Merge pull request #3342 from mtmail/tyops

[nominatim.git] / lib-sql / tokenizer / legacy_tokenizer.sql

diff --git a/lib-sql/tokenizer/legacy_tokenizer.sql b/lib-sql/tokenizer/legacy_tokenizer.sql
index a2c6b52073ec007e052b7775a148f6159fa1239d..c21d0510429910a94403203bfd2acf77df52fce1 100644 (file)
--- a/lib-sql/tokenizer/legacy_tokenizer.sql
+++ b/lib-sql/tokenizer/legacy_tokenizer.sql
@@ -1,3 +1,10 @@
+-- SPDX-License-Identifier: GPL-2.0-only
+--
+-- This file is part of Nominatim. (https://nominatim.org)
+--
+-- Copyright (C) 2022 by the Nominatim developer community.
+-- For a full list of authors see the git log.
+
 -- Get tokens used for searching the given place.
 --
 -- These are the tokens that will be saved in the search_name table.
@@ -34,17 +41,38 @@ AS $$
 $$ LANGUAGE SQL IMMUTABLE STRICT;
 
 
-CREATE OR REPLACE FUNCTION token_addr_street_match_tokens(info JSONB)
-  RETURNS INTEGER[]
+CREATE OR REPLACE FUNCTION token_is_street_address(info JSONB)
+  RETURNS BOOLEAN
+AS $$
+  SELECT info->>'street' is not null or info->>'place_search' is null;
+$$ LANGUAGE SQL IMMUTABLE;
+
+
+CREATE OR REPLACE FUNCTION token_has_addr_street(info JSONB)
+  RETURNS BOOLEAN
+AS $$
+  SELECT info->>'street' is not null and info->>'street' != '{}';
+$$ LANGUAGE SQL IMMUTABLE;
+
+
+CREATE OR REPLACE FUNCTION token_has_addr_place(info JSONB)
+  RETURNS BOOLEAN
+AS $$
+  SELECT info->>'place_match' is not null;
+$$ LANGUAGE SQL IMMUTABLE;
+
+
+CREATE OR REPLACE FUNCTION token_matches_street(info JSONB, street_tokens INTEGER[])
+  RETURNS BOOLEAN
 AS $$
-  SELECT (info->>'street')::INTEGER[]
+  SELECT (info->>'street')::INTEGER[] && street_tokens
 $$ LANGUAGE SQL IMMUTABLE STRICT;
 
 
-CREATE OR REPLACE FUNCTION token_addr_place_match_tokens(info JSONB)
-  RETURNS INTEGER[]
+CREATE OR REPLACE FUNCTION token_matches_place(info JSONB, place_tokens INTEGER[])
+  RETURNS BOOLEAN
 AS $$
-  SELECT (info->>'place_match')::INTEGER[]
+  SELECT (info->>'place_match')::INTEGER[] && place_tokens
 $$ LANGUAGE SQL IMMUTABLE STRICT;
 
 
@@ -55,26 +83,31 @@ AS $$
 $$ LANGUAGE SQL IMMUTABLE STRICT;
 
 
-DROP TYPE IF EXISTS token_addresstoken CASCADE;
-CREATE TYPE token_addresstoken AS (
-  key TEXT,
-  match_tokens INT[],
-  search_tokens INT[]
-);
+CREATE OR REPLACE FUNCTION token_get_address_keys(info JSONB)
+  RETURNS SETOF TEXT
+AS $$
+  SELECT * FROM jsonb_object_keys(info->'addr');
+$$ LANGUAGE SQL IMMUTABLE STRICT;
+
+
+CREATE OR REPLACE FUNCTION token_get_address_search_tokens(info JSONB, key TEXT)
+  RETURNS INTEGER[]
+AS $$
+  SELECT (info->'addr'->key->>0)::INTEGER[];
+$$ LANGUAGE SQL IMMUTABLE STRICT;
+
 
-CREATE OR REPLACE FUNCTION token_get_address_tokens(info JSONB)
-  RETURNS SETOF token_addresstoken
+CREATE OR REPLACE FUNCTION token_matches_address(info JSONB, key TEXT, tokens INTEGER[])
+  RETURNS BOOLEAN
 AS $$
-  SELECT key, (value->>1)::int[] as match_tokens,
-         (value->>0)::int[] as search_tokens
-  FROM jsonb_each(info->'addr');
+  SELECT (info->'addr'->key->>1)::INTEGER[] && tokens;
 $$ LANGUAGE SQL IMMUTABLE STRICT;
 
 
-CREATE OR REPLACE FUNCTION token_normalized_postcode(postcode TEXT)
+CREATE OR REPLACE FUNCTION token_get_postcode(info JSONB)
   RETURNS TEXT
 AS $$
-  SELECT CASE WHEN postcode SIMILAR TO '%(,|;)%' THEN NULL ELSE upper(trim(postcode))END;
+  SELECT info->>'postcode';
 $$ LANGUAGE SQL IMMUTABLE STRICT;
 
 
@@ -314,7 +347,7 @@ BEGIN
       END LOOP;
     END IF;
 
-    -- consider parts before an opening braket a full word as well
+    -- consider parts before an opening bracket a full word as well
     words := regexp_split_to_array(value, E'[(]');
     IF array_upper(words, 1) > 1 THEN
       s := make_standard_name(words[1]);