use new transliteration in initial housenumber word computation

The new create_housenumber_id() function splits housenumber
lists correctly. Otherwise there is no difference.
This commit is contained in:
Sarah Hoffmann 2021-04-01 18:52:06 +02:00
parent 55fcc44c8c
commit 6cbef84cad
3 changed files with 7 additions and 4 deletions

View File

@ -47,7 +47,10 @@ END;
$$
LANGUAGE plpgsql;
-- Create housenumber tokens from an OSM addr:housenumber.
-- The housnumber is split at comma and semicolon as necessary.
-- The function returns the normalized form of the housenumber suitable
-- for comparison.
CREATE OR REPLACE FUNCTION create_housenumber_id(housenumber TEXT)
RETURNS TEXT
AS $$
@ -56,7 +59,7 @@ DECLARE
BEGIN
SELECT array_to_string(array_agg(trans), ';')
INTO normtext
FROM (SELECT transliteration(lookup_word) as trans, getorcreate_housenumber_id(lookup_word)
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;

View File

@ -667,7 +667,7 @@ BEGIN
IF NEW.address is not NULL THEN
IF NEW.address ? 'conscriptionnumber' THEN
IF NEW.address ? 'streetnumber' THEN
NEW.housenumber := (NEW.address->'conscriptionnumber') || ';' || (NEW.address->'streetnumber');
NEW.housenumber := (NEW.address->'conscriptionnumber') || '/' || (NEW.address->'streetnumber');
ELSE
NEW.housenumber := NEW.address->'conscriptionnumber';
END IF;

View File

@ -5,7 +5,7 @@ CREATE TABLE word_frequencies AS
GROUP BY id);
select count(getorcreate_postcode_id(v)) from (select distinct address->'postcode' as v from place where address ? 'postcode') as w where v is not null;
select count(getorcreate_housenumber_id(make_standard_name(v))) from (select distinct address->'housenumber' as v from place where address ? 'housenumber') as w;
select count(create_housenumber_id(v)) from (select distinct address->'housenumber' as v from place where address ? 'housenumber') as w;
-- copy the word frequencies
update word set search_name_count = count from word_frequencies wf where wf.id = word.word_id;