From: Sarah Hoffmann <lonvia@denofr.de>
Date: Mon, 1 Apr 2024 12:25:51 +0000 (+0200)
Subject: minimum counts for tokens should always be 1
X-Git-Tag: deploy~4^2
X-Git-Url: https://git.openstreetmap.org/nominatim.git/commitdiff_plain/78c19bc006d5957f183968d086226be4b2b65fe3

minimum counts for tokens should always be 1

to avoid accidental devision by 0.
---

diff --git a/nominatim/api/search/icu_tokenizer.py b/nominatim/api/search/icu_tokenizer.py
index 05ec7690..eb90c122 100644
--- a/nominatim/api/search/icu_tokenizer.py
+++ b/nominatim/api/search/icu_tokenizer.py
@@ -122,10 +122,10 @@ class ICUToken(qmod.Token):
         else:
             lookup_word = row.word_token
 
-        return ICUToken(penalty=penalty, token=row.word_id, count=count,
+        return ICUToken(penalty=penalty, token=row.word_id, count=max(1, count),
                         lookup_word=lookup_word, is_indexed=True,
                         word_token=row.word_token, info=row.info,
-                        addr_count=addr_count)
+                        addr_count=max(1, addr_count))
 
 
 
diff --git a/nominatim/api/search/legacy_tokenizer.py b/nominatim/api/search/legacy_tokenizer.py
index bd17706e..031f2456 100644
--- a/nominatim/api/search/legacy_tokenizer.py
+++ b/nominatim/api/search/legacy_tokenizer.py
@@ -209,7 +209,7 @@ class LegacyQueryAnalyzer(AbstractQueryAnalyzer):
                 is_indexed = False
 
         return LegacyToken(penalty=penalty, token=row.word_id,
-                           count=row.search_name_count or 1,
+                           count=max(1, row.search_name_count or 1),
                            addr_count=1, # not supported
                            lookup_word=lookup_word,
                            word_token=row.word_token.strip(),