osm-search/Nominatim
1
1
Fork 0
mirror of https://github.com/osm-search/Nominatim.git synced 2024-11-23 05:35:13 +03:00
Code Issues Packages Projects Releases Wiki Activity

remove country restriction from tokenizer

Browse Source 
Restricting tokens due to the search context is better done in
the generic search part instead of repeating the same test in
every tokenizer implementation.
This commit is contained in:
Sarah Hoffmann 2021-07-29 21:25:59 +02:00
parent 78d11fe628
commit 0fb8eade13
5 changed files with 21 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
lib-php/Geocode.php
View File

@ -498,7 +498,6 @@ class Geocode
if ($this->aCountryCodes) {
$oCtx->setCountryList($this->aCountryCodes);
}
$this->oTokenizer->setCountryRestriction($this->aCountryCodes);
Debug::newSection('Query Preprocessing');

16
lib-php/SearchContext.php
View File

@ -28,6 +28,8 @@ class SearchContext
public $sqlViewboxLarge = '';
/// Reference along a route (as SQL).
public $sqlViewboxCentre = '';
/// List of countries to restrict search to (as array).
public $aCountryList = null;
/// List of countries to restrict search to (as SQL).
public $sqlCountryList = '';
/// List of place IDs to exclude (as SQL).
@ -187,6 +189,7 @@ class SearchContext
public function setCountryList($aCountries)
{
$this->sqlCountryList = '('.join(',', array_map('addQuotes', $aCountries)).')';
$this->aCountryList = $aCountries;
}
/**
@ -279,6 +282,19 @@ class SearchContext
return '';
}
/**
* Check if the given country is covered by the search context.
*
* @param string $sCountryCode Country code of the country to check.
*
* @return True, if no country code restrictions are set or the
* country is included in the country list.
*/
public function isCountryApplicable($sCountryCode)
{
return $this->aCountryList === null || in_array($sCountryCode, $this->aCountryList);
}
public function debugInfo()
{
return array(

4
lib-php/TokenCountry.php
View File

@ -36,7 +36,9 @@ class Country
*/
public function isExtendable($oSearch, $oPosition)
{
return !$oSearch->hasCountry() && $oPosition->maybePhrase('country');
return !$oSearch->hasCountry()
&& $oPosition->maybePhrase('country')
&& $oSearch->getContext()->isCountryApplicable($this->sCountryCode);
}
/**

12
lib-php/tokenizer/legacy_icu_tokenizer.php
View File

@ -8,7 +8,6 @@ class Tokenizer
private $oNormalizer;
private $oTransliterator;
private $aCountryRestriction;
public function __construct(&$oDB)
{
@ -30,12 +29,6 @@ class Tokenizer
}
public function setCountryRestriction($aCountries)
{
$this->aCountryRestriction = $aCountries;
}
public function normalizeString($sTerm)
{
if ($this->oNormalizer === null) {
@ -162,10 +155,7 @@ class Tokenizer
switch ($aWord['type']) {
case 'C': // country name tokens
if ($aWord['word'] !== null
&& (!$this->aCountryRestriction
|| in_array($aWord['word'], $this->aCountryRestriction))
) {
if ($aWord['word'] !== null) {
$oValidTokens->addToken(
$sTok,
new Token\Country($iId, $aWord['word'])

14
lib-php/tokenizer/legacy_tokenizer.php
View File

@ -7,7 +7,6 @@ class Tokenizer
private $oDB;
private $oNormalizer = null;
private $aCountryRestriction = null;
public function __construct(&$oDB)
{
@ -37,12 +36,6 @@ class Tokenizer
}
public function setCountryRestriction($aCountries)
{
$this->aCountryRestriction = $aCountries;
}
public function normalizeString($sTerm)
{
if ($this->oNormalizer === null) {
@ -206,12 +199,7 @@ class Tokenizer
);
}
} elseif ($aWord['country_code']) {
// Filter country tokens that do not match restricted countries.
if (!$this->aCountryRestriction
|| in_array($aWord['country_code'], $this->aCountryRestriction)
) {
$oToken = new Token\Country($iId, $aWord['country_code']);
}
$oToken = new Token\Country($iId, $aWord['country_code']);
} elseif ($aWord['word_token'][0] == ' ') {
$oToken = new Token\Word(
$iId,