tonyair
/
silverstripe-fulltextsearch
mirror of https://github.com/silverstripe/silverstripe-fulltextsearch
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

ENHANCEMENT Per-field analyser support for SolrIndex

This commit is contained in:
Ingo Schommer 2012-09-03 22:53:23 +02:00
parent b6fd7b7b89
commit 0ef78f905c
3 changed files with 68 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
code/solr/SolrIndex.php
View File

@ -24,6 +24,8 @@ abstract class SolrIndex extends SearchIndex {
static $sortTypeMap = array();
protected $analyzerFields = array();
function generateSchema() {
return $this->renderWith(Director::baseFolder() . '/fulltextsearch/conf/templates/schema.ss');
}
@ -36,6 +38,24 @@ abstract class SolrIndex extends SearchIndex {
return $this->renderWith(Director::baseFolder() . '/fulltextsearch/conf/templates/types.ss');
}
/**
* Index-time analyzer which is applied to a specific field.
* Can be used to remove HTML tags, apply stemming, etc.
*
* @see http://wiki.apache.org/solr/AnalyzersTokenizersTokenFilters#solr.WhitespaceTokenizerFactory
*
* @param String $field
* @param String $type
* @param Array $params Parameters for the analyzer, usually at least a "class"
*/
function addAnalyzer($field, $type, $params) {
$fullFields = $this->fieldData($field);
if($fullFields) foreach($fullFields as $fullField => $spec) {
if(!isset($this->analyzerFields[$fullField])) $this->analyzerFields[$fullField] = array();
$this->analyzerFields[$fullField][$type] = $params;
}
}
function getFieldDefinitions() {
$xml = array();
$stored = Director::isDev() ? "stored='true'" : "stored='false'";
@ -84,6 +104,13 @@ abstract class SolrIndex extends SearchIndex {
$multiValued = (isset($spec['multi_valued']) && $spec['multi_valued']) ? "true" : '';
$type = isset($typeMap[$spec['type']]) ? $typeMap[$spec['type']] : $typeMap['*'];
$analyzerXml = '';
if(isset($this->analyzerFields[$name])) {
foreach($this->analyzerFields[$name] as $analyzerType => $analyzerParams) {
$analyzerXml .= $this->toXmlTag($analyzerType, $analyzerParams);
}
}
$fieldParams = array_merge(
array(
'name' => $name,
@ -97,7 +124,8 @@ abstract class SolrIndex extends SearchIndex {
return $this->toXmlTag(
"field",
$fieldParams
$fieldParams,
$analyzerXml ? "<analyzer>$analyzerXml</analyzer>" : null
);
}

23
docs/Solr.md
View File

@ -76,6 +76,27 @@ You can also copy the `thirdparty/`solr directory somewhere else,
just set the path value in `mysite/_config.php` to point to the new location.
And of course run `java -jar start.jar` from the new directory.
### Adding Analyzers, Tokenizers and Token Filters
When a document is indexed, its individual fields are subject to the analyzing and tokenizing filters that can transform and normalize the data in the fields. For example — removing blank spaces, removing html code, stemming, removing a particular character and replacing it with another
(see [Solr Wiki](http://wiki.apache.org/solr/AnalyzersTokenizersTokenFilters)).
Example: Replace synonyms on indexing (e.g. "i-pad" with "iPad")
<?php
class MyIndex extends SolrIndex {
function init() {
$this->addClass('Page');
$this->addField('Content');
$this->addAnalyzer('Content', 'filter', array('class' => 'solr.SynonymFilterFactory'));
}
}
// Generates the following XML schema definition:
// <field name="Page_Content" ...>
// <filter class="solr.SynonymFilterFactory" synonyms="syn.txt" ignoreCase="true" expand="false"/>
// </field>
## Debugging
### Using the web admin interface
@ -103,4 +124,4 @@ In order to query the field, reverse the search conditions and exclude the range
// Wrong: Filter will ignore all empty field values
$myQuery->filter(<field>, new SearchQuery_Range('*', <date>));
// Better: Exclude the opposite range
$myQuery->exclude(<field>, new SearchQuery_Range(<date>, '*'));
$myQuery->exclude(<field>, new SearchQuery_Range(<date>, '*'));

17
tests/SolrIndexTest.php
View File

@ -65,6 +65,23 @@ class SolrIndexTest extends SapphireTest {
Director::set_environment_type($origMode);
}
function testAddAnalyzer() {
$index = new SolrIndexTest_FakeIndex();
$defs = simplexml_load_string('<fields>' . $index->getFieldDefinitions() . '</fields>');
$defField1 = $defs->xpath('field[@name="SearchUpdaterTest_Container_Field1"]');
$analyzers = $defField1[0]->analyzer;
$this->assertFalse((bool)$analyzers);
$index->addAnalyzer('Field1', 'charFilter', array('class' => 'solr.HTMLStripCharFilterFactory'));
$defs = simplexml_load_string('<fields>' . $index->getFieldDefinitions() . '</fields>');
$defField1 = $defs->xpath('field[@name="SearchUpdaterTest_Container_Field1"]');
$analyzers = $defField1[0]->analyzer;
$this->assertTrue((bool)$analyzers);
$this->assertEquals('solr.HTMLStripCharFilterFactory', $analyzers[0]->charFilter[0]['class']);
}
protected function getServiceMock() {
$serviceMock = Phockito::mock('SolrService');
$fakeResponse = new Apache_Solr_Response(new Apache_Solr_HttpTransport_Response(null, null, null));