Upstream/WordPress
1
0
Fork 0
mirror of https://github.com/WordPress/WordPress.git synced 2024-12-22 17:18:32 +01:00
Code Issues Projects Releases Wiki Activity

Order search results by relevance, rather than by date.

Browse Source 
The ordering logic is as follows:
 * Full sentence matches in post titles.
 * All search terms in post titles.
 * Any search terms in post titles.
 * Full sentence matches in post content.

Each section and any remaining posts are then sorted by date.

Introduces some filters:
 * wp_search_stopwords, to filter stop words ignored in WHERE.
 * posts_search_orderby, to filter the ORDER BY when ordering search results.

props azaozz, wonderboymusic.
fixes #7394.

Built from https://develop.svn.wordpress.org/trunk@25632


git-svn-id: http://core.svn.wordpress.org/trunk@25549 1a063a9b-81f0-0310-95a4-ce76da25c4cd
This commit is contained in:
Andrew Nacin 2013-09-27 17:34:09 +00:00
parent dfb46e5e2c
commit d7fc6d5d49
3 changed files with 223 additions and 42 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
wp-includes/deprecated.php
View File

@ -3316,6 +3316,18 @@ function user_pass_ok($user_login, $user_pass) {
*/
function _save_post_hook() {}
/**
* Formerly used internally to tidy up the search terms.
*
* @access private
* @since 2.9.0
* @deprecated 3.7.0
*/
function _search_terms_tidy( $t ) {
_deprecated_function( __FUNCTION__, '3.5', '' );
return trim( $t, "\"'\n\r " );
}
/**
* Check if the installed version of GD supports particular image type
*

13
wp-includes/functions.php
View File

@ -3702,19 +3702,6 @@ function get_file_data( $file, $default_headers, $context = '' ) {
return $all_headers;
}
/**
* Used internally to tidy up the search terms.
*
* @access private
* @since 2.9.0
*
* @param string $t
* @return string
*/
function _search_terms_tidy($t) {
return trim($t, "\"'\n\r ");
}
/**
* Returns true.
*

240
wp-includes/query.php
View File

@ -1289,6 +1289,14 @@ class WP_Query {
*/
var $thumbnails_cached = false;
/**
* Cached list of search stopwords.
*
* @since 3.7.0
* @var array
*/
private $stopwords;
/**
* Resets query flags to false.
*
@ -1470,6 +1478,10 @@ class WP_Query {
if ( '' !== $qv['second'] ) $qv['second'] = absint($qv['second']);
if ( '' !== $qv['menu_order'] ) $qv['menu_order'] = absint($qv['menu_order']);
// Fairly insane upper bound for search string lengths.
if ( ! empty( $qv['s'] ) && strlen( $qv['s'] ) > 1600 )
$qv['s'] = '';
// Compat. Map subpost to attachment.
if ( '' != $qv['subpost'] )
$qv['attachment'] = $qv['subpost'];
@ -1894,6 +1906,183 @@ class WP_Query {
do_action( 'parse_tax_query', $this );
}
/**
* Generate SQL for the WHERE clause based on passed search terms.
*
* @since 3.7.0
*
* @global type $wpdb
* @param array $q Query variables.
*/
protected function parse_search( &$q ) {
global $wpdb;
$search = '';
// added slashes screw with quote grouping when done early, so done later
$q['s'] = stripslashes( $q['s'] );
if ( empty( $_GET['s'] ) && $this->is_main_query() )
$q['s'] = urldecode( $q['s'] );
// there are no line breaks in <input /> fields
$q['s'] = str_replace( array( "\r", "\n" ), '', $q['s'] );
$q['search_terms_count'] = 1;
if ( ! empty( $q['sentence'] ) ) {
$q['search_terms'] = array( $q['s'] );
} else {
if ( preg_match_all( '/".*?("|$)|((?<=[\t ",+])|^)[^\t ",+]+/', $q['s'], $matches ) ) {
$q['search_terms_count'] = count( $matches[0] );
$q['search_terms'] = $this->parse_search_terms( $matches[0] );
// if the search string has only short terms or stopwords, or is 10+ terms long, match it as sentence
if ( empty( $q['search_terms'] ) || count( $q['search_terms'] ) > 9 )
$q['search_terms'] = array( $q['s'] );
} else {
$q['search_terms'] = array( $q['s'] );
}
}
$n = ! empty( $q['exact'] ) ? '' : '%';
$searchand = '';
$q['search_orderby_title'] = array();
foreach ( $q['search_terms'] as $term ) {
$term = like_escape( esc_sql( $term ) );
if ( $n )
$q['search_orderby_title'][] = "$wpdb->posts.post_title LIKE '%$term%'";
$search .= "{$searchand}(($wpdb->posts.post_title LIKE '{$n}{$term}{$n}') OR ($wpdb->posts.post_content LIKE '{$n}{$term}{$n}'))";
$searchand = ' AND ';
}
if ( ! empty( $search ) ) {
$search = " AND ({$search}) ";
if ( ! is_user_logged_in() )
$search .= " AND ($wpdb->posts.post_password = '') ";
}
/**
* Filter the search SQL that is used in the WHERE clause of WP_Query.
*
* @since 3.0.0
*
* @param string $search Search SQL for WHERE clause.
* @param WP_Query $this The current WP_Query object.
*/
return apply_filters_ref_array( 'posts_search', array( $search, &$this ) );
}
/**
* Check if the terms are suitable for searching.
*
* Uses an array of stopwords (terms) that are excluded from the separate
* term matching when searching for posts. The list of English stopwords is
* the approximate search engines list, and is translatable.
*
* @since 3.7.0
*
* @param array Terms to check.
* @return array Terms that are not stopwords.
*/
protected function parse_search_terms( $terms ) {
$strtolower = function_exists( 'mb_strtolower' ) ? 'mb_strtolower' : 'strtolower';
$checked = array();
$stopwords = $this->get_search_stopwords();
foreach ( $terms as $term ) {
// keep before/after spaces when term is for exact match
if ( preg_match( '/^".+"$/', $term ) )
$term = trim( $term, "\"'" );
else
$term = trim( $term, "\"' " );
// \p{L} matches a single letter that is not a Chinese, Japanese, etc. char
if ( ! $term || preg_match( '/^\p{L}$/u', $term ) )
continue;
if ( in_array( call_user_func( $strtolower, $term ), $stopwords, true ) )
continue;
$checked[] = $term;
}
return $checked;
}
/**
* Retrieve stopwords used when parsing search terms.
*
* @since 3.7.0
*
* @return array Stopwords.
*/
protected function get_search_stopwords() {
if ( isset( $this->stopwords ) )
return $this->stopwords;
/* translators: This is a comma-separated list of very common words that should be excluded from a search,
* like a, an, and the. These are usually called "stopwords". You should not simply translate these individual
* words into your language. Instead, look for and provide commonly accepted stopwords in your language.
*/
$words = explode( ',', _x( 'about,an,are,as,at,be,by,com,for,from,how,in,is,it,of,on,or,that,the,this,to,was,what,when,where,who,will,with,www',
'Comma-separated list of search stopwords in your language' ) );
foreach( $words as $word ) {
$word = trim( $word, "\r\n\t " );
if ( $word )
$stopwords[] = $word;
}
/**
* Filter stopwords used when parsing search terms.
*
* @since 3.7.0
*
* @param array $stopwords Stopwords.
*/
$this->stopwords = apply_filters( 'wp_search_stopwords', $stopwords );
return $this->stopwords;
}
/**
* Generate SQL for the ORDER BY condition based on passed search terms.
*
* @global wpdb $wpdb
* @param array $q Query variables.
* @return string ORDER BY clause.
*/
protected function parse_search_order( &$q ) {
global $wpdb;
$search_orderby = '';
if ( $q['search_terms_count'] > 1 ) {
$num_terms = count( $q['search_orderby_title'] );
$search_orderby_s = like_escape( esc_sql( $q['s'] ) );
$search_orderby = '(CASE ';
// sentence match in 'post_title'
$search_orderby .= "WHEN $wpdb->posts.post_title LIKE '%{$search_orderby_s}%' THEN 1 ";
// sanity limit, sort as sentence when more than 6 terms
// (few searches are longer than 6 terms and most titles are not)
if ( $num_terms < 7 ) {
// all words in title
$search_orderby .= 'WHEN ' . implode( ' AND ', $q['search_orderby_title'] ) . ' THEN 2 ';
// any word in title, not needed when $num_terms == 1
if ( $num_terms > 1 )
$search_orderby .= 'WHEN ' . implode( ' OR ', $q['search_orderby_title'] ) . ' THEN 3 ';
}
// sentence match in 'post_content'
$search_orderby .= "WHEN $wpdb->posts.post_content LIKE '%{$search_orderby_s}%' THEN 4 ";
$search_orderby .= 'ELSE 5 END)';
} else {
// single word or sentence search
$search_orderby = reset( $q['search_orderby_title'] ) . ' DESC';
}
return $search_orderby;
}
/**
* Sets the 404 property and saves whether query is feed.
*
@ -2235,35 +2424,9 @@ class WP_Query {
}
}
// If a search pattern is specified, load the posts that match
if ( !empty($q['s']) ) {
// added slashes screw with quote grouping when done early, so done later
$q['s'] = stripslashes($q['s']);
if ( empty( $_GET['s'] ) && $this->is_main_query() )
$q['s'] = urldecode($q['s']);
if ( !empty($q['sentence']) ) {
$q['search_terms'] = array($q['s']);
} else {
preg_match_all('/".*?("|$)|((?<=[\r\n\t ",+])|^)[^\r\n\t ",+]+/', $q['s'], $matches);
$q['search_terms'] = array_map('_search_terms_tidy', $matches[0]);
}
$n = !empty($q['exact']) ? '' : '%';
$searchand = '';
foreach( (array) $q['search_terms'] as $term ) {
$term = esc_sql( like_escape( $term ) );
$search .= "{$searchand}(($wpdb->posts.post_title LIKE '{$n}{$term}{$n}') OR ($wpdb->posts.post_content LIKE '{$n}{$term}{$n}'))";
$searchand = ' AND ';
}
if ( !empty($search) ) {
$search = " AND ({$search}) ";
if ( !is_user_logged_in() )
$search .= " AND ($wpdb->posts.post_password = '') ";
}
}
// Allow plugins to contextually add/remove/modify the search section of the database query
$search = apply_filters_ref_array('posts_search', array( $search, &$this ) );
// If a search pattern is specified, load the posts that match.
if ( ! empty( $q['s'] ) )
$search = $this->parse_search( $q );
// Taxonomies
if ( !$this->is_singular ) {
@ -2461,6 +2624,25 @@ class WP_Query {
$orderby .= " {$q['order']}";
}
// Order search results by relevance only when another "orderby" is not specified in the query.
if ( ! empty( $q['s'] ) ) {
$search_orderby = '';
if ( ! empty( $q['search_orderby_title'] ) && empty( $q['orderby'] ) )
$search_orderby = $this->parse_search_order( $q );
/**
* Filter the ORDER BY used when ordering search results.
*
* @since 3.7.0
*
* @param string $search_orderby The ORDER BY clause.
* @param WP_Query $this The current WP_Query instance.
*/
$search_orderby = apply_filters( 'posts_search_orderby', $search_orderby, $this );
if ( $search_orderby )
$orderby = $orderby ? $search_orderby . ', ' . $orderby : $search_orderby;
}
if ( is_array( $post_type ) && count( $post_type ) > 1 ) {
$post_type_cap = 'multiple_post_type';
} else {