hans/Nominatim
1
0
Fork 1
mirror of https://github.com/osm-search/Nominatim.git synced 2026-02-26 11:08:13 +00:00
Code Issues Packages Projects Releases Wiki Activity

use word_token length when penalizing against postcodes

Browse Source
This commit is contained in:
Sarah Hoffmann
2025-03-18 23:07:27 +01:00
parent 7b3c725f2a
commit 2c61fe08a0
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

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

@@ -269,10 +269,10 @@ class ICUQueryAnalyzer(AbstractQueryAnalyzer):
""" """
for i, node, tlist in query.iter_token_lists(): for i, node, tlist in query.iter_token_lists():
if tlist.ttype == qmod.TOKEN_POSTCODE: if tlist.ttype == qmod.TOKEN_POSTCODE:
tlen = len(cast(ICUToken, tlist.tokens[0]).word_token)
for repl in node.starting: for repl in node.starting:
if repl.end == tlist.end and repl.ttype != qmod.TOKEN_POSTCODE \ if repl.end == tlist.end and repl.ttype != qmod.TOKEN_POSTCODE \
and (repl.ttype != qmod.TOKEN_HOUSENUMBER and (repl.ttype != qmod.TOKEN_HOUSENUMBER or tlen > 4):
or len(tlist.tokens[0].lookup_word) > 4):
repl.add_penalty(0.39) repl.add_penalty(0.39)
elif (tlist.ttype == qmod.TOKEN_HOUSENUMBER elif (tlist.ttype == qmod.TOKEN_HOUSENUMBER
and len(tlist.tokens[0].lookup_word) <= 3): and len(tlist.tokens[0].lookup_word) <= 3):