Ticket #7394: 7394.3.diff

src/wp-includes/functions.php

@@ -3703 +3703 @@
 }
 
 /**
- * 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.
  *
  * Useful for returning true to filters easily.
@@ -4042 +4029 @@
 }
 
 /**
+ * Check if the terms are suitable for searching.
+ *
+ * Includes 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. MySQL has a much longer default list of full-text stopwords.
+ *
+ * @since 3.6.0
+ * @access private
+ *
+ * @param array Terms to check
+ * @return array
+ */
+function _check_search_terms($terms) {
+        $strtolower_func = function_exists( 'mb_strtolower' ) ? 'mb_strtolower' : 'strtolower';
+        $checked = $stopwords = array();
+
+        $_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,the,www', 'Comma separated list of common words to exclude when searching (stopwords).' ) );
+
+        foreach( $_words as $word ) {
+                $word = trim( $word, "\r\n\t " );
+                if ( ! $word )
+                        continue;
+                $stopwords[] = $word;
+        }
+
+        $stopwords = apply_filters( 'wp_search_stopwords', $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( $strtolower_func( $term ), $stopwords, true ) )
+                        continue;
+
+                $checked[] = $term;
+        }
+
+        return $checked;
+}
+
+/**
  * Load the auth check for monitoring whether the user is still logged in.
  *
  * Can be disabled with remove_action( 'admin_enqueue_scripts', 'wp_auth_check_load' );

src/wp-includes/query.php

@@ -2230 +2230 @@
                 }
 
                 // If a search pattern is specified, load the posts that match
-                if ( !empty($q['s']) ) {
+                // Sanity check: search string shouldn't be more than 1600 characters.
+                // See ticket #21688 for more info.
+                if ( ! empty( $q['s'] ) && strlen( $q['s'] ) < 1600 ) {
                         // 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'] );
+                        $num_all_terms = 1;
                         if ( !empty($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', $matches[0]);
+                                if ( preg_match_all( '/".*?("|$)|((?<=[\t ",+])|^)[^\t ",+]+/', $q['s'], $matches ) ) {
+                                        $num_all_terms = count( $matches[0] );
+                                        $q['search_terms'] = _check_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 = '';
-                        foreach( (array) $q['search_terms'] as $term ) {
+                        $search_orderby_title = array();
+                        foreach ( $q['search_terms'] as $term ) {
                                 $term = esc_sql( like_escape( $term ) );
+                                if ( $n )
+                                        $search_orderby_title[] = "$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 ';
                         }
@@ -3666 +3683 @@
                 else
                         $post_type = 'post';
 
+                // Order search results by relevance when another "orderby" is not specified in the query
+                if ( ! empty( $search_orderby_title ) && empty( $q['orderby'] ) ) {
+                        $search_orderby = '';
+                        // Allow plugins to override sorting on 'post_title' and/or 'post_content'.
+                        // Passing an empty array would disable sorting.
+                        $search_orderby_on = array( 'post_title', 'post_content' );
+                        $search_orderby_on = apply_filters( 'posts_search_orderby_on', $search_orderby_on, $this );
+
+                        if ( $num_all_terms > 1 ) {
+                                $num_terms = count( $search_orderby_title );
+                                $search_orderby_s = esc_sql( like_escape( $q['s'] ) );
+
+                                if ( in_array( 'post_title', $search_orderby_on, true ) ) {
+                                        $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 ', $search_orderby_title ) . ' THEN 2 ';
+                                                // any word in title, not needed when $num_terms == 1
+                                                if ( $num_terms > 1 )
+                                                        $search_orderby .= 'WHEN ' . implode( ' OR ', $search_orderby_title ) . ' THEN 3 ';
+                                        }
+
+                                        // sentence match in 'post_content'
+                                        if ( in_array( 'post_content', $search_orderby_on, true ) )
+                                                $search_orderby .= "WHEN $wpdb->posts.post_content LIKE '%{$search_orderby_s}%' THEN 4 ";
+
+                                        $search_orderby .= 'ELSE 5 END)';
+                                } elseif ( in_array( 'post_content', $search_orderby_on, true ) ) {
+                                        // Not sorting on 'post_title', order by sentence matches in 'post_content'
+                                        $search_orderby = "$wpdb->posts.post_content LIKE '%{$search_orderby_s}%' DESC";
+                                }
+                        } else {
+                                // single word or sentence search
+                                if ( in_array( 'post_title', $search_orderby_on, true ) )
+                                        $search_orderby = reset( $search_orderby_title ) . ' DESC';
+                        }
+                        // Allow plugins to add/remove/modify the 'order by' for the search section of the database query
+                        $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 ) ) {
                         if ( count( $post_type ) > 1 )
                                 return;

src/wp-includes/deprecated.php

@@ -3317 +3317 @@
 function _save_post_hook() {}
 
 /**
+ * Formerly used internally to tidy up the search terms.
+ *
+ * @access private
+ * @since 2.9.0
+ * @deprecated 3.6.0
+ */
+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
  *
  * @since 2.9.0