]> git.openstreetmap.org Git - nominatim.git/commitdiff
use address counts for improving index lookup
authorSarah Hoffmann <lonvia@denofr.de>
Sat, 16 Mar 2024 15:56:04 +0000 (16:56 +0100)
committerSarah Hoffmann <lonvia@denofr.de>
Mon, 18 Mar 2024 10:25:48 +0000 (11:25 +0100)
nominatim/api/search/db_search_builder.py
nominatim/api/search/db_search_fields.py

index f8eabad1497c189f23e8c52d8e68ddacd12c10ef..97e7ac0282a79b40e7015bcd1069c8edbadeb09e 100644 (file)
@@ -227,28 +227,73 @@ class SearchBuilder:
         if name_fulls:
             fulls_count = sum(t.count for t in name_fulls)
             if len(name_partials) == 1:
-                penalty += min(1, max(0, (exp_count - 50 * fulls_count) / (1000 * fulls_count)))
-            # At this point drop unindexed partials from the address.
-            # This might yield wrong results, nothing we can do about that.
-            if not partials_indexed:
-                addr_tokens = [t.token for t in addr_partials if t.is_indexed]
+                penalty += min(0.5, max(0, (exp_count - 50 * fulls_count) / (2000 * fulls_count)))
+            if partials_indexed:
                 penalty += 1.2 * sum(t.penalty for t in addr_partials if not t.is_indexed)
-            # Any of the full names applies with all of the partials from the address
-            yield penalty, fulls_count / (2**len(addr_tokens)),\
-                  dbf.lookup_by_any_name([t.token for t in name_fulls],
-                                         addr_tokens,
-                                         fulls_count > 30000 / max(1, 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)))
 
         # To catch remaining results, lookup by name and address
         # We only do this if there is a reasonable number of results expected.
         exp_count = exp_count / (2**len(addr_tokens)) if addr_tokens else exp_count
         if exp_count < 10000 and all(t.is_indexed for t in name_partials.values()):
-            lookup = [dbf.FieldLookup('name_vector', list(name_partials.keys()), lookups.LookupAll)]
-            if addr_tokens:
-                lookup.append(dbf.FieldLookup('nameaddress_vector', addr_tokens, lookups.LookupAll))
             penalty += 0.35 * max(1 if name_fulls else 0.1,
                                   5 - len(name_partials) - len(addr_tokens))
-            yield penalty, exp_count, lookup
+            yield penalty, exp_count,\
+                  self.get_name_address_ranking(list(name_partials.keys()), addr_partials)
+
+
+    def get_name_address_ranking(self, name_tokens: List[int],
+                                 addr_partials: List[Token]) -> List[dbf.FieldLookup]:
+        """ Create a ranking expression looking up by name and address.
+        """
+        lookup = [dbf.FieldLookup('name_vector', name_tokens, lookups.LookupAll)]
+
+        addr_restrict_tokens = []
+        addr_lookup_tokens = []
+        for t in addr_partials:
+            if t.is_indexed:
+                if t.addr_count > 20000:
+                    addr_restrict_tokens.append(t.token)
+                else:
+                    addr_lookup_tokens.append(t.token)
+
+        if addr_restrict_tokens:
+            lookup.append(dbf.FieldLookup('nameaddress_vector',
+                                          addr_restrict_tokens, lookups.Restrict))
+        if addr_lookup_tokens:
+            lookup.append(dbf.FieldLookup('nameaddress_vector',
+                                          addr_lookup_tokens, lookups.LookupAll))
+
+        return lookup
+
+
+    def get_full_name_ranking(self, name_fulls: List[Token], addr_partials: List[Token],
+                              use_lookup: bool) -> List[dbf.FieldLookup]:
+        """ Create a ranking expression with full name terms and
+            additional address lookup. When 'use_lookup' is true, then
+            address lookups will use the index, when the occurences are not
+            too many.
+        """
+        # At this point drop unindexed partials from the address.
+        # This might yield wrong results, nothing we can do about that.
+        if use_lookup:
+            addr_restrict_tokens = []
+            addr_lookup_tokens = []
+            for t in addr_partials:
+                if t.is_indexed:
+                    if t.addr_count > 20000:
+                        addr_restrict_tokens.append(t.token)
+                    else:
+                        addr_lookup_tokens.append(t.token)
+        else:
+            addr_restrict_tokens = [t.token for t in addr_partials if t.is_indexed]
+            addr_lookup_tokens = []
+
+        return dbf.lookup_by_any_name([t.token for t in name_fulls],
+                                      addr_restrict_tokens, addr_lookup_tokens)
 
 
     def get_name_ranking(self, trange: TokenRange,
index cd5717753ba722616084dc06bbfb76dda901c0fb..7f775277e6960a7dc37d76a7181b5ad4e4273084 100644 (file)
@@ -231,16 +231,17 @@ def lookup_by_names(name_tokens: List[int], addr_tokens: List[int]) -> List[Fiel
     return lookup
 
 
-def lookup_by_any_name(name_tokens: List[int], addr_tokens: List[int],
-                       use_index_for_addr: bool) -> List[FieldLookup]:
+def lookup_by_any_name(name_tokens: List[int], addr_restrict_tokens: List[int],
+                       addr_lookup_tokens: List[int]) -> List[FieldLookup]:
     """ Create a lookup list where name tokens are looked up via index
         and only one of the name tokens must be present.
         Potential address tokens are used to restrict the search further.
     """
     lookup = [FieldLookup('name_vector', name_tokens, lookups.LookupAny)]
-    if addr_tokens:
-        lookup.append(FieldLookup('nameaddress_vector', addr_tokens,
-                                  lookups.LookupAll if use_index_for_addr else lookups.Restrict))
+    if addr_restrict_tokens:
+        lookup.append(FieldLookup('nameaddress_vector', addr_restrict_tokens, lookups.Restrict))
+    if addr_lookup_tokens:
+        lookup.append(FieldLookup('nameaddress_vector', addr_lookup_tokens, lookups.LookupAll))
 
     return lookup