Wordpress 3.3.1-scripts
[autoinstalls/wordpress.git] / wp-includes / canonical.php
1 <?php
2 /**
3  * Canonical API to handle WordPress Redirecting
4  *
5  * Based on "Permalink Redirect" from Scott Yang and "Enforce www. Preference"
6  * by Mark Jaquith
7  *
8  * @package WordPress
9  * @since 2.3.0
10  */
11
12 /**
13  * Redirects incoming links to the proper URL based on the site url.
14  *
15  * Search engines consider www.somedomain.com and somedomain.com to be two
16  * different URLs when they both go to the same location. This SEO enhancement
17  * prevents penalty for duplicate content by redirecting all incoming links to
18  * one or the other.
19  *
20  * Prevents redirection for feeds, trackbacks, searches, comment popup, and
21  * admin URLs. Does not redirect on IIS, page/post previews, and on form data.
22  *
23  * Will also attempt to find the correct link when a user enters a URL that does
24  * not exist based on exact WordPress query. Will instead try to parse the URL
25  * or query in an attempt to figure the correct page to go to.
26  *
27  * @since 2.3.0
28  * @uses $wp_rewrite
29  * @uses $is_IIS
30  *
31  * @param string $requested_url Optional. The URL that was requested, used to
32  *              figure if redirect is needed.
33  * @param bool $do_redirect Optional. Redirect to the new URL.
34  * @return null|false|string Null, if redirect not needed. False, if redirect
35  *              not needed or the string of the URL
36  */
37 function redirect_canonical( $requested_url = null, $do_redirect = true ) {
38         global $wp_rewrite, $is_IIS, $wp_query, $wpdb;
39
40         if ( is_trackback() || is_search() || is_comments_popup() || is_admin() || !empty($_POST) || is_preview() || is_robots() || $is_IIS )
41                 return;
42
43         if ( !$requested_url ) {
44                 // build the URL in the address bar
45                 $requested_url  = is_ssl() ? 'https://' : 'http://';
46                 $requested_url .= $_SERVER['HTTP_HOST'];
47                 $requested_url .= $_SERVER['REQUEST_URI'];
48         }
49
50         $original = @parse_url($requested_url);
51         if ( false === $original )
52                 return;
53
54         // Some PHP setups turn requests for / into /index.php in REQUEST_URI
55         // See: http://trac.wordpress.org/ticket/5017
56         // See: http://trac.wordpress.org/ticket/7173
57         // Disabled, for now:
58         // $original['path'] = preg_replace('|/index\.php$|', '/', $original['path']);
59
60         $redirect = $original;
61         $redirect_url = false;
62
63         // Notice fixing
64         if ( !isset($redirect['path']) )
65                 $redirect['path'] = '';
66         if ( !isset($redirect['query']) )
67                 $redirect['query'] = '';
68
69         if ( is_singular() && 1 > $wp_query->post_count && ($id = get_query_var('p')) ) {
70
71                 $vars = $wpdb->get_results( $wpdb->prepare("SELECT post_type, post_parent FROM $wpdb->posts WHERE ID = %d", $id) );
72
73                 if ( isset($vars[0]) && $vars = $vars[0] ) {
74                         if ( 'revision' == $vars->post_type && $vars->post_parent > 0 )
75                                 $id = $vars->post_parent;
76
77                         if ( $redirect_url = get_permalink($id) )
78                                 $redirect['query'] = remove_query_arg(array('p', 'page_id', 'attachment_id', 'post_type'), $redirect['query']);
79                 }
80         }
81
82         // These tests give us a WP-generated permalink
83         if ( is_404() ) {
84
85                 // Redirect ?page_id, ?p=, ?attachment_id= to their respective url's
86                 $id = max( get_query_var('p'), get_query_var('page_id'), get_query_var('attachment_id') );
87                 if ( $id && $redirect_post = get_post($id) ) {
88                         $post_type_obj = get_post_type_object($redirect_post->post_type);
89                         if ( $post_type_obj->public ) {
90                                 $redirect_url = get_permalink($redirect_post);
91                                 $redirect['query'] = remove_query_arg(array('p', 'page_id', 'attachment_id', 'post_type'), $redirect['query']);
92                         }
93                 }
94
95                 if ( ! $redirect_url )
96                         $redirect_url = redirect_guess_404_permalink();
97
98         } elseif ( is_object($wp_rewrite) && $wp_rewrite->using_permalinks() ) {
99                 // rewriting of old ?p=X, ?m=2004, ?m=200401, ?m=20040101
100                 if ( is_attachment() && !empty($_GET['attachment_id']) && ! $redirect_url ) {
101                         if ( $redirect_url = get_attachment_link(get_query_var('attachment_id')) )
102                                 $redirect['query'] = remove_query_arg('attachment_id', $redirect['query']);
103                 } elseif ( is_single() && !empty($_GET['p']) && ! $redirect_url ) {
104                         if ( $redirect_url = get_permalink(get_query_var('p')) )
105                                 $redirect['query'] = remove_query_arg(array('p', 'post_type'), $redirect['query']);
106                 } elseif ( is_single() && !empty($_GET['name'])  && ! $redirect_url ) {
107                         if ( $redirect_url = get_permalink( $wp_query->get_queried_object_id() ) )
108                                 $redirect['query'] = remove_query_arg('name', $redirect['query']);
109                 } elseif ( is_page() && !empty($_GET['page_id']) && ! $redirect_url ) {
110                         if ( $redirect_url = get_permalink(get_query_var('page_id')) )
111                                 $redirect['query'] = remove_query_arg('page_id', $redirect['query']);
112                 } elseif ( is_page() && !is_feed() && isset($wp_query->queried_object) && 'page' == get_option('show_on_front') && $wp_query->queried_object->ID == get_option('page_on_front')  && ! $redirect_url ) {
113                         $redirect_url = home_url('/');
114                 } elseif ( is_home() && !empty($_GET['page_id']) && 'page' == get_option('show_on_front') && get_query_var('page_id') == get_option('page_for_posts')  && ! $redirect_url ) {
115                         if ( $redirect_url = get_permalink(get_option('page_for_posts')) )
116                                 $redirect['query'] = remove_query_arg('page_id', $redirect['query']);
117                 } elseif ( !empty($_GET['m']) && ( is_year() || is_month() || is_day() ) ) {
118                         $m = get_query_var('m');
119                         switch ( strlen($m) ) {
120                                 case 4: // Yearly
121                                         $redirect_url = get_year_link($m);
122                                         break;
123                                 case 6: // Monthly
124                                         $redirect_url = get_month_link( substr($m, 0, 4), substr($m, 4, 2) );
125                                         break;
126                                 case 8: // Daily
127                                         $redirect_url = get_day_link(substr($m, 0, 4), substr($m, 4, 2), substr($m, 6, 2));
128                                         break;
129                         }
130                         if ( $redirect_url )
131                                 $redirect['query'] = remove_query_arg('m', $redirect['query']);
132                 // now moving on to non ?m=X year/month/day links
133                 } elseif ( is_day() && get_query_var('year') && get_query_var('monthnum') && !empty($_GET['day']) ) {
134                         if ( $redirect_url = get_day_link(get_query_var('year'), get_query_var('monthnum'), get_query_var('day')) )
135                                 $redirect['query'] = remove_query_arg(array('year', 'monthnum', 'day'), $redirect['query']);
136                 } elseif ( is_month() && get_query_var('year') && !empty($_GET['monthnum']) ) {
137                         if ( $redirect_url = get_month_link(get_query_var('year'), get_query_var('monthnum')) )
138                                 $redirect['query'] = remove_query_arg(array('year', 'monthnum'), $redirect['query']);
139                 } elseif ( is_year() && !empty($_GET['year']) ) {
140                         if ( $redirect_url = get_year_link(get_query_var('year')) )
141                                 $redirect['query'] = remove_query_arg('year', $redirect['query']);
142                 } elseif ( is_author() && !empty($_GET['author']) && preg_match( '|^[0-9]+$|', $_GET['author'] ) ) {
143                         $author = get_userdata(get_query_var('author'));
144                         if ( ( false !== $author ) && $wpdb->get_var( $wpdb->prepare( "SELECT ID FROM $wpdb->posts WHERE $wpdb->posts.post_author = %d AND $wpdb->posts.post_status = 'publish' LIMIT 1", $author->ID ) ) ) {
145                                 if ( $redirect_url = get_author_posts_url($author->ID, $author->user_nicename) )
146                                         $redirect['query'] = remove_query_arg('author', $redirect['query']);
147                         }
148                 } elseif ( is_category() || is_tag() || is_tax() ) { // Terms (Tags/categories)
149
150                         $term_count = 0;
151                         foreach ( $wp_query->tax_query->queries as $tax_query )
152                                 $term_count += count( $tax_query['terms'] );
153
154                         $obj = $wp_query->get_queried_object();
155                         if ( $term_count <= 1 && !empty($obj->term_id) && ( $tax_url = get_term_link((int)$obj->term_id, $obj->taxonomy) ) && !is_wp_error($tax_url) ) {
156                                 if ( !empty($redirect['query']) ) {
157                                         // Strip taxonomy query vars off the url.
158                                         $qv_remove = array( 'term', 'taxonomy');
159                                         if ( is_category() ) {
160                                                 $qv_remove[] = 'category_name';
161                                                 $qv_remove[] = 'cat';
162                                         } elseif ( is_tag() ) {
163                                                 $qv_remove[] = 'tag';
164                                                 $qv_remove[] = 'tag_id';
165                                         } else { // Custom taxonomies will have a custom query var, remove those too:
166                                                 $tax_obj = get_taxonomy( $obj->taxonomy );
167                                                 if ( false !== $tax_obj->query_var )
168                                                         $qv_remove[] = $tax_obj->query_var;
169                                         }
170
171                                         $rewrite_vars = array_diff( array_keys($wp_query->query), array_keys($_GET) );
172
173                                         if ( !array_diff($rewrite_vars, array_keys($_GET))  ) { // Check to see if all the Query vars are coming from the rewrite, none are set via $_GET
174                                                 $redirect['query'] = remove_query_arg($qv_remove, $redirect['query']); //Remove all of the per-tax qv's
175
176                                                 // Create the destination url for this taxonomy
177                                                 $tax_url = parse_url($tax_url);
178                                                 if ( ! empty($tax_url['query']) ) { // Taxonomy accessible via ?taxonomy=..&term=.. or any custom qv..
179                                                         parse_str($tax_url['query'], $query_vars);
180                                                         $redirect['query'] = add_query_arg($query_vars, $redirect['query']);
181                                                 } else { // Taxonomy is accessible via a "pretty-URL"
182                                                         $redirect['path'] = $tax_url['path'];
183                                                 }
184
185                                         } else { // Some query vars are set via $_GET. Unset those from $_GET that exist via the rewrite
186                                                 foreach ( $qv_remove as $_qv ) {
187                                                         if ( isset($rewrite_vars[$_qv]) )
188                                                                 $redirect['query'] = remove_query_arg($_qv, $redirect['query']);
189                                                 }
190                                         }
191                                 }
192
193                         }
194                 } elseif ( is_single() && strpos($wp_rewrite->permalink_structure, '%category%') !== false ) {
195                         $category = get_category_by_path(get_query_var('category_name'));
196                         $post_terms = wp_get_object_terms($wp_query->get_queried_object_id(), 'category', array('fields' => 'tt_ids'));
197                         if ( (!$category || is_wp_error($category)) || ( !is_wp_error($post_terms) && !empty($post_terms) && !in_array($category->term_taxonomy_id, $post_terms) ) )
198                                 $redirect_url = get_permalink($wp_query->get_queried_object_id());
199                 }
200
201                 // Post Paging
202                 if ( is_singular() && get_query_var('page') && $redirect_url ) {
203                         $redirect_url = trailingslashit( $redirect_url ) . user_trailingslashit( get_query_var( 'page' ), 'single_paged' );
204                         $redirect['query'] = remove_query_arg( 'page', $redirect['query'] );
205                 }
206
207                 // paging and feeds
208                 if ( get_query_var('paged') || is_feed() || get_query_var('cpage') ) {
209                         while ( preg_match( "#/$wp_rewrite->pagination_base/?[0-9]+?(/+)?$#", $redirect['path'] ) || preg_match( '#/(comments/?)?(feed|rss|rdf|atom|rss2)(/+)?$#', $redirect['path'] ) || preg_match( '#/comment-page-[0-9]+(/+)?$#', $redirect['path'] ) ) {
210                                 // Strip off paging and feed
211                                 $redirect['path'] = preg_replace("#/$wp_rewrite->pagination_base/?[0-9]+?(/+)?$#", '/', $redirect['path']); // strip off any existing paging
212                                 $redirect['path'] = preg_replace('#/(comments/?)?(feed|rss2?|rdf|atom)(/+|$)#', '/', $redirect['path']); // strip off feed endings
213                                 $redirect['path'] = preg_replace('#/comment-page-[0-9]+?(/+)?$#', '/', $redirect['path']); // strip off any existing comment paging
214                         }
215
216                         $addl_path = '';
217                         if ( is_feed() && in_array( get_query_var('feed'), $wp_rewrite->feeds ) ) {
218                                 $addl_path = !empty( $addl_path ) ? trailingslashit($addl_path) : '';
219                                 if ( get_query_var( 'withcomments' ) )
220                                         $addl_path .= 'comments/';
221                                 if ( ( 'rss' == get_default_feed() && 'feed' == get_query_var('feed') ) || 'rss' == get_query_var('feed') )
222                                         $addl_path .= user_trailingslashit( 'feed/' . ( ( get_default_feed() == 'rss2' ) ? '' : 'rss2' ), 'feed' );
223                                 else
224                                         $addl_path .= user_trailingslashit( 'feed/' . ( ( get_default_feed() ==  get_query_var('feed') || 'feed' == get_query_var('feed') ) ? '' : get_query_var('feed') ), 'feed' );
225                                 $redirect['query'] = remove_query_arg( 'feed', $redirect['query'] );
226                         } elseif ( is_feed() && 'old' == get_query_var('feed') ) {
227                                 $old_feed_files = array(
228                                         'wp-atom.php'         => 'atom',
229                                         'wp-commentsrss2.php' => 'comments_rss2',
230                                         'wp-feed.php'         => get_default_feed(),
231                                         'wp-rdf.php'          => 'rdf',
232                                         'wp-rss.php'          => 'rss2',
233                                         'wp-rss2.php'         => 'rss2',
234                                 );
235                                 if ( isset( $old_feed_files[ basename( $redirect['path'] ) ] ) ) {
236                                         $redirect_url = get_feed_link( $old_feed_files[ basename( $redirect['path'] ) ] );
237                                         wp_redirect( $redirect_url, 301 );
238                                         die();
239                                 }
240                         }
241
242                         if ( get_query_var('paged') > 0 ) {
243                                 $paged = get_query_var('paged');
244                                 $redirect['query'] = remove_query_arg( 'paged', $redirect['query'] );
245                                 if ( !is_feed() ) {
246                                         if ( $paged > 1 && !is_single() ) {
247                                                 $addl_path = ( !empty( $addl_path ) ? trailingslashit($addl_path) : '' ) . user_trailingslashit("$wp_rewrite->pagination_base/$paged", 'paged');
248                                         } elseif ( !is_single() ) {
249                                                 $addl_path = !empty( $addl_path ) ? trailingslashit($addl_path) : '';
250                                         }
251                                 } elseif ( $paged > 1 ) {
252                                         $redirect['query'] = add_query_arg( 'paged', $paged, $redirect['query'] );
253                                 }
254                         }
255
256                         if ( get_option('page_comments') && ( ( 'newest' == get_option('default_comments_page') && get_query_var('cpage') > 0 ) || ( 'newest' != get_option('default_comments_page') && get_query_var('cpage') > 1 ) ) ) {
257                                 $addl_path = ( !empty( $addl_path ) ? trailingslashit($addl_path) : '' ) . user_trailingslashit( 'comment-page-' . get_query_var('cpage'), 'commentpaged' );
258                                 $redirect['query'] = remove_query_arg( 'cpage', $redirect['query'] );
259                         }
260
261                         $redirect['path'] = user_trailingslashit( preg_replace('|/index.php/?$|', '/', $redirect['path']) ); // strip off trailing /index.php/
262                         if ( !empty( $addl_path ) && $wp_rewrite->using_index_permalinks() && strpos($redirect['path'], '/index.php/') === false )
263                                 $redirect['path'] = trailingslashit($redirect['path']) . 'index.php/';
264                         if ( !empty( $addl_path ) )
265                                 $redirect['path'] = trailingslashit($redirect['path']) . $addl_path;
266                         $redirect_url = $redirect['scheme'] . '://' . $redirect['host'] . $redirect['path'];
267                 }
268         }
269
270         // tack on any additional query vars
271         $redirect['query'] = preg_replace( '#^\??&*?#', '', $redirect['query'] );
272         if ( $redirect_url && !empty($redirect['query']) ) {
273                 parse_str( $redirect['query'], $_parsed_query );
274                 $redirect = @parse_url($redirect_url);
275
276                 if ( ! empty( $_parsed_query['name'] ) && ! empty( $redirect['query'] ) ) {
277                         parse_str( $redirect['query'], $_parsed_redirect_query );
278
279                         if ( empty( $_parsed_redirect_query['name'] ) )
280                                 unset( $_parsed_query['name'] );
281                 }
282
283                 $_parsed_query = array_map( 'rawurlencode', $_parsed_query );
284                 $redirect_url = add_query_arg( $_parsed_query, $redirect_url );
285         }
286
287         if ( $redirect_url )
288                 $redirect = @parse_url($redirect_url);
289
290         // www.example.com vs example.com
291         $user_home = @parse_url(home_url());
292         if ( !empty($user_home['host']) )
293                 $redirect['host'] = $user_home['host'];
294         if ( empty($user_home['path']) )
295                 $user_home['path'] = '/';
296
297         // Handle ports
298         if ( !empty($user_home['port']) )
299                 $redirect['port'] = $user_home['port'];
300         else
301                 unset($redirect['port']);
302
303         // trailing /index.php
304         $redirect['path'] = preg_replace('|/index.php/*?$|', '/', $redirect['path']);
305
306         // Remove trailing spaces from the path
307         $redirect['path'] = preg_replace( '#(%20| )+$#', '', $redirect['path'] );
308
309         if ( !empty( $redirect['query'] ) ) {
310                 // Remove trailing spaces from certain terminating query string args
311                 $redirect['query'] = preg_replace( '#((p|page_id|cat|tag)=[^&]*?)(%20| )+$#', '$1', $redirect['query'] );
312
313                 // Clean up empty query strings
314                 $redirect['query'] = trim(preg_replace( '#(^|&)(p|page_id|cat|tag)=?(&|$)#', '&', $redirect['query']), '&');
315
316                 // Redirect obsolete feeds
317                 $redirect['query'] = preg_replace( '#(^|&)feed=rss(&|$)#', '$1feed=rss2$3', $redirect['query'] );
318
319                 // Remove redundant leading ampersands
320                 $redirect['query'] = preg_replace( '#^\??&*?#', '', $redirect['query'] );
321         }
322
323         // strip /index.php/ when we're not using PATHINFO permalinks
324         if ( !$wp_rewrite->using_index_permalinks() )
325                 $redirect['path'] = str_replace('/index.php/', '/', $redirect['path']);
326
327         // trailing slashes
328         if ( is_object($wp_rewrite) && $wp_rewrite->using_permalinks() && !is_404() && (!is_front_page() || ( is_front_page() && (get_query_var('paged') > 1) ) ) ) {
329                 $user_ts_type = '';
330                 if ( get_query_var('paged') > 0 ) {
331                         $user_ts_type = 'paged';
332                 } else {
333                         foreach ( array('single', 'category', 'page', 'day', 'month', 'year', 'home') as $type ) {
334                                 $func = 'is_' . $type;
335                                 if ( call_user_func($func) ) {
336                                         $user_ts_type = $type;
337                                         break;
338                                 }
339                         }
340                 }
341                 $redirect['path'] = user_trailingslashit($redirect['path'], $user_ts_type);
342         } elseif ( is_front_page() ) {
343                 $redirect['path'] = trailingslashit($redirect['path']);
344         }
345
346         // Strip multiple slashes out of the URL
347         if ( strpos($redirect['path'], '//') > -1 )
348                 $redirect['path'] = preg_replace('|/+|', '/', $redirect['path']);
349
350         // Always trailing slash the Front Page URL
351         if ( trailingslashit( $redirect['path'] ) == trailingslashit( $user_home['path'] ) )
352                 $redirect['path'] = trailingslashit($redirect['path']);
353
354         // Ignore differences in host capitalization, as this can lead to infinite redirects
355         // Only redirect no-www <=> yes-www
356         if ( strtolower($original['host']) == strtolower($redirect['host']) ||
357                 ( strtolower($original['host']) != 'www.' . strtolower($redirect['host']) && 'www.' . strtolower($original['host']) != strtolower($redirect['host']) ) )
358                 $redirect['host'] = $original['host'];
359
360         $compare_original = array($original['host'], $original['path']);
361
362         if ( !empty( $original['port'] ) )
363                 $compare_original[] = $original['port'];
364
365         if ( !empty( $original['query'] ) )
366                 $compare_original[] = $original['query'];
367
368         $compare_redirect = array($redirect['host'], $redirect['path']);
369
370         if ( !empty( $redirect['port'] ) )
371                 $compare_redirect[] = $redirect['port'];
372
373         if ( !empty( $redirect['query'] ) )
374                 $compare_redirect[] = $redirect['query'];
375
376         if ( $compare_original !== $compare_redirect ) {
377                 $redirect_url = $redirect['scheme'] . '://' . $redirect['host'];
378                 if ( !empty($redirect['port']) )
379                         $redirect_url .= ':' . $redirect['port'];
380                 $redirect_url .= $redirect['path'];
381                 if ( !empty($redirect['query']) )
382                         $redirect_url .= '?' . $redirect['query'];
383         }
384
385         if ( !$redirect_url || $redirect_url == $requested_url )
386                 return false;
387
388         // Hex encoded octets are case-insensitive.
389         if ( false !== strpos($requested_url, '%') ) {
390                 if ( !function_exists('lowercase_octets') ) {
391                         function lowercase_octets($matches) {
392                                 return strtolower( $matches[0] );
393                         }
394                 }
395                 $requested_url = preg_replace_callback('|%[a-fA-F0-9][a-fA-F0-9]|', 'lowercase_octets', $requested_url);
396         }
397
398         // Note that you can use the "redirect_canonical" filter to cancel a canonical redirect for whatever reason by returning FALSE
399         $redirect_url = apply_filters('redirect_canonical', $redirect_url, $requested_url);
400
401         if ( !$redirect_url || $redirect_url == $requested_url ) // yes, again -- in case the filter aborted the request
402                 return false;
403
404         if ( $do_redirect ) {
405                 // protect against chained redirects
406                 if ( !redirect_canonical($redirect_url, false) ) {
407                         wp_redirect($redirect_url, 301);
408                         exit();
409                 } else {
410                         // Debug
411                         // die("1: $redirect_url<br />2: " . redirect_canonical( $redirect_url, false ) );
412                         return false;
413                 }
414         } else {
415                 return $redirect_url;
416         }
417 }
418
419 /**
420  * Attempts to guess correct post based on query vars.
421  *
422  * @since 2.3.0
423  * @uses $wpdb
424  *
425  * @return bool|string Returns False, if it can't find post, returns correct
426  *              location on success.
427  */
428 function redirect_guess_404_permalink() {
429         global $wpdb;
430
431         if ( !get_query_var('name') )
432                 return false;
433
434         $where = $wpdb->prepare("post_name LIKE %s", like_escape( get_query_var('name') ) . '%');
435
436         // if any of post_type, year, monthnum, or day are set, use them to refine the query
437         if ( get_query_var('post_type') )
438                 $where .= $wpdb->prepare(" AND post_type = %s", get_query_var('post_type'));
439         if ( get_query_var('year') )
440                 $where .= $wpdb->prepare(" AND YEAR(post_date) = %d", get_query_var('year'));
441         if ( get_query_var('monthnum') )
442                 $where .= $wpdb->prepare(" AND MONTH(post_date) = %d", get_query_var('monthnum'));
443         if ( get_query_var('day') )
444                 $where .= $wpdb->prepare(" AND DAYOFMONTH(post_date) = %d", get_query_var('day'));
445
446         $post_id = $wpdb->get_var("SELECT ID FROM $wpdb->posts WHERE $where AND post_status = 'publish'");
447         if ( !$post_id )
448                 return false;
449         return get_permalink($post_id);
450 }
451
452 add_action('template_redirect', 'redirect_canonical');
453
454 ?>