Wordpress 3.0-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 penality 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() || $is_IIS || ( isset($_POST) && count($_POST) ) || is_preview() || is_robots() )
41                 return;
42
43         if ( !$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                         if ( get_query_var( 'page' ) ) {
107                                 $redirect_url = trailingslashit( $redirect_url ) . user_trailingslashit( get_query_var( 'page' ), 'single_paged' );
108                                 $redirect['query'] = remove_query_arg( 'page', $redirect['query'] );
109                         }
110                 } elseif ( is_single() && !empty($_GET['name'])  && ! $redirect_url ) {
111                         if ( $redirect_url = get_permalink( $wp_query->get_queried_object_id() ) )
112                                 $redirect['query'] = remove_query_arg('name', $redirect['query']);
113                 } elseif ( is_page() && !empty($_GET['page_id']) && ! $redirect_url ) {
114                         if ( $redirect_url = get_permalink(get_query_var('page_id')) )
115                                 $redirect['query'] = remove_query_arg('page_id', $redirect['query']);
116                 } elseif ( is_page() && !is_feed() && isset($wp_query->queried_object) && 'page' == get_option('show_on_front') && $wp_query->queried_object->ID == get_option('page_on_front')  && ! $redirect_url ) {
117                         $redirect_url = home_url('/');
118                 } elseif ( is_home() && !empty($_GET['page_id']) && 'page' == get_option('show_on_front') && get_query_var('page_id') == get_option('page_for_posts')  && ! $redirect_url ) {
119                         if ( $redirect_url = get_permalink(get_option('page_for_posts')) )
120                                 $redirect['query'] = remove_query_arg('page_id', $redirect['query']);
121                 } elseif ( !empty($_GET['m']) && ( is_year() || is_month() || is_day() ) ) {
122                         $m = get_query_var('m');
123                         switch ( strlen($m) ) {
124                                 case 4: // Yearly
125                                         $redirect_url = get_year_link($m);
126                                         break;
127                                 case 6: // Monthly
128                                         $redirect_url = get_month_link( substr($m, 0, 4), substr($m, 4, 2) );
129                                         break;
130                                 case 8: // Daily
131                                         $redirect_url = get_day_link(substr($m, 0, 4), substr($m, 4, 2), substr($m, 6, 2));
132                                         break;
133                         }
134                         if ( $redirect_url )
135                                 $redirect['query'] = remove_query_arg('m', $redirect['query']);
136                 // now moving on to non ?m=X year/month/day links
137                 } elseif ( is_day() && get_query_var('year') && get_query_var('monthnum') && !empty($_GET['day']) ) {
138                         if ( $redirect_url = get_day_link(get_query_var('year'), get_query_var('monthnum'), get_query_var('day')) )
139                                 $redirect['query'] = remove_query_arg(array('year', 'monthnum', 'day'), $redirect['query']);
140                 } elseif ( is_month() && get_query_var('year') && !empty($_GET['monthnum']) ) {
141                         if ( $redirect_url = get_month_link(get_query_var('year'), get_query_var('monthnum')) )
142                                 $redirect['query'] = remove_query_arg(array('year', 'monthnum'), $redirect['query']);
143                 } elseif ( is_year() && !empty($_GET['year']) ) {
144                         if ( $redirect_url = get_year_link(get_query_var('year')) )
145                                 $redirect['query'] = remove_query_arg('year', $redirect['query']);
146                 } elseif ( is_author() && !empty($_GET['author']) && preg_match( '|^[0-9]+$|', $_GET['author'] ) ) {
147                         $author = get_userdata(get_query_var('author'));
148                         if ( false !== $author && $redirect_url = get_author_posts_url($author->ID, $author->user_nicename) )
149                                 $redirect['query'] = remove_query_arg('author', $redirect['query']);
150                 } elseif ( is_category() || is_tag() || is_tax() ) { // Terms (Tags/categories)
151
152                         $term_count = 0;
153                         foreach ( array('category__in', 'category__not_in', 'category__and', 'post__in', 'post__not_in',
154                         'tag__in', 'tag__not_in', 'tag__and', 'tag_slug__in', 'tag_slug__and') as $key )
155                                 $term_count += count($wp_query->query_vars[$key]);
156
157                         $obj = $wp_query->get_queried_object();
158
159                         if ( $term_count <= 1 && !empty($obj->term_id) && ( $tax_url = get_term_link((int)$obj->term_id, $obj->taxonomy) ) && !is_wp_error($tax_url) ) {
160
161                                 if ( is_category() ) {
162                                         $redirect['query'] = remove_query_arg( array( 'category_name', 'category', 'cat'), $redirect['query']);
163                                 } elseif ( is_tag() ) {
164                                         $redirect['query'] = remove_query_arg( array( 'tag', 'tag_id'), $redirect['query']);
165                                 } elseif ( is_tax() ) { // Custom taxonomies will have a custom query var, remove those too:
166                                         $tax = get_taxonomy( $obj->taxonomy );
167                                         if ( false !== $tax->query_var)
168                                                 $redirect['query'] = remove_query_arg($tax->query_var, $redirect['query']);
169                                         else
170                                                 $redirect['query'] = remove_query_arg( array( 'term', 'taxonomy'), $redirect['query']);
171                                 }
172
173                                 $tax_url = parse_url($tax_url);
174                                 if ( ! empty($tax_url['query']) ) { // Custom taxonomies may only be accessable via ?taxonomy=..&term=..
175                                         parse_str($tax_url['query'], $query_vars);
176                                         $redirect['query'] = add_query_arg($query_vars, $redirect['query']);
177                                 } else { // Taxonomy is accessable via a "pretty-URL"
178                                         $redirect['path'] = $tax_url['path'];
179                                 }
180
181                         }
182                 } elseif ( is_single() && strpos($wp_rewrite->permalink_structure, '%category%') !== false ) {
183                         $category = get_term_by('slug', get_query_var('category_name'), 'category');
184                         $post_terms = wp_get_object_terms($wp_query->get_queried_object_id(), 'category', array('fields' => 'tt_ids'));
185                         if ( (!$category || is_wp_error($category)) || ( !is_wp_error($post_terms) && !empty($post_terms) && !in_array($category->term_taxonomy_id, $post_terms) ) )
186                                 $redirect_url = get_permalink($wp_query->get_queried_object_id());
187                 }
188
189                 // paging and feeds
190                 if ( get_query_var('paged') || is_feed() || get_query_var('cpage') ) {
191                         if ( !$redirect_url )
192                                 $redirect_url = $requested_url;
193                         $paged_redirect = @parse_url($redirect_url);
194                         while ( preg_match( '#/page/?[0-9]+?(/+)?$#', $paged_redirect['path'] ) || preg_match( '#/(comments/?)?(feed|rss|rdf|atom|rss2)(/+)?$#', $paged_redirect['path'] ) || preg_match( '#/comment-page-[0-9]+(/+)?$#', $paged_redirect['path'] ) ) {
195                                 // Strip off paging and feed
196                                 $paged_redirect['path'] = preg_replace('#/page/?[0-9]+?(/+)?$#', '/', $paged_redirect['path']); // strip off any existing paging
197                                 $paged_redirect['path'] = preg_replace('#/(comments/?)?(feed|rss2?|rdf|atom)(/+|$)#', '/', $paged_redirect['path']); // strip off feed endings
198                                 $paged_redirect['path'] = preg_replace('#/comment-page-[0-9]+?(/+)?$#', '/', $paged_redirect['path']); // strip off any existing comment paging
199                         }
200
201                         $addl_path = '';
202                         if ( is_feed() ) {
203                                 $addl_path = !empty( $addl_path ) ? trailingslashit($addl_path) : '';
204                                 if ( get_query_var( 'withcomments' ) )
205                                         $addl_path .= 'comments/';
206                                 $addl_path .= user_trailingslashit( 'feed/' . ( ( get_default_feed() ==  get_query_var('feed') || 'feed' == get_query_var('feed') ) ? '' : get_query_var('feed') ), 'feed' );
207                                 $redirect['query'] = remove_query_arg( 'feed', $redirect['query'] );
208                         }
209
210                         if ( get_query_var('paged') > 0 ) {
211                                 $paged = get_query_var('paged');
212                                 $redirect['query'] = remove_query_arg( 'paged', $redirect['query'] );
213                                 if ( !is_feed() ) {
214                                         if ( $paged > 1 && !is_single() ) {
215                                                 $addl_path = ( !empty( $addl_path ) ? trailingslashit($addl_path) : '' ) . user_trailingslashit("page/$paged", 'paged');
216                                         } elseif ( !is_single() ) {
217                                                 $addl_path = !empty( $addl_path ) ? trailingslashit($addl_path) : '';
218                                         }
219                                 } elseif ( $paged > 1 ) {
220                                         $redirect['query'] = add_query_arg( 'paged', $paged, $redirect['query'] );
221                                 }
222                         }
223
224                         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 ) ) ) {
225                                 $addl_path = ( !empty( $addl_path ) ? trailingslashit($addl_path) : '' ) . user_trailingslashit( 'comment-page-' . get_query_var('cpage'), 'commentpaged' );
226                                 $redirect['query'] = remove_query_arg( 'cpage', $redirect['query'] );
227                         }
228
229                         $paged_redirect['path'] = user_trailingslashit( preg_replace('|/index.php/?$|', '/', $paged_redirect['path']) ); // strip off trailing /index.php/
230                         if ( !empty( $addl_path ) && $wp_rewrite->using_index_permalinks() && strpos($paged_redirect['path'], '/index.php/') === false )
231                                 $paged_redirect['path'] = trailingslashit($paged_redirect['path']) . 'index.php/';
232                         if ( !empty( $addl_path ) )
233                                 $paged_redirect['path'] = trailingslashit($paged_redirect['path']) . $addl_path;
234                         $redirect_url = $paged_redirect['scheme'] . '://' . $paged_redirect['host'] . $paged_redirect['path'];
235                         $redirect['path'] = $paged_redirect['path'];
236                 }
237         }
238
239         // tack on any additional query vars
240         $redirect['query'] = preg_replace( '#^\??&*?#', '', $redirect['query'] );
241         if ( $redirect_url && !empty($redirect['query']) ) {
242                 if ( strpos($redirect_url, '?') !== false )
243                         $redirect_url .= '&';
244                 else
245                         $redirect_url .= '?';
246                 $redirect_url .= $redirect['query'];
247         }
248
249         if ( $redirect_url )
250                 $redirect = @parse_url($redirect_url);
251
252         // www.example.com vs example.com
253         $user_home = @parse_url(home_url());
254         if ( !empty($user_home['host']) )
255                 $redirect['host'] = $user_home['host'];
256         if ( empty($user_home['path']) )
257                 $user_home['path'] = '/';
258
259         // Handle ports
260         if ( !empty($user_home['port']) )
261                 $redirect['port'] = $user_home['port'];
262         else
263                 unset($redirect['port']);
264
265         // trailing /index.php
266         $redirect['path'] = preg_replace('|/index.php/*?$|', '/', $redirect['path']);
267
268         // Remove trailing spaces from the path
269         $redirect['path'] = preg_replace( '#(%20| )+$#', '', $redirect['path'] );
270
271         if ( !empty( $redirect['query'] ) ) {
272                 // Remove trailing spaces from certain terminating query string args
273                 $redirect['query'] = preg_replace( '#((p|page_id|cat|tag)=[^&]*?)(%20| )+$#', '$1', $redirect['query'] );
274
275                 // Clean up empty query strings
276                 $redirect['query'] = trim(preg_replace( '#(^|&)(p|page_id|cat|tag)=?(&|$)#', '&', $redirect['query']), '&');
277
278                 // Remove redundant leading ampersands
279                 $redirect['query'] = preg_replace( '#^\??&*?#', '', $redirect['query'] );
280         }
281
282         // strip /index.php/ when we're not using PATHINFO permalinks
283         if ( !$wp_rewrite->using_index_permalinks() )
284                 $redirect['path'] = str_replace('/index.php/', '/', $redirect['path']);
285
286         // trailing slashes
287         if ( is_object($wp_rewrite) && $wp_rewrite->using_permalinks() && !is_404() && (!is_front_page() || ( is_front_page() && (get_query_var('paged') > 1) ) ) ) {
288                 $user_ts_type = '';
289                 if ( get_query_var('paged') > 0 ) {
290                         $user_ts_type = 'paged';
291                 } else {
292                         foreach ( array('single', 'category', 'page', 'day', 'month', 'year', 'home') as $type ) {
293                                 $func = 'is_' . $type;
294                                 if ( call_user_func($func) ) {
295                                         $user_ts_type = $type;
296                                         break;
297                                 }
298                         }
299                 }
300                 $redirect['path'] = user_trailingslashit($redirect['path'], $user_ts_type);
301         } elseif ( is_front_page() ) {
302                 $redirect['path'] = trailingslashit($redirect['path']);
303         }
304
305         // Strip multiple slashes out of the URL
306         if ( strpos($redirect['path'], '//') > -1 )
307                 $redirect['path'] = preg_replace('|/+|', '/', $redirect['path']);
308
309         // Always trailing slash the Front Page URL
310         if ( trailingslashit( $redirect['path'] ) == trailingslashit( $user_home['path'] ) )
311                 $redirect['path'] = trailingslashit($redirect['path']);
312
313         // Ignore differences in host capitalization, as this can lead to infinite redirects
314         // Only redirect no-www <=> yes-www
315         if ( strtolower($original['host']) == strtolower($redirect['host']) ||
316                 ( strtolower($original['host']) != 'www.' . strtolower($redirect['host']) && 'www.' . strtolower($original['host']) != strtolower($redirect['host']) ) )
317                 $redirect['host'] = $original['host'];
318
319         $compare_original = array($original['host'], $original['path']);
320
321         if ( !empty( $original['port'] ) )
322                 $compare_original[] = $original['port'];
323
324         if ( !empty( $original['query'] ) )
325                 $compare_original[] = $original['query'];
326
327         $compare_redirect = array($redirect['host'], $redirect['path']);
328
329         if ( !empty( $redirect['port'] ) )
330                 $compare_redirect[] = $redirect['port'];
331
332         if ( !empty( $redirect['query'] ) )
333                 $compare_redirect[] = $redirect['query'];
334
335         if ( $compare_original !== $compare_redirect ) {
336                 $redirect_url = $redirect['scheme'] . '://' . $redirect['host'];
337                 if ( !empty($redirect['port']) )
338                         $redirect_url .= ':' . $redirect['port'];
339                 $redirect_url .= $redirect['path'];
340                 if ( !empty($redirect['query']) )
341                         $redirect_url .= '?' . $redirect['query'];
342         }
343
344         if ( !$redirect_url || $redirect_url == $requested_url )
345                 return false;
346
347         // Note that you can use the "redirect_canonical" filter to cancel a canonical redirect for whatever reason by returning FALSE
348         $redirect_url = apply_filters('redirect_canonical', $redirect_url, $requested_url);
349
350         if ( !$redirect_url || $redirect_url == $requested_url ) // yes, again -- in case the filter aborted the request
351                 return false;
352
353         if ( $do_redirect ) {
354                 // protect against chained redirects
355                 if ( !redirect_canonical($redirect_url, false) ) {
356                         wp_redirect($redirect_url, 301);
357                         exit();
358                 } else {
359                         // Debug
360                         // die("1: $redirect_url<br />2: " . redirect_canonical( $redirect_url, false ) );
361                         return false;
362                 }
363         } else {
364                 return $redirect_url;
365         }
366 }
367
368 /**
369  * Attempts to guess correct post based on query vars.
370  *
371  * @since 2.3.0
372  * @uses $wpdb
373  *
374  * @return bool|string Returns False, if it can't find post, returns correct
375  *              location on success.
376  */
377 function redirect_guess_404_permalink() {
378         global $wpdb;
379
380         if ( !get_query_var('name') )
381                 return false;
382
383         $where = $wpdb->prepare("post_name LIKE %s", get_query_var('name') . '%');
384
385         // if any of post_type, year, monthnum, or day are set, use them to refine the query
386         if ( get_query_var('post_type') )
387                 $where .= $wpdb->prepare(" AND post_type = %s", get_query_var('post_type'));
388         if ( get_query_var('year') )
389                 $where .= $wpdb->prepare(" AND YEAR(post_date) = %d", get_query_var('year'));
390         if ( get_query_var('monthnum') )
391                 $where .= $wpdb->prepare(" AND MONTH(post_date) = %d", get_query_var('monthnum'));
392         if ( get_query_var('day') )
393                 $where .= $wpdb->prepare(" AND DAYOFMONTH(post_date) = %d", get_query_var('day'));
394
395         $post_id = $wpdb->get_var("SELECT ID FROM $wpdb->posts WHERE $where AND post_status = 'publish'");
396         if ( !$post_id )
397                 return false;
398         return get_permalink($post_id);
399 }
400
401 add_action('template_redirect', 'redirect_canonical');
402
403 ?>