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

[fix] spacing in phrases

Browse Source
This commit is contained in:
Al
2016-01-23 21:59:49 -05:00
parent 3a8c3dfcf6
commit 1bb797f783
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
scripts/geodata/osm/osm_address_training_data.py
View File

@@ -412,7 +412,7 @@ def osm_abbreviate(gazetteer, s, language, abbreviate_prob=0.3, separate_prob=0.
token = random.choice(abbreviations) if abbreviations else canonical token = random.choice(abbreviations) if abbreviations else canonical
token = recase_abbreviation(token, tokens[i:i + len(t)]) token = recase_abbreviation(token, tokens[i:i + len(t)])
abbreviated.append(token) abbreviated.append(token)
if i + len(t) < n - 1 and raw_tokens[i + len(t)][0] > sum(raw_tokens[i + len(t) - 1][:2]): if i + len(t) < n and raw_tokens[i + len(t)][0] > sum(raw_tokens[i + len(t) - 1][:2]):
abbreviated.append(u' ') abbreviated.append(u' ')
break break
elif is_prefix: elif is_prefix: