safety-screen/Nominatim
1
0
Fork 0
forked from hans/Nominatim
Code Pull Requests Activity

postcode token should have transliterated term in word_token

Browse Source
This commit is contained in:
Sarah Hoffmann
2025-03-18 22:55:33 +01:00
parent edc5ada625
commit 7b3c725f2a
1 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/nominatim_api/search/icu_tokenizer.py
View File

@@ -193,10 +193,12 @@ class ICUQueryAnalyzer(AbstractQueryAnalyzer):
self.add_extra_tokens(query) self.add_extra_tokens(query)
for start, end, pc in self.postcode_parser.parse(query): for start, end, pc in self.postcode_parser.parse(query):
term = ' '.join(n.term_lookup for n in query.nodes[start + 1:end + 1])
query.add_token(qmod.TokenRange(start, end), query.add_token(qmod.TokenRange(start, end),
qmod.TOKEN_POSTCODE, qmod.TOKEN_POSTCODE,
ICUToken(penalty=0.1, token=0, count=1, addr_count=1, ICUToken(penalty=0.1, token=0, count=1, addr_count=1,
lookup_word=pc, word_token=pc, info=None)) lookup_word=pc, word_token=term,
info=None))
self.rerank_tokens(query) self.rerank_tokens(query)
log().table_dump('Word tokens', _dump_word_tokens(query)) log().table_dump('Word tokens', _dump_word_tokens(query))