drop category tokens when they make up a full phrase

[nominatim.git] / nominatim / api / v1 / helpers.py

diff --git a/nominatim/api/v1/helpers.py b/nominatim/api/v1/helpers.py
index 325e5bc629911dc446476c1499499eb641e741dd..896a131cc8c98da4eb982e69068e4f3df6cee2a8 100644 (file)
--- a/nominatim/api/v1/helpers.py
+++ b/nominatim/api/v1/helpers.py
@@ -37,7 +37,7 @@ def zoom_to_rank(zoom: int) -> int:
     return REVERSE_MAX_RANKS[max(0, min(18, zoom))]
 
 
     return REVERSE_MAX_RANKS[max(0, min(18, zoom))]
 
 

-FEATURE_TYPE_TO_RANK: Dict[Optional[str], Any] = {
+FEATURE_TYPE_TO_RANK: Dict[Optional[str], Tuple[int, int]] = {

     'country': (4, 4),
     'state': (8, 8),
     'city': (14, 16),
     'country': (4, 4),
     'state': (8, 8),
     'city': (14, 16),


@@ -108,15 +108,18 @@ def deduplicate_results(results: SearchResults, max_results: int) -> SearchResul
             assert result.names and 'ref' in result.names
             if any(_is_postcode_relation_for(r, result.names['ref']) for r in results):
                 continue
             assert result.names and 'ref' in result.names
             if any(_is_postcode_relation_for(r, result.names['ref']) for r in results):
                 continue

-        classification = (result.osm_object[0] if result.osm_object else None,
-                          result.category,
-                          result.display_name,
-                          result.rank_address)
-        if result.osm_object not in osm_ids_done \
-           and classification not in classification_done:
+        if result.source_table == SourceTable.PLACEX:
+            classification = (result.osm_object[0] if result.osm_object else None,
+                              result.category,
+                              result.display_name,
+                              result.rank_address)
+            if result.osm_object not in osm_ids_done \
+               and classification not in classification_done:
+                deduped.append(result)
+            osm_ids_done.add(result.osm_object)
+            classification_done.add(classification)
+        else:

             deduped.append(result)
             deduped.append(result)

-        osm_ids_done.add(result.osm_object)
-        classification_done.add(classification)

         if len(deduped) >= max_results:
             break
 
         if len(deduped) >= max_results:
             break
 


@@ -136,10 +139,10 @@ def _deg(axis:str) -> str:
     return f"(?P<{axis}_deg>\\d+\\.\\d+)°?"
 
 def _deg_min(axis: str) -> str:
     return f"(?P<{axis}_deg>\\d+\\.\\d+)°?"
 
 def _deg_min(axis: str) -> str:

-    return f"(?P<{axis}_deg>\\d+)[°\\s]+(?P<{axis}_min>[\\d.]+)?[′']*"
+    return f"(?P<{axis}_deg>\\d+)[°\\s]+(?P<{axis}_min>[\\d.]+)[′']*"

 
 def _deg_min_sec(axis: str) -> str:
 
 def _deg_min_sec(axis: str) -> str:

-    return f"(?P<{axis}_deg>\\d+)[°\\s]+(?P<{axis}_min>\\d+)[′'\\s]+(?P<{axis}_sec>[\\d.]+)?[\"″]*"
+    return f"(?P<{axis}_deg>\\d+)[°\\s]+(?P<{axis}_min>\\d+)[′'\\s]+(?P<{axis}_sec>[\\d.]+)[\"″]*"

 
 COORD_REGEX = [re.compile(r'(?:(?P<pre>.*?)\s+)??' + r + r'(?:\s+(?P<post>.*))?') for r in (
     r"(?P<ns>[NS])\s*" + _deg('lat') + r"[\s,]+" + r"(?P<ew>[EW])\s*" + _deg('lon'),
 
 COORD_REGEX = [re.compile(r'(?:(?P<pre>.*?)\s+)??' + r + r'(?:\s+(?P<post>.*))?') for r in (
     r"(?P<ns>[NS])\s*" + _deg('lat') + r"[\s,]+" + r"(?P<ew>[EW])\s*" + _deg('lon'),