. set dummy $p; word=$2 echo checking for $word IFS="${IFS= }"; saveifs="$IFS"; IFS="${IFS}:" for dir in $PATH; do test -z "$dir" && dir=. if test -f $dir/$word; then M4="$p" break fi done IFS="$saveifs" fi test -n "$M4" && test -n "$verbose" && echo " setting M4 to $M4" test -n "$M4" && break done test -n "$M4" || M4="m4" # Make sure to not get the incompatible SysV /etc/install and # /usr/sbin/install, which might be in PATH before a BSD-like install, # or the SunOS /usr/etc/install directory, or the AIX /bin/install, # or the AFS install, which mishandles nonexistent args, or # /usr/ucb/install on SVR4, which tries to use the nonexistent group # `staff'. On most BSDish systems install is in /usr/bin, not /usr/ucb # anyway. Sigh. if test "z${INSTALL}" = "z" ; then echo checking for install IFS="${IFS= }"; saveifs="$IFS"; IFS="${IFS}:" for dir in $PATH; do test -z "$dir" && dir=. case $dir in /etc|/usr/sbin|/usr/etc|/usr/afsws/bin|/usr/ucb) ;; *) if test -f $dir/installbsd; then INSTALL="$dir/installbsd -c" # OSF1 INSTALL_PROGRAM='$(INSTALL)' INSTALL_DATA='$(INSTALL) -m 644' break fi if test -f $dir/install; then if grep dspmsg $dir/install >/dev/null 2>&1; then : # AIX else INSTALL="$dir/install -c" INSTALL_PROGRAM='$(INSTALL)' INSTALL_DATA='$(INSTALL) -m 644' break fi fi ;; esac done IFS="$saveifs" fi INSTALL=${INSTALL-cp} test -n "$verbose" && echo " setting INSTALL to $INSTALL" INSTALL_PROGRAM=${INSTALL_PROGRAM-'$(INSTALL)'} test -n "$verbose" && echo " setting INSTALL_PROGRAM to $INSTALL_PROGRAM" INSTALL_DATA=${INSTALL_DATA-'$(INSTALL)'} test -n "$verbose" && echo " setting INSTALL_DATA to $INSTALL_DATA" # Set default prefixes. if test -n "$prefix"; then test -z "$exec_prefix" && exec_prefix='${prefix}' prsub="s%^prefix\\([ ]*\\)=\\([ ]*\\).*$%prefix\\1=\\2$prefix%" fi if test -n "$exec_prefix"; then prsub="$prsub s%^exec_prefix\\([ ]*\\)=\\([ ]*\\).*$%exec_prefix\\1=\\2$exec_prefix%" fi # Quote sed substitution magic chars in DEFS. cat >conftest.def < config.status </dev/null | sed 1q`: # # $0 $configure_args for arg do case "\$arg" in -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r) echo running \${CONFIG_SHELL-/bin/sh} $0 $configure_args exec \${CONFIG_SHELL-/bin/sh} $0 $configure_args ;; *) echo "Usage: config.status --recheck" 2>&1; exit 1 ;; esac done trap 'rm -f Makefile; exit 1' 1 3 15 M4='$M4' INSTALL='$INSTALL' INSTALL_PROGRAM='$INSTALL_PROGRAM' INSTALL_DATA='$INSTALL_DATA' LIBS='$LIBS' srcdir='$srcdir' DEFS='$DEFS' prefix='$prefix' exec_prefix='$exec_prefix' prsub='$prsub' extrasub='$extrasub' EOF cat >> config.status <<\EOF top_srcdir=$srcdir CONFIG_FILES=${CONFIG_FILES-"Makefile"} for file in .. ${CONFIG_FILES}; do if test "x$file" != x..; then srcdir=$top_srcdir # Remove last slash and all that follows it. Not all systems have dirname. dir=`echo $file|sed 's%/[^/][^/]*$%%'` if test "$dir" != "$file"; then test "$top_srcdir" != . && srcdir=$top_srcdir/$dir test ! -d $dir && mkdir $dir fi echo creating $file rm -f $file echo "# Generated automatically from `echo $file|sed 's|.*/||'`.in by configure." > $file sed -e " $prsub $extrasub s%@M4@%$M4%g s%@INSTALL@%$INSTALL%g s%@INSTALL_PROGRAM@%$INSTALL_PROGRAM%g s%@INSTALL_DATA@%$INSTALL_DATA%g s%@LIBS@%$LIBS%g s%@srcdir@%$srcdir%g s%@DEFS@%$DEFS% " $top_srcdir/${file}.in >> $file fi; done exit 0 EOF chmod +x config.status ${CONFIG_SHELL-/bin/sh} config.status autoconf-1.7/configure.in100666 21570 13 242 5455552127 13510 0ustar djmuserdnl Process this file with autoconf to produce a configure script. AC_INIT(acgeneral.m4) AC_PROGRAMS_CHECK(M4, gnum4 gm4, m4) AC_PROG_INSTALL AC_OUTPUT(Makefile) autoconf-1.7/autoheader.sh100777 12241 1 14025 5454335634 14666 0ustar rolanddaemon#!/bin/sh # autoheader -- create `config.h.in' from `configure.in' # Copyright (C) 1992, 1993 Free Software Foundation, Inc. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. # Written by Roland McGrath. # If given no args, create `config.h.in' from template file `configure.in'. # With one arg, create a header file on standard output from # the given template file. usage="Usage: autoheader [-h] [--help] [-m dir] [--macrodir=dir] [-v] [--version] [template-file]" # NLS nuisances. # These must not be set unconditionally because not all systems understand # e.g. LANG=C (notably SCO). if test "${LC_ALL+set}" = 'set' ; then LC_ALL=C; export LC_ALL; fi if test "${LANG+set}" = 'set' ; then LANG=C; export LANG; fi test -z "${AC_MACRODIR}" && AC_MACRODIR=@datadir@ test -z "${M4}" && M4=@M4@ print_version="" while test $# -gt 0 ; do case "z${1}" in z-h | z--help | z--h* ) echo "${usage}"; exit 0 ;; z--macrodir=* | z--m*=* ) AC_MACRODIR="`echo \"${1}\" | sed -e 's/^[^=]*=//'`" shift ;; z-m | z--macrodir | z--m* ) shift test $# -eq 0 && { echo "${usage}" 1>&2; exit 1; } AC_MACRODIR="${1}" shift ;; z-v | z--version | z--v* ) print_version="-DAC_PRINT_VERSION" shift ;; z-- ) # Stop option processing shift; break ;; z- ) # Use stdin as input. break ;; z-* ) echo "${usage}" 1>&2; exit 1 ;; * ) break ;; esac done TEMPLATES="${AC_MACRODIR}/acconfig.h" test -r acconfig.h && TEMPLATES="${TEMPLATES} acconfig.h" MACROFILES="${AC_MACRODIR}/acgeneral.m4 ${AC_MACRODIR}/acspecific.m4" test -r ${AC_MACRODIR}/aclocal.m4 \ && MACROFILES="${MACROFILES} ${AC_MACRODIR}/aclocal.m4" test -r aclocal.m4 && MACROFILES="${MACROFILES} aclocal.m4" MACROFILES="${print_version} ${MACROFILES}" case $# in 0) if test -n "$print_version" then infile=/dev/null else infile=configure.in; fi ;; 1) infile=$1 ;; *) echo "$usage" >&2; exit 1 ;; esac # These are the alternate definitions of the acgeneral.m4 macros we want to # redefine. They produce strings in the output marked with "@@@" so we can # easily extract the information we want. The `#' at the end of the first # line of each definition seems to be necessary to prevent m4 from eating # the newline, which makes the @@@ not always be at the beginning of a line. frob='define([AC_DEFINE],[# @@@syms="$syms $1"@@@ ])dnl define([AC_HAVE_FUNCS],[# @@@funcs="$funcs $1"@@@ ])dnl define([AC_HAVE_HEADERS],[# @@@headers="$headers $1"@@@ ])dnl define([AC_CONFIG_HEADER],[# @@@config_h=$1@@@ ])dnl define([AC_HAVE_LIBRARY], [# changequote(/,/)dnl define(/libname/, dnl patsubst(patsubst($1, /lib\([^\.]*\)\.a/, /\1/), /-l/, //))dnl changequote([,])dnl @@@libs="$libs libname"@@@ # If it was found, we do: $2 # If it was not found, we do: $3 ])dnl ' config_h=config.h syms= funcs= headers= libs= # We extract assignments of SYMS, FUNCS, HEADERS, and LIBS from the # modified autoconf processing of the input file. The sed hair is # necessary to win for multi-line macro invocations. eval "`echo \"$frob\" \ | $M4 $MACROFILES - $infile \ | sed -n -e ' : again /^@@@.*@@@$/s/^@@@\(.*\)@@@$/\1/p /^@@@/{ s/^@@@//p n s/^/@@@/ b again }'`" test -n "$print_version" && exit 0 # Make SYMS newline-separated rather than blank-separated, and remove dups. syms="`for sym in $syms; do echo $sym; done | sort | uniq`" if test $# -eq 0; then tmpout=autoh$$ trap "rm -f $tmpout; exit 1" 1 3 15 exec > $tmpout fi # Don't write "do not edit" -- it will get copied into the # config.h, which it's ok to edit. echo "/* ${config_h}.in. Generated automatically from $infile by autoheader. */" # This puts each paragraph on its own line, separated by @s. if test -n "$syms"; then # Make sure the boundary of template files is also the boundary # of the paragraph. Extra newlines don't hurt since they will # be removed. for t in $TEMPLATES; do cat $t; echo; echo; done | # The sed script is suboptimal because it has to take care of # some broken seds (e.g. AIX) that remove '\n' from the # pattern/hold space if the line is empty. (junio@twinsun.com). sed -n -e ' /^[ ]*$/{ x s/\n/@/g p s/.*/@/ x } H' | sed -e 's/@@*/@/g' | # Select each paragraph that refers to a symbol we picked out above. fgrep "$syms" | tr @ \\012 fi for func in `for x in $funcs; do echo $x; done | sort | uniq`; do sym="`echo ${func} | sed 's/[^a-zA-Z0-9_]/_/g' | tr '[a-z]' '[A-Z]'`" echo " /* Define if you have ${func}. */ #undef HAVE_${sym}" done for header in `for x in $headers; do echo $x; done | sort | uniq`; do sym="`echo ${header} | sed 's/[^a-zA-Z0-9_]/_/g' | tr '[a-z]' '[A-Z]'`" echo " /* Define if you have the <${header}> header file. */ #undef HAVE_${sym}" done for lib in `for x in $libs; do echo $x; done | sort | uniq`; do sym="`echo ${lib} | sed 's/[^a-zA-Z0-9_]/_/g' | tr '[a-z]' '[A-Z]'`" echo " /* Define if you have the ${lib} library (-l${lib}). */ #undef HAVE_LIB${sym}" done status=0 for sym in $syms; do if fgrep $sym $TEMPLATES >/dev/null; then : # All is well. else echo "$0: Symbol \`${sym}' is not covered by $TEMPLATES" >&2 status=1 fi done if test $# -eq 0; then if test $status -eq 0; then if cmp -s $tmpout ${config_h}.in; then rm -f $tmpout else mv -f $tmpout ${config_h}.in fi else rm -f $tmpout fi fi exit $status autoconf-1.7/mkinstalldirs100777 21570 13 1153 5432727042 14022 0ustar djmuser#!/bin/sh # Make directory hierarchy. # Written by Noah Friedman # Public domain. defaultIFS=' ' IFS="${IFS-${defaultIFS}}" errstatus=0 for file in ${1+"$@"} ; do oIFS="${IFS}" # Some sh's can't handle IFS=/ for some reason. IFS='%' set - `echo ${file} | sed -e 's@/@%@g' -e 's@^%@/@'` IFS="${oIFS}" pathcomp='' for d in ${1+"$@"} ; do pathcomp="${pathcomp}${d}" if test ! -d "${pathcomp}"; then echo "mkdir $pathcomp" 1>&2 mkdir "${pathcomp}" || errstatus=$? fi pathcomp="${pathcomp}/" done done exit $errstatus # eof