restrict deduplication to results from placex

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

diff --git a/nominatim/api/v1/helpers.py b/nominatim/api/v1/helpers.py
index ea7c125d7215540541bb81f4f02faeddfc592a86..13087280cefec24bcee197e94eadd6aa5bd653df 100644 (file)
--- a/nominatim/api/v1/helpers.py
+++ b/nominatim/api/v1/helpers.py
@@ -83,7 +83,7 @@ def extend_query_parts(queryparts: Dict[str, Any], details: Dict[str, Any],
     if parsed.countries:
         queryparts['countrycodes'] = ','.join(parsed.countries)
     queryparts['exclude_place_ids'] = \
-        ','.join(chain(excluded, map(str, parsed.excluded)))
+        ','.join(chain(excluded, map(str, (e for e in parsed.excluded if e > 0))))
     if parsed.viewbox:
         queryparts['viewbox'] = ','.join(f"{c:.7g}" for c in parsed.viewbox.coords)
     if parsed.bounded_viewbox:
@@ -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
-        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)
-        osm_ids_done.add(result.osm_object)
-        classification_done.add(classification)
         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+)[°\\s]+(?P<{axis}_min>[\\d.]+)?[′']*"
+    return f"(?P<{axis}_deg>\\d+)[°\\s]+(?P<{axis}_min>[\\d.]+)[′']*"
 
 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'),