From: Sarah Hoffmann <lonvia@denofr.de>
Date: Tue, 17 Nov 2020 17:03:33 +0000 (+0100)
Subject: split addr: tags into words before adding to the search index
X-Git-Tag: v3.6.0~29^2
X-Git-Url: https://git.openstreetmap.org./nominatim.git/commitdiff_plain/30a6b6bdac5806051741b771d5ad8ad2c976aac9?hp=cc345f531ac81b3ebc85452accc433a753c61868

split addr: tags into words before adding to the search index

Address parts are only matched by single partial words. If
the addr: names are not split, then multi-word names cannot
be found.
---

diff --git a/sql/functions/normalization.sql b/sql/functions/normalization.sql
index b2fecab7..a5560cb9 100644
--- a/sql/functions/normalization.sql
+++ b/sql/functions/normalization.sql
@@ -207,19 +207,31 @@ CREATE OR REPLACE FUNCTION addr_ids_from_name(lookup_word TEXT)
   RETURNS INTEGER[]
   AS $$
 DECLARE
-  lookup_token TEXT;
+  words TEXT[];
   id INTEGER;
   return_word_id INTEGER[];
+  word_ids INTEGER[];
+  j INTEGER;
 BEGIN
-  lookup_token := make_standard_name(lookup_word);
-  SELECT array_agg(word_id) FROM word
-    WHERE word_token = lookup_token and class is null and type is null
-    INTO return_word_id;
-  IF return_word_id IS NULL THEN
-    id := nextval('seq_word');
-    INSERT INTO word VALUES (id, lookup_token, null, null, null, null, 0);
-    return_word_id = ARRAY[id];
+  words := string_to_array(make_standard_name(lookup_word), ' ');
+  IF array_upper(words, 1) IS NOT NULL THEN
+    FOR j IN 1..array_upper(words, 1) LOOP
+      IF (words[j] != '') THEN
+        SELECT array_agg(word_id) INTO word_ids
+          FROM word
+         WHERE word_token = words[j] and class is null and type is null;
+
+        IF word_ids IS NULL THEN
+          id := nextval('seq_word');
+          INSERT INTO word VALUES (id, words[j], null, null, null, null, 0);
+          return_word_id := return_word_id || id;
+        ELSE
+          return_word_id := array_merge(return_word_id, word_ids);
+        END IF;
+      END IF;
+    END LOOP;
   END IF;
+
   RETURN return_word_id;
 END;
 $$
diff --git a/test/bdd/db/import/search_name.feature b/test/bdd/db/import/search_name.feature
index 866a597d..beae3444 100644
--- a/test/bdd/db/import/search_name.feature
+++ b/test/bdd/db/import/search_name.feature
@@ -30,7 +30,24 @@ Feature: Creation of search terms
          | osm_type | osm_id | name |
          | N        | 1      | 23, Rose Street |
 
-    Scenario: Unnamed POI has no search entry when it has known addr: tags
+    Scenario: Searching for unknown addr: tags also works for multiple words
+        Given the scene roads-with-pois
+        And the places
+         | osm | class   | type        | housenr | addr+city        | geometry |
+         | N1  | place   | house       | 23      | Little Big Town  | :p-N1 |
+        And the places
+         | osm | class   | type        | name+name   | geometry |
+         | W1  | highway | residential | Rose Street | :w-north |
+        When importing
+        Then search_name contains
+         | object | name_vector | nameaddress_vector |
+         | N1     | #23         | Rose Street, Little, Big, Town |
+        When searching for "23 Rose Street, Little Big Town"
+        Then results contain
+         | osm_type | osm_id | name |
+         | N        | 1      | 23, Rose Street |
+
+     Scenario: Unnamed POI has no search entry when it has known addr: tags
         Given the scene roads-with-pois
         And the places
          | osm | class   | type        | housenr | addr+city | geometry |
@@ -197,7 +214,7 @@ Feature: Creation of search terms
         When importing
         Then search_name contains
          | object | nameaddress_vector |
-         | W1     | bonn, new york, smalltown |
+         | W1     | bonn, new, york, smalltown |
 
     Scenario: A known addr:* tag is added even if the name is unknown
         Given the scene roads-with-pois