Merge remote-tracking branch 'upstream/master'

[nominatim.git] / lib-sql / functions / normalization.sql

diff --git a/lib-sql/functions/normalization.sql b/lib-sql/functions/normalization.sql
index 0300dab401a96323112060f032830496c4151792..f283f9165d9e324590ecbdaaa5fc9fc872651e16 100644 (file)
--- a/lib-sql/functions/normalization.sql
+++ b/lib-sql/functions/normalization.sql
@@ -47,6 +47,25 @@ END;
 $$
 LANGUAGE plpgsql;
 
+-- Create housenumber tokens from an OSM addr:housenumber.
+-- The housnumber is split at comma and semicolon as necessary.
+-- The function returns the normalized form of the housenumber suitable
+-- for comparison.
+CREATE OR REPLACE FUNCTION create_housenumber_id(housenumber TEXT)
+  RETURNS TEXT
+  AS $$
+DECLARE
+  normtext TEXT;
+BEGIN
+  SELECT array_to_string(array_agg(trans), ';')
+    INTO normtext
+    FROM (SELECT lookup_word as trans, getorcreate_housenumber_id(lookup_word)
+          FROM (SELECT make_standard_name(h) as lookup_word
+                FROM regexp_split_to_table(housenumber, '[,;]') h) x) y;
+
+  return normtext;
+END;
+$$ LANGUAGE plpgsql STABLE STRICT;
 
 CREATE OR REPLACE FUNCTION getorcreate_housenumber_id(lookup_word TEXT)
   RETURNS INTEGER
@@ -118,7 +137,7 @@ $$
 LANGUAGE plpgsql;
 
 
-CREATE OR REPLACE FUNCTION getorcreate_amenity(lookup_word TEXT,
+CREATE OR REPLACE FUNCTION getorcreate_amenity(lookup_word TEXT, normalized_word TEXT,
                                                lookup_class text, lookup_type text)
   RETURNS INTEGER
   AS $$
@@ -128,12 +147,12 @@ DECLARE
 BEGIN
   lookup_token := ' '||trim(lookup_word);
   SELECT min(word_id) FROM word
-  WHERE word_token = lookup_token and word = lookup_word
+  WHERE word_token = lookup_token and word = normalized_word
         and class = lookup_class and type = lookup_type
   INTO return_word_id;
   IF return_word_id IS NULL THEN
     return_word_id := nextval('seq_word');
-    INSERT INTO word VALUES (return_word_id, lookup_token, lookup_word,
+    INSERT INTO word VALUES (return_word_id, lookup_token, normalized_word,
                              lookup_class, lookup_type, null, 0);
   END IF;
   RETURN return_word_id;
@@ -143,6 +162,7 @@ LANGUAGE plpgsql;
 
 
 CREATE OR REPLACE FUNCTION getorcreate_amenityoperator(lookup_word TEXT,
+                                                       normalized_word TEXT,
                                                        lookup_class text,
                                                        lookup_type text,
                                                        op text)
@@ -154,12 +174,12 @@ DECLARE
 BEGIN
   lookup_token := ' '||trim(lookup_word);
   SELECT min(word_id) FROM word
-  WHERE word_token = lookup_token and word = lookup_word
+  WHERE word_token = lookup_token and word = normalized_word
         and class = lookup_class and type = lookup_type and operator = op
   INTO return_word_id;
   IF return_word_id IS NULL THEN
     return_word_id := nextval('seq_word');
-    INSERT INTO word VALUES (return_word_id, lookup_token, lookup_word,
+    INSERT INTO word VALUES (return_word_id, lookup_token, normalized_word,
                              lookup_class, lookup_type, null, 0, op);
   END IF;
   RETURN return_word_id;