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

[languages] Non-default language canonicals, more test cases

Browse Source
This commit is contained in:
Al
2015-08-24 02:21:53 -04:00
parent c1ce91abbf
commit e1d336716c
2 changed files with 10 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
scripts/geodata/language_id/disambiguation.py
View File

@@ -195,7 +195,8 @@ def disambiguate_language(text, languages):
elif not seen_languages and len(potentials) == 1 and len(t[0][1]) > 1:
possible_lang = lang if possible_lang is None or possible_lang == lang else None
if seen_languages and valid and not any((l in seen_languages for l in valid)):
if seen_languages and valid and not any((l in seen_languages for l in valid)) and \
(not any((valid_languages.get(l) for l in valid)) or any((valid_languages.get(l) for l in seen_languages))):
return AMBIGUOUS_LANGUAGE
if len(valid) == 1: