osm-search/Nominatim
1
1
Fork 0
mirror of https://github.com/osm-search/Nominatim.git synced 2024-09-19 15:08:31 +03:00
Code Issues Packages Projects Releases Wiki Activity

switch housenumber tokens to new word table layout

Browse Source
This commit is contained in:
Sarah Hoffmann 2021-07-20 11:36:20 +02:00
parent 1618aba5f2
commit 5ab0a63fd6
3 changed files with 12 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
lib-php/tokenizer/legacy_icu_tokenizer.php
View File

@ -156,6 +156,8 @@ class Tokenizer
$aDBWords = $this->oDB->getAll($sSQL, null, 'Could not get word tokens.');
foreach ($aDBWords as $aWord) {
$iId = (int) $aWord['word_id'];
switch ($aWord['type']) {
'C': // country name tokens
if ($aWord['country'] === null
@ -166,12 +168,13 @@ class Tokenizer
}
$oToken = new Token\Country($iId, $aWord['country'])
break;
'H': // house number tokens
$oToken = new Token\HouseNumber($iId, $aWord['word_token']);
break;
default:
continue;
}
/* $iId = (int) $aWord['word_id'];
if ($aWord['class']) {
/* if ($aWord['class']) {
// Special terms need to appear in their normalized form.
// (postcodes are not normalized in the word table)
$sNormWord = $this->normalizeString($aWord['word']);

10
lib-sql/tokenizer/legacy_icu_tokenizer.sql
View File

@ -140,15 +140,13 @@ CREATE OR REPLACE FUNCTION getorcreate_hnr_id(lookup_term TEXT)
DECLARE
return_id INTEGER;
BEGIN
SELECT min(word_id) INTO return_id
FROM word
WHERE word_token = ' ' || lookup_term
and class = 'place' and type = 'house';
SELECT min(word_id) INTO return_id FROM word
WHERE word_token = lookup_term and type = 'H';
IF return_id IS NULL THEN
return_id := nextval('seq_word');
INSERT INTO word (word_id, word_token, class, type, search_name_count)
VALUES (return_id, ' ' || lookup_term, 'place', 'house', 0);
INSERT INTO word (word_id, word_token, type)
VALUES (return_id, lookup_term, 'H');
END IF;
RETURN return_id;

3
nominatim/tokenizer/legacy_icu_tokenizer.py
View File

@ -601,7 +601,8 @@ class _TokenCache:
def get_hnr_tokens(self, conn, terms):
""" Get token ids for a list of housenumbers, looking them up in the
database if necessary.
database if necessary. `terms` is an iterable of normalized
housenumbers.
"""
tokens = []
askdb = []