Wordpress 3.0.2
[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'), $redirect['query']);
106                 } elseif ( is_single() && !empty($_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 && $redirect_url = get_author_posts_url($author->ID, $author->user_nicename) )
145                                 $redirect['query'] = remove_query_arg('author', $redirect['query']);
146                 } elseif ( is_category() || is_tag() || is_tax() ) { // Terms (Tags/categories)
147
148                         $term_count = 0;
149                         foreach ( array('category__in', 'category__not_in', 'category__and', 'post__in', 'post__not_in',
150                         'tag__in', 'tag__not_in', 'tag__and', 'tag_slug__in', 'tag_slug__and') as $key )
151                                 $term_count += count($wp_query->query_vars[$key]);
152
153                         $obj = $wp_query->get_queried_object();
154
155                         if ( $term_count <= 1 && !empty($obj->term_id) && ( $tax_url = get_term_link((int)$obj->term_id, $obj->taxonomy) )
156                                         && !is_wp_error($tax_url) && $redirect['query'] ) {
157                                 if ( is_category() ) {
158                                         $redirect['query'] = remove_query_arg( array( 'category_name', 'category', 'cat'), $redirect['query']);
159                                 } elseif ( is_tag() ) {
160                                         $redirect['query'] = remove_query_arg( array( 'tag', 'tag_id'), $redirect['query']);
161                                 } elseif ( is_tax() ) { // Custom taxonomies will have a custom query var, remove those too:
162                                         $tax = get_taxonomy( $obj->taxonomy );
163                                         if ( false !== $tax->query_var)
164                                                 $redirect['query'] = remove_query_arg($tax->query_var, $redirect['query']);
165                                         else
166                                                 $redirect['query'] = remove_query_arg( array( 'term', 'taxonomy'), $redirect['query']);
167                                 }
168
169                                 $tax_url = parse_url($tax_url);
170                                 if ( ! empty($tax_url['query']) ) { // Custom taxonomies may only be accessable via ?taxonomy=..&term=..
171                                         parse_str($tax_url['query'], $query_vars);
172                                         $redirect['query'] = add_query_arg($query_vars, $redirect['query']);
173                                 } else { // Taxonomy is accessable via a "pretty-URL"
174                                         $redirect['path'] = $tax_url['path'];
175                                 }
176
177                         }
178                 } elseif ( is_single() && strpos($wp_rewrite->permalink_structure, '%category%') !== false ) {
179                         $category = get_category_by_path(get_query_var('category_name'));
180                         $post_terms = wp_get_object_terms($wp_query->get_queried_object_id(), 'category', array('fields' => 'tt_ids'));
181                         if ( (!$category || is_wp_error($category)) || ( !is_wp_error($post_terms) && !empty($post_terms) && !in_array($category->term_taxonomy_id, $post_terms) ) )
182                                 $redirect_url = get_permalink($wp_query->get_queried_object_id());
183                 }
184
185                 // Post Paging
186                 if ( is_singular() && get_query_var('page') && $redirect_url ) {
187                         $redirect_url = trailingslashit( $redirect_url ) . user_trailingslashit( get_query_var( 'page' ), 'single_paged' );
188                         $redirect['query'] = remove_query_arg( 'page', $redirect['query'] );
189                 }
190
191                 // paging and feeds
192                 if ( get_query_var('paged') || is_feed() || get_query_var('cpage') ) {
193                         if ( !$redirect_url )
194                                 $redirect_url = $requested_url;
195                         $paged_redirect = @parse_url($redirect_url);
196                         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'] ) ) {
197                                 // Strip off paging and feed
198                                 $paged_redirect['path'] = preg_replace('#/page/?[0-9]+?(/+)?$#', '/', $paged_redirect['path']); // strip off any existing paging
199                                 $paged_redirect['path'] = preg_replace('#/(comments/?)?(feed|rss2?|rdf|atom)(/+|$)#', '/', $paged_redirect['path']); // strip off feed endings
200                                 $paged_redirect['path'] = preg_replace('#/comment-page-[0-9]+?(/+)?$#', '/', $paged_redirect['path']); // strip off any existing comment paging
201                         }
202
203                         $addl_path = '';
204                         if ( is_feed() ) {
205                                 $addl_path = !empty( $addl_path ) ? trailingslashit($addl_path) : '';
206                                 if ( get_query_var( 'withcomments' ) )
207                                         $addl_path .= 'comments/';
208                                 $addl_path .= user_trailingslashit( 'feed/' . ( ( get_default_feed() ==  get_query_var('feed') || 'feed' == get_query_var('feed') ) ? '' : get_query_var('feed') ), 'feed' );
209                                 $redirect['query'] = remove_query_arg( 'feed', $redirect['query'] );
210                         }
211
212                         if ( get_query_var('paged') > 0 ) {
213                                 $paged = get_query_var('paged');
214                                 $redirect['query'] = remove_query_arg( 'paged', $redirect['query'] );
215                                 if ( !is_feed() ) {
216                                         if ( $paged > 1 && !is_single() ) {
217                                                 $addl_path = ( !empty( $addl_path ) ? trailingslashit($addl_path) : '' ) . user_trailingslashit("page/$paged", 'paged');
218                                         } elseif ( !is_single() ) {
219                                                 $addl_path = !empty( $addl_path ) ? trailingslashit($addl_path) : '';
220                                         }
221                                 } elseif ( $paged > 1 ) {
222                                         $redirect['query'] = add_query_arg( 'paged', $paged, $redirect['query'] );
223                                 }
224                         }
225
226                         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 ) ) ) {
227                                 $addl_path = ( !empty( $addl_path ) ? trailingslashit($addl_path) : '' ) . user_trailingslashit( 'comment-page-' . get_query_var('cpage'), 'commentpaged' );
228                                 $redirect['query'] = remove_query_arg( 'cpage', $redirect['query'] );
229                         }
230
231                         $paged_redirect['path'] = user_trailingslashit( preg_replace('|/index.php/?$|', '/', $paged_redirect['path']) ); // strip off trailing /index.php/
232                         if ( !empty( $addl_path ) && $wp_rewrite->using_index_permalinks() && strpos($paged_redirect['path'], '/index.php/') === false )
233                                 $paged_redirect['path'] = trailingslashit($paged_redirect['path']) . 'index.php/';
234                         if ( !empty( $addl_path ) )
235                                 $paged_redirect['path'] = trailingslashit($paged_redirect['path']) . $addl_path;
236                         $redirect_url = $paged_redirect['scheme'] . '://' . $paged_redirect['host'] . $paged_redirect['path'];
237                         $redirect['path'] = $paged_redirect['path'];
238                 }
239         }
240
241         // tack on any additional query vars
242         $redirect['query'] = preg_replace( '#^\??&*?#', '', $redirect['query'] );
243         if ( $redirect_url && !empty($redirect['query']) ) {
244                 if ( strpos($redirect_url, '?') !== false )
245                         $redirect_url .= '&';
246                 else
247                         $redirect_url .= '?';
248                 $redirect_url .= $redirect['query'];
249         }
250
251         if ( $redirect_url )
252                 $redirect = @parse_url($redirect_url);
253
254         // www.example.com vs example.com
255         $user_home = @parse_url(home_url());
256         if ( !empty($user_home['host']) )
257                 $redirect['host'] = $user_home['host'];
258         if ( empty($user_home['path']) )
259                 $user_home['path'] = '/';
260
261         // Handle ports
262         if ( !empty($user_home['port']) )
263                 $redirect['port'] = $user_home['port'];
264         else
265                 unset($redirect['port']);
266
267         // trailing /index.php
268         $redirect['path'] = preg_replace('|/index.php/*?$|', '/', $redirect['path']);
269
270         // Remove trailing spaces from the path
271         $redirect['path'] = preg_replace( '#(%20| )+$#', '', $redirect['path'] );
272
273         if ( !empty( $redirect['query'] ) ) {
274                 // Remove trailing spaces from certain terminating query string args
275                 $redirect['query'] = preg_replace( '#((p|page_id|cat|tag)=[^&]*?)(%20| )+$#', '$1', $redirect['query'] );
276
277                 // Clean up empty query strings
278                 $redirect['query'] = trim(preg_replace( '#(^|&)(p|page_id|cat|tag)=?(&|$)#', '&', $redirect['query']), '&');
279
280                 // Remove redundant leading ampersands
281                 $redirect['query'] = preg_replace( '#^\??&*?#', '', $redirect['query'] );
282         }
283
284         // strip /index.php/ when we're not using PATHINFO permalinks
285         if ( !$wp_rewrite->using_index_permalinks() )
286                 $redirect['path'] = str_replace('/index.php/', '/', $redirect['path']);
287
288         // trailing slashes
289         if ( is_object($wp_rewrite) && $wp_rewrite->using_permalinks() && !is_404() && (!is_front_page() || ( is_front_page() && (get_query_var('paged') > 1) ) ) ) {
290                 $user_ts_type = '';
291                 if ( get_query_var('paged') > 0 ) {
292                         $user_ts_type = 'paged';
293                 } else {
294                         foreach ( array('single', 'category', 'page', 'day', 'month', 'year', 'home') as $type ) {
295                                 $func = 'is_' . $type;
296                                 if ( call_user_func($func) ) {
297                                         $user_ts_type = $type;
298                                         break;
299                                 }
300                         }
301                 }
302                 $redirect['path'] = user_trailingslashit($redirect['path'], $user_ts_type);
303         } elseif ( is_front_page() ) {
304                 $redirect['path'] = trailingslashit($redirect['path']);
305         }
306
307         // Strip multiple slashes out of the URL
308         if ( strpos($redirect['path'], '//') > -1 )
309                 $redirect['path'] = preg_replace('|/+|', '/', $redirect['path']);
310
311         // Always trailing slash the Front Page URL
312         if ( trailingslashit( $redirect['path'] ) == trailingslashit( $user_home['path'] ) )
313                 $redirect['path'] = trailingslashit($redirect['path']);
314
315         // Ignore differences in host capitalization, as this can lead to infinite redirects
316         // Only redirect no-www <=> yes-www
317         if ( strtolower($original['host']) == strtolower($redirect['host']) ||
318                 ( strtolower($original['host']) != 'www.' . strtolower($redirect['host']) && 'www.' . strtolower($original['host']) != strtolower($redirect['host']) ) )
319                 $redirect['host'] = $original['host'];
320
321         $compare_original = array($original['host'], $original['path']);
322
323         if ( !empty( $original['port'] ) )
324                 $compare_original[] = $original['port'];
325
326         if ( !empty( $original['query'] ) )
327                 $compare_original[] = $original['query'];
328
329         $compare_redirect = array($redirect['host'], $redirect['path']);
330
331         if ( !empty( $redirect['port'] ) )
332                 $compare_redirect[] = $redirect['port'];
333
334         if ( !empty( $redirect['query'] ) )
335                 $compare_redirect[] = $redirect['query'];
336
337         if ( $compare_original !== $compare_redirect ) {
338                 $redirect_url = $redirect['scheme'] . '://' . $redirect['host'];
339                 if ( !empty($redirect['port']) )
340                         $redirect_url .= ':' . $redirect['port'];
341                 $redirect_url .= $redirect['path'];
342                 if ( !empty($redirect['query']) )
343                         $redirect_url .= '?' . $redirect['query'];
344         }
345
346         if ( !$redirect_url || $redirect_url == $requested_url )
347                 return false;
348                 
349         // Hex encoded octets are case-insensitive. 
350         if ( false !== strpos($requested_url, '%') ) {
351                 if ( !function_exists('lowercase_octets') ) {
352                         function lowercase_octets($matches) { 
353                                 return strtolower( $matches[0] ); 
354                         } 
355                 }
356                 $requested_url = preg_replace_callback('|%[a-fA-F0-9][a-fA-F0-9]|', 'lowercase_octets', $requested_url);
357         }
358
359         // Note that you can use the "redirect_canonical" filter to cancel a canonical redirect for whatever reason by returning FALSE
360         $redirect_url = apply_filters('redirect_canonical', $redirect_url, $requested_url);
361
362         if ( !$redirect_url || $redirect_url == $requested_url ) // yes, again -- in case the filter aborted the request
363                 return false;
364
365         if ( $do_redirect ) {
366                 // protect against chained redirects
367                 if ( !redirect_canonical($redirect_url, false) ) {
368                         wp_redirect($redirect_url, 301);
369                         exit();
370                 } else {
371                         // Debug
372                         // die("1: $redirect_url<br />2: " . redirect_canonical( $redirect_url, false ) );
373                         return false;
374                 }
375         } else {
376                 return $redirect_url;
377         }
378 }
379
380 /**
381  * Attempts to guess correct post based on query vars.
382  *
383  * @since 2.3.0
384  * @uses $wpdb
385  *
386  * @return bool|string Returns False, if it can't find post, returns correct
387  *              location on success.
388  */
389 function redirect_guess_404_permalink() {
390         global $wpdb;
391
392         if ( !get_query_var('name') )
393                 return false;
394
395         $where = $wpdb->prepare("post_name LIKE %s", get_query_var('name') . '%');
396
397         // if any of post_type, year, monthnum, or day are set, use them to refine the query
398         if ( get_query_var('post_type') )
399                 $where .= $wpdb->prepare(" AND post_type = %s", get_query_var('post_type'));
400         if ( get_query_var('year') )
401                 $where .= $wpdb->prepare(" AND YEAR(post_date) = %d", get_query_var('year'));
402         if ( get_query_var('monthnum') )
403                 $where .= $wpdb->prepare(" AND MONTH(post_date) = %d", get_query_var('monthnum'));
404         if ( get_query_var('day') )
405                 $where .= $wpdb->prepare(" AND DAYOFMONTH(post_date) = %d", get_query_var('day'));
406
407         $post_id = $wpdb->get_var("SELECT ID FROM $wpdb->posts WHERE $where AND post_status = 'publish'");
408         if ( !$post_id )
409                 return false;
410         return get_permalink($post_id);
411 }
412
413 add_action('template_redirect', 'redirect_canonical');
414
415 ?>