safety-screen/Nominatim
1
0
Fork 0
forked from hans/Nominatim
Code Pull Requests Activity

do not write any word counts on initial word insert

Browse Source
This commit is contained in:
Sarah Hoffmann
2025-03-20 21:33:27 +01:00
parent 1705bb5f57
commit 58ef032a2b
1 changed files with 5 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
lib-sql/tokenizer/icu_tokenizer.sql
View File

@@ -128,16 +128,14 @@ DECLARE
partial_terms TEXT[] = '{}'::TEXT[]; partial_terms TEXT[] = '{}'::TEXT[];
term TEXT; term TEXT;
term_id INTEGER; term_id INTEGER;
term_count INTEGER;
BEGIN BEGIN
SELECT min(word_id) INTO full_token SELECT min(word_id) INTO full_token
FROM word WHERE word = norm_term and type = 'W'; FROM word WHERE word = norm_term and type = 'W';
IF full_token IS NULL THEN IF full_token IS NULL THEN
full_token := nextval('seq_word'); full_token := nextval('seq_word');
INSERT INTO word (word_id, word_token, type, word, info) INSERT INTO word (word_id, word_token, type, word)
SELECT full_token, lookup_term, 'W', norm_term, SELECT full_token, lookup_term, 'W', norm_term
json_build_object('count', 0)
FROM unnest(lookup_terms) as lookup_term; FROM unnest(lookup_terms) as lookup_term;
END IF; END IF;
@@ -150,14 +148,13 @@ BEGIN
partial_tokens := '{}'::INT[]; partial_tokens := '{}'::INT[];
FOR term IN SELECT unnest(partial_terms) LOOP FOR term IN SELECT unnest(partial_terms) LOOP
SELECT min(word_id), max(info->>'count') INTO term_id, term_count SELECT min(word_id) INTO term_id
FROM word WHERE word_token = term and type = 'w'; FROM word WHERE word_token = term and type = 'w';
IF term_id IS NULL THEN IF term_id IS NULL THEN
term_id := nextval('seq_word'); term_id := nextval('seq_word');
term_count := 0; INSERT INTO word (word_id, word_token, type)
INSERT INTO word (word_id, word_token, type, info) VALUES (term_id, term, 'w');
VALUES (term_id, term, 'w', json_build_object('count', term_count));
END IF; END IF;
partial_tokens := array_merge(partial_tokens, ARRAY[term_id]); partial_tokens := array_merge(partial_tokens, ARRAY[term_id]);