hans/Nominatim
1
0
Fork 1
mirror of https://github.com/osm-search/Nominatim.git synced 2026-03-12 22:04:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

add unit tests for legacy ICU tokenizer

Browse Source
This commit is contained in:
Sarah Hoffmann
2021-05-04 18:32:57 +02:00
parent d55fc39275
commit 18c99a5c5f
3 changed files with 262 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
nominatim/tokenizer/legacy_icu_tokenizer.py
View File

@@ -185,7 +185,9 @@ class LegacyICUTokenizer:
# copy them back into the word table
copystr = io.StringIO(''.join(('{}\t{}\n'.format(*args) for args in words.items())))
with conn.cursor() as cur:
copystr.seek(0)
cur.copy_from(copystr, 'word', columns=['word_token', 'search_name_count'])
cur.execute("""UPDATE word SET word_id = nextval('seq_word')
WHERE word_id is null""")
@@ -206,7 +208,6 @@ class LegacyICUNameAnalyzer:
self.normalizer = normalizer
self.transliterator = transliterator
self.abbreviations = abbreviations
#psycopg2.extras.register_hstore(self.conn)
self._cache = _TokenCache()
@@ -267,6 +268,7 @@ class LegacyICUNameAnalyzer:
copystr.write(self.transliterator.transliterate(postcode))
copystr.write('\tplace\tpostcode\t0\n')
copystr.seek(0)
cur.copy_from(copystr, 'word',
columns=['word', 'word_token', 'class', 'type',
'search_name_count'])
@@ -309,6 +311,7 @@ class LegacyICUNameAnalyzer:
copystr.write(oper if oper in ('in', 'near') else '\\N')
copystr.write('\t0\n')
copystr.seek(0)
cur.copy_from(copystr, 'word',
columns=['word', 'word_token', 'class', 'type',
'operator', 'search_name_count'])