hans/Nominatim
1
0
Fork 1
mirror of https://github.com/osm-search/Nominatim.git synced 2026-03-07 02:24:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

add normalized version of special search terms on import

Browse Source 
Requires the PHP bindings for libicu, so add that as a requirement.
This commit is contained in:
Sarah Hoffmann
2017-06-01 19:34:16 +02:00
parent 2dd7f0156a
commit e3fb706c65
6 changed files with 15 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
docs/Installation.md
View File

@@ -39,6 +39,7 @@ For running Nominatim:
* [PostGIS](http://postgis.refractions.net) (2.0 or later) * [PostGIS](http://postgis.refractions.net) (2.0 or later)
* [PHP](http://php.net) (5.4 or later) * [PHP](http://php.net) (5.4 or later)
* PHP-pgsql * PHP-pgsql
* PHP-intl (bundled with PHP)
* [PEAR::DB](http://pear.php.net/package/DB) * [PEAR::DB](http://pear.php.net/package/DB)
* a webserver (apache or nginx are recommended) * a webserver (apache or nginx are recommended)

12
sql/functions.sql
View File

@@ -101,7 +101,7 @@ END;
$$ $$
LANGUAGE plpgsql; LANGUAGE plpgsql;
CREATE OR REPLACE FUNCTION getorcreate_amenity(lookup_word TEXT, lookup_class text, lookup_type text) CREATE OR REPLACE FUNCTION getorcreate_amenity(lookup_word TEXT, normalized_word TEXT, lookup_class text, lookup_type text)
RETURNS INTEGER RETURNS INTEGER
AS $$ AS $$
DECLARE DECLARE
@@ -109,17 +109,17 @@ DECLARE
return_word_id INTEGER; return_word_id INTEGER;
BEGIN BEGIN
lookup_token := ' '||trim(lookup_word); lookup_token := ' '||trim(lookup_word);
SELECT min(word_id) FROM word WHERE word_token = lookup_token and class=lookup_class and type = lookup_type into return_word_id; SELECT min(word_id) FROM 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 IF return_word_id IS NULL THEN
return_word_id := nextval('seq_word'); return_word_id := nextval('seq_word');
INSERT INTO word VALUES (return_word_id, lookup_token, null, lookup_class, lookup_type, null, 0); INSERT INTO word VALUES (return_word_id, lookup_token, normalized_word, lookup_class, lookup_type, null, 0);
END IF; END IF;
RETURN return_word_id; RETURN return_word_id;
END; END;
$$ $$
LANGUAGE plpgsql; LANGUAGE plpgsql;
CREATE OR REPLACE FUNCTION getorcreate_amenityoperator(lookup_word TEXT, lookup_class text, lookup_type text, op text) CREATE OR REPLACE FUNCTION getorcreate_amenityoperator(lookup_word TEXT, normalized_word TEXT, lookup_class text, lookup_type text, op text)
RETURNS INTEGER RETURNS INTEGER
AS $$ AS $$
DECLARE DECLARE
@@ -127,10 +127,10 @@ DECLARE
return_word_id INTEGER; return_word_id INTEGER;
BEGIN BEGIN
lookup_token := ' '||trim(lookup_word); lookup_token := ' '||trim(lookup_word);
SELECT min(word_id) FROM word WHERE word_token = lookup_token and class=lookup_class and type = lookup_type and operator = op into return_word_id; SELECT min(word_id) FROM 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 IF return_word_id IS NULL THEN
return_word_id := nextval('seq_word'); return_word_id := nextval('seq_word');
INSERT INTO word VALUES (return_word_id, lookup_token, null, lookup_class, lookup_type, null, 0, op); INSERT INTO word VALUES (return_word_id, lookup_token, normalized_word, lookup_class, lookup_type, null, 0, op);
END IF; END IF;
RETURN return_word_id; RETURN return_word_id;
END; END;

8
utils/specialphrases.php
View File

@@ -19,6 +19,7 @@ getCmdOpt($_SERVER['argv'], $aCMDOptions, $aCMDResult, true, true);
include(CONST_InstallPath.'/settings/phrase_settings.php'); include(CONST_InstallPath.'/settings/phrase_settings.php');
if ($aCMDResult['wiki-import']) { if ($aCMDResult['wiki-import']) {
$oNormalizer = Transliterator::createFromRules(":: NFD (); [:Nonspacing Mark:] >; :: lower (); [[:Punctuation:][:Space:]]+ > ' '; :: NFC ();");
$aPairs = array(); $aPairs = array();
$sLanguageIn = CONST_Languages ? CONST_Languages : $sLanguageIn = CONST_Languages ? CONST_Languages :
@@ -31,6 +32,7 @@ if ($aCMDResult['wiki-import']) {
if (preg_match_all('#\\| ([^|]+) \\|\\| ([^|]+) \\|\\| ([^|]+) \\|\\| ([^|]+) \\|\\| ([\\-YN])#', $sWikiPageXML, $aMatches, PREG_SET_ORDER)) { if (preg_match_all('#\\| ([^|]+) \\|\\| ([^|]+) \\|\\| ([^|]+) \\|\\| ([^|]+) \\|\\| ([\\-YN])#', $sWikiPageXML, $aMatches, PREG_SET_ORDER)) {
foreach ($aMatches as $aMatch) { foreach ($aMatches as $aMatch) {
$sLabel = trim($aMatch[1]); $sLabel = trim($aMatch[1]);
$sTrans = pg_escape_string($oNormalizer->transliterate($sLabel));
$sClass = trim($aMatch[2]); $sClass = trim($aMatch[2]);
$sType = trim($aMatch[3]); $sType = trim($aMatch[3]);
// hack around a bug where building=yes was imported with // hack around a bug where building=yes was imported with
@@ -57,13 +59,13 @@ if ($aCMDResult['wiki-import']) {
switch (trim($aMatch[4])) { switch (trim($aMatch[4])) {
case 'near': case 'near':
echo "select getorcreate_amenityoperator(make_standard_name('".pg_escape_string($sLabel)."'), '$sClass', '$sType', 'near');\n"; echo "select getorcreate_amenityoperator(make_standard_name('".pg_escape_string($sLabel)."'), '$sTrans', '$sClass', '$sType', 'near');\n";
break; break;
case 'in': case 'in':
echo "select getorcreate_amenityoperator(make_standard_name('".pg_escape_string($sLabel)."'), '$sClass', '$sType', 'in');\n"; echo "select getorcreate_amenityoperator(make_standard_name('".pg_escape_string($sLabel)."'), '$sTrans', '$sClass', '$sType', 'in');\n";
break; break;
default: default:
echo "select getorcreate_amenity(make_standard_name('".pg_escape_string($sLabel)."'), '$sClass', '$sType');\n"; echo "select getorcreate_amenity(make_standard_name('".pg_escape_string($sLabel)."'), '$sTrans', '$sClass', '$sType');\n";
break; break;
} }
} }

2
vagrant/install-on-centos-7.sh
View File

@@ -21,7 +21,7 @@
sudo yum install -y postgresql-server postgresql-contrib postgresql-devel postgis postgis-utils \ sudo yum install -y postgresql-server postgresql-contrib postgresql-devel postgis postgis-utils \
git cmake make gcc gcc-c++ libtool policycoreutils-python \ git cmake make gcc gcc-c++ libtool policycoreutils-python \
php-pgsql php php-pear php-pear-DB libpqxx-devel proj-epsg \ php-pgsql php php-pear php-pear-DB php-intl libpqxx-devel proj-epsg \
bzip2-devel proj-devel geos-devel libxml2-devel boost-devel expat-devel zlib-devel bzip2-devel proj-devel geos-devel libxml2-devel boost-devel expat-devel zlib-devel
# If you want to run the test suite, you need to install the following # If you want to run the test suite, you need to install the following

2
vagrant/install-on-travis-ci.sh
View File

@@ -16,7 +16,7 @@ sudo apt-get install -y -qq libboost-dev libboost-system-dev \
libboost-filesystem-dev libexpat1-dev zlib1g-dev libxml2-dev\ libboost-filesystem-dev libexpat1-dev zlib1g-dev libxml2-dev\
libbz2-dev libpq-dev libgeos-c1 libgeos++-dev libproj-dev \ libbz2-dev libpq-dev libgeos-c1 libgeos++-dev libproj-dev \
postgresql-server-dev-9.6 postgresql-9.6-postgis-2.3 postgresql-contrib-9.6 \ postgresql-server-dev-9.6 postgresql-9.6-postgis-2.3 postgresql-contrib-9.6 \
apache2 php5 php5-pgsql php-pear php-db apache2 php5 php5-pgsql php-pear php-db php-intl
sudo apt-get install -y -qq python3-dev python3-pip python3-psycopg2 phpunit php5-cgi sudo apt-get install -y -qq python3-dev python3-pip python3-psycopg2 phpunit php5-cgi

2
vagrant/install-on-ubuntu-16.sh
View File

@@ -28,7 +28,7 @@ export DEBIAN_FRONTEND=noninteractive #DOCS:
libbz2-dev libpq-dev libgeos-dev libgeos++-dev libproj-dev \ libbz2-dev libpq-dev libgeos-dev libgeos++-dev libproj-dev \
postgresql-server-dev-9.5 postgresql-9.5-postgis-2.2 postgresql-contrib-9.5 \ postgresql-server-dev-9.5 postgresql-9.5-postgis-2.2 postgresql-contrib-9.5 \
apache2 php php-pgsql libapache2-mod-php php-pear php-db \ apache2 php php-pgsql libapache2-mod-php php-pear php-db \
git php-intl git
# If you want to run the test suite, you need to install the following # If you want to run the test suite, you need to install the following
# additional packages: # additional packages: