add lookup word to variants in word table

src/nominatim_db/tokenizer/icu_tokenizer.py

@@ -585,10 +585,14 @@ class ICUNameAnalyzer(AbstractAnalyzer):
             if word_id:
                 result = self._cache.housenumbers.get(word_id, result)
                 if result[0] is None:
-                    variants = analyzer.compute_variants(word_id)
+                    varout = analyzer.compute_variants(word_id)
+                    if isinstance(varout, tuple):
+                        variants = varout[0]
+                    else:
+                        variants = varout
                     if variants:
                         hid = execute_scalar(self.conn, "SELECT create_analyzed_hnr_id(%s, %s)",
-                                             (word_id, list(variants)))
+                                             (word_id, variants))
                         result = hid, variants[0]
                         self._cache.housenumbers[word_id] = result
 
@@ -633,13 +637,17 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
             full, part = self._cache.names.get(token_id, (None, None))
             if full is None:
-                variants = analyzer.compute_variants(word_id)
+                varset = analyzer.compute_variants(word_id)
+                if isinstance(varset, tuple):
+                    variants, lookups = varset
+                else:
+                    variants, lookups = varset, None
                 if not variants:
                     continue
 
                 with self.conn.cursor() as cur:
-                    cur.execute("SELECT * FROM getorcreate_full_word(%s, %s)",
-                                (token_id, variants))
+                    cur.execute("SELECT * FROM getorcreate_full_word(%s, %s, %s)",
+                                (token_id, variants, lookups))
                     full, part = cast(Tuple[int, List[int]], cur.fetchone())
 
                 self._cache.names[token_id] = (full, part)

src/nominatim_db/tokenizer/token_analysis/base.py

@@ -7,7 +7,7 @@
 """
 Common data types and protocols for analysers.
 """
-from typing import Mapping, List, Any
+from typing import Mapping, List, Any, Union, Tuple
 
 from ...typing import Protocol
 from ...data.place_name import PlaceName
@@ -33,7 +33,7 @@ class Analyzer(Protocol):
                     for example because the character set in use does not match.
         """
 
-    def compute_variants(self, canonical_id: str) -> List[str]:
+    def compute_variants(self, canonical_id: str) -> Union[List[str], Tuple[List[str], List[str]]]:
         """ Compute the transliterated spelling variants for the given
             canonical ID.
 

src/nominatim_db/tokenizer/token_analysis/generic.py

@@ -7,7 +7,7 @@
 """
 Generic processor for names that creates abbreviation variants.
 """
-from typing import Mapping, Dict, Any, Iterable, Iterator, Optional, List, cast
+from typing import Mapping, Dict, Any, Iterable, Optional, List, cast, Tuple
 import itertools
 
 from ...errors import UsageError
@@ -78,7 +78,7 @@ class GenericTokenAnalysis:
         """
         return cast(str, self.norm.transliterate(name.name)).strip()
 
-    def compute_variants(self, norm_name: str) -> List[str]:
+    def compute_variants(self, norm_name: str) -> Tuple[List[str], List[str]]:
         """ Compute the spelling variants for the given normalized name
             and transliterate the result.
         """
@@ -87,18 +87,20 @@ class GenericTokenAnalysis:
         for mutation in self.mutations:
             variants = mutation.generate(variants)
 
-        return [name for name in self._transliterate_unique_list(norm_name, variants) if name]
-
-    def _transliterate_unique_list(self, norm_name: str,
-                                   iterable: Iterable[str]) -> Iterator[Optional[str]]:
-        seen = set()
+        varset = set(map(str.strip, variants))
         if self.variant_only:
-            seen.add(norm_name)
+            varset.discard(norm_name)
 
-        for variant in map(str.strip, iterable):
-            if variant not in seen:
-                seen.add(variant)
-                yield self.to_ascii.transliterate(variant).strip()
+        trans = []
+        norm = []
+
+        for var in varset:
+            t = self.to_ascii.transliterate(var).strip()
+            if t:
+                trans.append(t)
+                norm.append(var)
+
+        return trans, norm
 
     def _generate_word_variants(self, norm_name: str) -> Iterable[str]:
         baseform = '^ ' + norm_name + ' ^'