// Rules database // (C) Copyright Beman G. Dawes 1998. All rights reserved. // This file contains licensed material. All uses are prohibited except // as specified in a Software License Agreement from the copyright holder. // $Id: rules.txt 1.1 1998/09/14 16:25:01 Beman Rel $ // Format: // // rule-number, country-wt, state-wt, loc-wt, sec-loc-wt, street-wt, // str-parse-wt, str-parse2-wt, str-parser, ground-zero-lat, ground-zero-lon // Requirements: // // * The country-wt entry must be the same for any rules. (See Notes, below.) // * Ground zero lat/lon must be less or equal the minimum value which // actually appears in those fields. The maximum offset for latitude // is +45 degrees, the maximum offset of longitude is +90 degrees. // Filenames: // // * The traditional names are used for the US word tables. // * International word table names are the same as the traditional names, // but prefixed with the ISO-3166 two character identifier (and an underscore) // for a representative country which uses the table. Thus ar_strparse.wt // is the street parser word table for Argentina. It may also be used for other // countries with similar address usage, such as Uruguay and Paraguay. // * If a country requires multiple tables because of multiple languages, // a second prefix (and underscore) specifies the language via the ISO-639 // two character identifier. Thus ca_fr_strparse.wt would name a Canadian // French language street parser word table. // Notes: // // * The chicken-and-egg paradox (word table contents required to choose word table) // is broken as follows: // * Country is determined using rule 1. // * State is determined using the country rule. // * Locality is determined using the state rule. // * Everything else determined using the locality rule. 6 // number of rules // 1 - US except AK, HI, and PR. country, state, locality, secloc, street, USParser, strparse, strpass2, 20000000,-130000000 // 2 - US, AK country, state, locality, secloc, street, USParser, strparse, strpass2, 40000000,-180000000 // 3 - US, HI country, state, locality, secloc, street, USParser, strparse, strpass2, 0,-180000000 // 4 - US, PR // Note: PR should be using Spanish parting tables (not the english ones) country, state, locality, secloc, street, USParser, strparse, strpass2, 0,-70000000 // 5 - US, VI country, state, locality, secloc, street, USParser, strparse, strpass2, 0,-70000000 // 6 - US, GU country, state, locality, secloc, street, USParser, strparse, strpass2, 0,140000000