tommy/libpostal
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[fix] abbreviations

Browse Source
This commit is contained in:
Al
2015-10-02 23:48:21 -04:00
parent db71b65412
commit 0aa6950b6c
2 changed files with 173678 additions and 247857 deletions
Download Patch File Download Diff File Expand all files Collapse all files

421531
src/scanner.c
View File

File diff suppressed because it is too large Load Diff

4
src/scanner.re
View File

@@ -137,9 +137,9 @@ apos_word = ("'"?({latinish_letter}+"'"){latinish_letter}+"'"?);
ellipsis = ("\."{2,}|"\u2026"); ellipsis = ("\."{2,}|"\u2026");
acronym = ({letter}"\."){2,}{letter}?; acronym = ({letter}"\."){2,}{letter}?;
multi_punct_abbreviation = ({letter}+"\.")+{letter}?; multi_punct_abbreviation = ({letter}+"\.")+{letter}*;
abbrev_word = (({letter}|{possible_word_char})+"\.")+({letter}|{possible_word_char}*); abbrev_word = (({letter}|{possible_word_char})+"\.")+({letter}|{possible_word_char})*;
hyphen_plus_abbreviation = ((({abbrev_word}|{any_word}){hyphen})+({abbrev_word}))|(({abbrev_word}|{any_word}){hyphen})*({abbrev_word}{hyphen})+(({abbrev_word}|{any_word}){hyphen})*({abbrev_word}|{any_word}); hyphen_plus_abbreviation = ((({abbrev_word}|{any_word}){hyphen})+({abbrev_word}))|(({abbrev_word}|{any_word}){hyphen})*({abbrev_word}{hyphen})+(({abbrev_word}|{any_word}){hyphen})*({abbrev_word}|{any_word});