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

[fix] only commas and hyphens need to be eliminated at the end of phrases in untagged address formatting

Browse Source
This commit is contained in:
Al
2015-09-27 19:25:28 -04:00
parent d3bfaf6b43
commit 93b3110a49
1 changed files with 5 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
scripts/geodata/address_formatting/formatter.py
View File

@@ -167,18 +167,21 @@ class AddressFormatter(object):
def strip_component(self, value, tagged=False): def strip_component(self, value, tagged=False):
if not tagged: if not tagged:
comma = token_types.COMMA.value
hyphen = token_types.HYPHEN.value
start = end = 0 start = end = 0
tokens = tokenize_raw(value.strip()) tokens = tokenize_raw(value.strip())
for token_start, token_length, token_type in tokens: for token_start, token_length, token_type in tokens:
start = token_start start = token_start
if token_type < token_types.PERIOD.value: if token_type in (comma, hyphen):
break break
else: else:
start = token_start + token_length start = token_start + token_length
for token_start, token_length, token_type in reversed(tokens): for token_start, token_length, token_type in reversed(tokens):
end = token_start + token_length end = token_start + token_length
if token_type < token_types.PERIOD.value: if token_type in (comma, hyphen):
break break
else: else:
end = token_start end = token_start