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

[names] Name affixes respect hyphens and lack of whitespace (for ideographic languages)

Browse Source
This commit is contained in:
Al
2016-05-16 10:22:09 -04:00
parent 64ad15fa36
commit 8b9e351961
1 changed files with 14 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
scripts/geodata/names/normalization.py
View File

@@ -28,18 +28,23 @@ class NameAffixes(object):
conf = yaml.load(open(os.path.join(config_dir, filename)))
prefixes = conf.get('prefixes', [])
name_prefixes = [safe_decode(phrase).lower() for phrase in prefixes]
self.language_prefixes[lang] = name_prefixes
prefixes = [safe_decode(phrase).lower() for phrase in conf.get('prefixes', [])]
prefixes_no_whitespace = [safe_decode(phrase).lower() for phrase in conf.get('prefixes_no_whitespace', [])]
suffixes = conf.get('suffixes', [])
name_suffixes = [safe_decode(phrase).lower() for phrase in suffixes]
self.language_suffixes[lang] = name_suffixes
self.language_prefixes[lang] = prefixes + prefixes_no_whitespace
whitespace_phrase = six.u(' ') if conf.get('whitespace', True) else six.u('')
suffixes = [safe_decode(phrase).lower() for phrase in conf.get('suffixes', [])]
suffixes_no_whitespace = [safe_decode(phrase).lower() for phrase in conf.get('suffixes_no_whitespace', [])]
prefix_regex = six.u('^(?:{})').format(six.u('|').join(['{}{}'.format(s, whitespace_phrase) for s in name_prefixes]))
suffix_regex = six.u('(?:{})$').format(six.u('|').join(['{}{}'.format(whitespace_phrase, s) for s in name_suffixes]))
self.language_suffixes[lang] = suffixes + suffixes_no_whitespace
whitespace_phrase = six.u('[ \-]')
all_prefixes = [six.u('{}{}').format(s, whitespace_phrase) for s in prefixes] + prefixes_no_whitespace
all_suffixes = [six.u('{}{}').format(whitespace_phrase, s) for s in suffixes] + suffixes_no_whitespace
prefix_regex = six.u('^(?:{})').format(six.u('|').join(all_prefixes))
suffix_regex = six.u('(?:{})$').format(six.u('|').join(all_suffixes))
self.language_prefix_regexes[lang] = re.compile(prefix_regex, re.I | re.UNICODE)
self.language_suffix_regexes[lang] = re.compile(suffix_regex, re.I | re.UNICODE)