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

[openaddresses] Omitting fields with the value "unknown"

Browse Source
This commit is contained in:
Al
2016-08-27 00:46:21 -04:00
parent 44221cf1f1
commit 4ed394cc1c
1 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
scripts/geodata/openaddresses/formatter.py
View File

@@ -29,6 +29,7 @@ OPENADDRESS_FORMAT_DATA_FILENAME = 'openaddresses_formatted_addresses.tsv'
numeric_range_regex = re.compile('[\s]*\-[\s]*') numeric_range_regex = re.compile('[\s]*\-[\s]*')
null_regex = re.compile('^\s*(?:null|none)\s*$', re.I) null_regex = re.compile('^\s*(?:null|none)\s*$', re.I)
unknown_regex = re.compile('^\s*(?:unknown)\s*$', re.I)
not_applicable_regex = re.compile('^\s*n\.?\s*/?\s*a\.?\s*$', re.I) not_applicable_regex = re.compile('^\s*n\.?\s*/?\s*a\.?\s*$', re.I)
@@ -212,7 +213,7 @@ class OpenAddressesFormatter(object):
if value and len(value) < 2 or is_numeric(value): if value and len(value) < 2 or is_numeric(value):
continue continue
if not_applicable_regex.match(value) or null_regex.match(value): if not_applicable_regex.match(value) or null_regex.match(value) or unknown_regex.match(value):
continue continue
for exp, sub_val in self.all_field_regex_replacements: for exp, sub_val in self.all_field_regex_replacements: