From: Sarah Hoffmann <lonvia@denofr.de>
Date: Wed, 13 Nov 2024 18:35:54 +0000 (+0100)
Subject: Merge remote-tracking branch 'upstream/master'
X-Git-Url: https://git.openstreetmap.org./nominatim.git/commitdiff_plain/e1dc4379e0cd100200ac53752442143ca4846fc5

Merge remote-tracking branch 'upstream/master'
---

e1dc4379e0cd100200ac53752442143ca4846fc5
diff --cc src/nominatim_api/search/db_search_builder.py
index 0d7487a4,632270ef..1fbb7168
--- a/src/nominatim_api/search/db_search_builder.py
+++ b/src/nominatim_api/search/db_search_builder.py
@@@ -230,8 -221,8 +221,8 @@@ class SearchBuilder
          if name_fulls:
              fulls_count = sum(t.count for t in name_fulls)
  
 -            if fulls_count < 50000 or addr_count < 30000:
 +            if fulls_count < 80000 or addr_count < 50000:
-                 yield penalty,fulls_count / (2**len(addr_tokens)), \
+                 yield penalty, fulls_count / (2**len(addr_tokens)), \
                      self.get_full_name_ranking(name_fulls, addr_partials,
                                                 fulls_count > 30000 / max(1, len(addr_tokens)))
  
diff --cc src/nominatim_api/search/icu_tokenizer.py
index c2a26510,fa14531a..c18dd8be
--- a/src/nominatim_api/search/icu_tokenizer.py
+++ b/src/nominatim_api/search/icu_tokenizer.py
@@@ -208,14 -202,8 +202,13 @@@ class ICUQueryAnalyzer(AbstractQueryAna
              standardized form search will work with. All information removed
              at this stage is inevitably lost.
          """
 -        return cast(str, self.normalizer.transliterate(text))
 +        norm = cast(str, self.normalizer.transliterate(text))
 +        numspaces = norm.count(' ')
 +        if numspaces > 4 and len(norm) <= (numspaces + 1) * 3:
 +            return ''
 +
 +        return norm
  
- 
      def split_query(self, query: qmod.QueryStruct) -> Tuple[QueryParts, WordDict]:
          """ Transliterate the phrases and split them into tokens.