From ce065bb9ecb09e076c6998105b30862905841d2e Mon Sep 17 00:00:00 2001 From: Al Date: Sun, 22 Nov 2015 18:01:07 -0500 Subject: [PATCH] [fix] using new pypostal tokenize API --- scripts/geodata/language_id/disambiguation.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/geodata/language_id/disambiguation.py b/scripts/geodata/language_id/disambiguation.py index e1e7d715..57d92f0d 100644 --- a/scripts/geodata/language_id/disambiguation.py +++ b/scripts/geodata/language_id/disambiguation.py @@ -197,7 +197,7 @@ def disambiguate_language(text, languages): read_len += script_len num_defaults = sum((1 for lang, default in valid_languages.iteritems() if default)) - tokens = [(c, t.rstrip('.')) for c, t in tokenize(safe_decode(text).replace(u'-', u' ').lower())] + tokens = [(c, t.rstrip('.')) for t, c in tokenize(safe_decode(text).replace(u'-', u' ').lower())] current_lang = None possible_lang = None