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

when rematching only distinguish between perfect, somewhat and bad match

Browse Source
This commit is contained in:
Sarah Hoffmann
2025-06-01 11:53:23 +02:00
parent 831fccdaee
commit f2236f68f1
1 changed files with 4 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
src/nominatim_api/search/geocoder.py
View File

@@ -153,11 +153,10 @@ class ForwardGeocoder:
if not words:
continue
for qword in qwords:
wdist = max(difflib.SequenceMatcher(a=qword, b=w).quick_ratio() for w in words)
if wdist < 0.5:
distance += len(qword)
else:
distance += (1.0 - wdist) * len(qword)
# only add distance penalty if there is no perfect match
if qword not in words:
wdist = max(difflib.SequenceMatcher(a=qword, b=w).quick_ratio() for w in words)
distance += len(qword) if wdist < 0.4 else 1
# Compensate for the fact that country names do not get a
# match penalty yet by the tokenizer.
# Temporary hack that needs to be removed!