首页 / PHP

用PHP和xapian构建全文检索

互联网，请记住我 2010-04-23 10:25:42 浏览 4,363 次

标签 xapian 全文检索检索

#ffa500">/**

* Search method

public functionsearch($params){

$this->xapian_init_readonly();

$start = microtime(true);

// queries array to later construct full query

$arr_queries = array();

// from date

if(!empty($params['date_from'])) {

$arr_queries[] = new XapianQuery(XapianQuery::OP_VALUE_GE, 6, date('Ymd', strtotime($params['date_from'])));

}

// to date

if(!empty($params['date_to'])) {

$arr_queries[] = new XapianQuery(XapianQuery::OP_VALUE_LE, 6, date('Ymd', strtotime($params['date_to'])));

}

// unique key

if(!empty($params['unique_key'])) {

$arr_queries[] = new XapianQuery(self::XAPIAN_PREFIX_UID . $params['unique_key']);

}

// normal search query parsed

if(!empty($params['search'])) {

$qp = new XapianQueryParser();

$qp->set_stemmer($this->xapian_stemmer);

$qp->set_database($this->xapian_read_db);

$qp->set_stemming_strategy(XapianQueryParser::STEM_SOME);

$arr_queries[] = $qp->parse_query($params['search']);

}

// Find the results for the query.

// construct final query

$query = array_pop($arr_queries);

foreach($arr_queries as$sq){

$query = new XapianQuery(XapianQuery::OP_AND, $query, $sq);

}

$this->xapian_enquire->set_query($query);

// set the count to the specified params

$offset = (isset($params['offset'])) ? intval($params['offset']) : 0;

$count = (isset($params['count'])) ? intval($params['count']) : self::DEFAULT_COUNT;

$matches = $this->xapian_enquire->get_mset($offset, $count);

$response = new stdClass();

$response->result_count = $matches->get_matches_estimated();

$results = array();

$i = $matches->begin();

while (!$i->equals($matches->end())){

$m = array();

$n = $i->get_rank() + 1;

$doc = $i->get_document();

$m['position'] = $n;

$m['url'] = $doc->get_value(self::XAPIAN_FIELD_URL);

$m['name'] = $doc->get_value(self::XAPIAN_FIELD_NAME);

$m['summary'] = $doc->get_value(self::XAPIAN_FIELD_SUMMARY);

$m['date'] = $doc->get_value(self::XAPIAN_FIELD_DATE);

$m['unique_key'] = $doc->get_value(self::XAPIAN_FIELD_UID);

$m['percent'] = $i->get_percent();

$results[count($results)] = $m;

$i->next();

}

$response->results = $results;

$end = microtime(true);

// runtime info

$response->execute = new stdClass();

$response->execute->call = 'search';

$response->execute->offset = $offset;

$response->execute->count = $count;

$response->execute->start = $start;

$response->execute->end = $end;

$response->execute->time = $end - $start;

// debug stuff

$response->execute->debug = $query->get_description();

return $response;

}

index.php

<?php
require_once 'XapianWrapper.php';
$x = new XapianWrapper();
$res = $x->index(array());
print_r($res);

Search.php

<?php
require_once 'XapianWrapper.php';
$x = new XapianWrapper();
$params = array('search' => 'foo');
$res = $x->search($params);
print_r($res);

delete.php

<?php
require_once 'XapianWrapper.php';
$x = new XapianWrapper();
$params = array(
'items' => array('foo'),
);
$res = $x->delete($params);
print_r($res);

使用示例:
您下载刚才的源码包后,就可以导入db.sql,并在命令里运行程序;

bash$ php index.php
stdClass Object
(
[indexed] => Array
(
[0] => Array
(
[name] => foo
[guid] => foo
[url] =>
)
[1] => Array
(
[name] => bar
[guid] => bar
[url] =>
)
)
)
bash$ php search.php
stdClass Object
(
[result_count] => 2
[results] => Array
(
[0] => Array
(
[position] => 1
[url] =>
[name] => foo
[summary] => foo bar test
[date] => 20081105
[unique_key] => foo
[percent] => 100
)
[1] => Array
(
[position] => 2
[url] =>
[name] => bar
[summary] => test foo bar
[date] => 20091105
[unique_key] => bar
[percent] => 50
)
)
[execute] => stdClass Object
(
[call] => search
[offset] => 0
[count] => 10
[start] => 1256674866.79
[end] => 1256674866.79
[time] => 0.000944852828979
[debug] => Xapian::Query(Zfoo:(pos=1))
)
)
bash$ php delete.php
Array
(
[0] => foo
)
bash$ php search.php
stdClass Object
(
[result_count] => 1
[results] => Array
(
[0] => Array
(
[position] => 1
[url] =>
[name] => bar
[summary] => test foo bar
[date] => 20091105
[unique_key] => bar
[percent] => 100
)
)
[execute] => stdClass Object
(
[call] => search
[offset] => 0
[count] => 10
[start] => 1256674876.02
[end] => 1256674876.02
[time] => 0.000872850418091
[debug] => Xapian::Query(Zfoo:(pos=1))
)
)

接下来,扩展您自己的程序来满足您的各种需求吧,欢迎反馈.好好地享受检索的乐趣吧.

建议继续学习

Xapian搜索体系结构（阅读 5,163）
Hermes：来自腾讯的实时检索分析平台（阅读 3,880）
Xapian的查询分析器（阅读 3,622）
检测文本正文是否包含有特定词的PHP扩展（阅读 3,401）
基于Lucene/XML的站内全文检索解决方案：WebLucene （阅读 3,322）
利用Sphinx实现实时全文检索（阅读 3,184）
Xapian 术语表（阅读 3,122）
分布式全文检索系统SolrCloud简介（阅读 2,820）
检索结果聚类展望（阅读 2,722）