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  * @author Scott Yang
   9  * @author Mark Jaquith
  10  * @package WordPress
  11  * @since 2.3
  12  */
  13
  14 /**
  15  * Redirects incoming links to the proper URL based on the site url
  16  *
  17  * Search engines consider www.somedomain.com and somedomain.com to be two
  18  * different URLs when they both go to the same location. This SEO enhancement
  19  * prevents penality for duplicate content by redirecting all incoming links to
  20  * one or the other.
  21  *
  22  * Prevents redirection for feeds, trackbacks, searches, comment popup, and
  23  * admin URLs. Does not redirect on IIS, page/post previews, and on form data.
  24  *
  25  * Will also attempt to find the correct link when a user enters a URL that does
  26  * not exist based on exact WordPress query. Will instead try to parse the URL
  27  * or query in an attempt to figure the correct page to go to.
  28  *
  29  * @since 2.3
  30  * @uses $wp_rewrite
  31  * @uses $is_IIS
  32  *
  33  * @param string $requested_url Optional. The URL that was requested, used to
  34  *              figure if redirect is needed.
  35  * @param bool $do_redirect Optional. Redirect to the new URL.
  36  * @return null|false|string Null, if redirect not needed. False, if redirect
  37  *              not needed or the string of the URL
  38  */
  39 function redirect_canonical($requested_url=null, $do_redirect=true) {
  40         global $wp_rewrite, $is_IIS;
  41
  42         if ( is_feed() || is_trackback() || is_search() || is_comments_popup() || is_admin() || $is_IIS || ( isset($_POST) && count($_POST) ) || is_preview() || is_robots() )
  43                 return;
  44
  45         if ( !$requested_url ) {
  46                 // build the URL in the address bar
  47                 $requested_url  = ( isset($_SERVER['HTTPS'] ) && strtolower($_SERVER['HTTPS']) == 'on' ) ? 'https://' : 'http://';
  48                 $requested_url .= $_SERVER['HTTP_HOST'];
  49                 $requested_url .= $_SERVER['REQUEST_URI'];
  50         }
  51
  52         $original = @parse_url($requested_url);
  53         if ( false === $original )
  54                 return;
  55
  56         // Some PHP setups turn requests for / into /index.php in REQUEST_URI
  57         $original['path'] = preg_replace('|/index\.php$|', '/', $original['path']);
  58
  59         $redirect = $original;
  60         $redirect_url = false;
  61
  62         // These tests give us a WP-generated permalink
  63         if ( is_404() ) {
  64                 $redirect_url = redirect_guess_404_permalink();
  65         } elseif ( is_object($wp_rewrite) && $wp_rewrite->using_permalinks() ) {
  66                 // rewriting of old ?p=X, ?m=2004, ?m=200401, ?m=20040101
  67                 if ( is_single() && isset($_GET['p']) ) {
  68                         if ( $redirect_url = get_permalink(get_query_var('p')) )
  69                                 $redirect['query'] = remove_query_arg('p', $redirect['query']);
  70                 } elseif ( is_page() && isset($_GET['page_id']) ) {
  71                         if ( $redirect_url = get_permalink(get_query_var('page_id')) )
  72                                 $redirect['query'] = remove_query_arg('page_id', $redirect['query']);
  73                 } elseif ( isset($_GET['m']) && ( is_year() || is_month() || is_day() ) ) {
  74                         $m = get_query_var('m');
  75                         switch ( strlen($m) ) {
  76                                 case 4: // Yearly
  77                                         $redirect_url = get_year_link($m);
  78                                         break;
  79                                 case 6: // Monthly
  80                                         $redirect_url = get_month_link( substr($m, 0, 4), substr($m, 4, 2) );
  81                                         break;
  82                                 case 8: // Daily
  83                                         $redirect_url = get_day_link(substr($m, 0, 4), substr($m, 4, 2), substr($m, 6, 2));
  84                                         break;
  85                         }
  86                         if ( $redirect_url )
  87                                 $redirect['query'] = remove_query_arg('m', $redirect['query']);
  88                 // now moving on to non ?m=X year/month/day links
  89                 } elseif ( is_day() && get_query_var('year') && get_query_var('monthnum') && isset($_GET['day']) ) {
  90                         if ( $redirect_url = get_day_link(get_query_var('year'), get_query_var('monthnum'), get_query_var('day')) )
  91                                 $redirect['query'] = remove_query_arg(array('year', 'monthnum', 'day'), $redirect['query']);
  92                 } elseif ( is_month() && get_query_var('year') && isset($_GET['monthnum']) ) {
  93                         if ( $redirect_url = get_month_link(get_query_var('year'), get_query_var('monthnum')) )
  94                                 $redirect['query'] = remove_query_arg(array('year', 'monthnum'), $redirect['query']);
  95                 } elseif ( is_year() && isset($_GET['year']) ) {
  96                         if ( $redirect_url = get_year_link(get_query_var('year')) )
  97                                 $redirect['query'] = remove_query_arg('year', $redirect['query']);
  98                 } elseif ( is_category() && isset($_GET['cat']) ) {
  99                         if ( $redirect_url = get_category_link(get_query_var('cat')) )
 100                                 $redirect['query'] = remove_query_arg('cat', $redirect['query']);
 101                 } elseif ( is_author() && isset($_GET['author']) ) {
 102                         $author = get_userdata(get_query_var('author'));
 103                         if ( false !== $author && $redirect_url = get_author_link(false, $author->ID, $author->user_nicename) )
 104                                 $redirect['query'] = remove_query_arg('author', $redirect['author']);
 105                 }
 106
 107         // paging
 108                 if ( $paged = get_query_var('paged') ) {
 109                         if ( $paged > 0 ) {
 110                                 if ( !$redirect_url )
 111                                         $redirect_url = $requested_url;
 112                                 $paged_redirect = @parse_url($redirect_url);
 113                                 $paged_redirect['path'] = preg_replace('|/page/[0-9]+?(/+)?$|', '/', $paged_redirect['path']); // strip off any existing paging
 114                                 $paged_redirect['path'] = preg_replace('|/index.php/?$|', '/', $paged_redirect['path']); // strip off trailing /index.php/
 115                                 if ( $paged > 1 && !is_single() ) {
 116                                         $paged_redirect['path'] = trailingslashit($paged_redirect['path']);
 117                                         if ( $wp_rewrite->using_index_permalinks() && strpos($paged_redirect['path'], '/index.php/') === false )
 118                                                 $paged_redirect['path'] .= 'index.php/';
 119                                         $paged_redirect['path'] .= user_trailingslashit("page/$paged", 'paged');
 120                                 } elseif ( !is_home() && !is_single() ){
 121                                         $paged_redirect['path'] = user_trailingslashit($paged_redirect['path'], 'paged');
 122                                 }
 123                                 $redirect_url = $paged_redirect['scheme'] . '://' . $paged_redirect['host'] . $paged_redirect['path'];
 124                                 $redirect['path'] = $paged_redirect['path'];
 125                         }
 126                         $redirect['query'] = remove_query_arg('paged', $redirect['query']);
 127                 }
 128         }
 129
 130         // tack on any additional query vars
 131         if ( $redirect_url && $redirect['query'] ) {
 132                 if ( strpos($redirect_url, '?') !== false )
 133                         $redirect_url .= '&';
 134                 else
 135                         $redirect_url .= '?';
 136                 $redirect_url .= $redirect['query'];
 137         }
 138
 139         if ( $redirect_url )
 140                 $redirect = @parse_url($redirect_url);
 141
 142         // www.example.com vs example.com
 143         $user_home = @parse_url(get_option('home'));
 144         if ( isset($user_home['host']) )
 145                 $redirect['host'] = $user_home['host'];
 146
 147         // Handle ports
 148         if ( isset($user_home['port']) )
 149                 $redirect['port'] = $user_home['port'];
 150         else
 151                 unset($redirect['port']);
 152
 153         // trailing /index.php/
 154         $redirect['path'] = preg_replace('|/index.php/$|', '/', $redirect['path']);
 155
 156         // strip /index.php/ when we're not using PATHINFO permalinks
 157         if ( !$wp_rewrite->using_index_permalinks() )
 158                 $redirect['path'] = str_replace('/index.php/', '/', $redirect['path']);
 159
 160         // trailing slashes
 161         if ( is_object($wp_rewrite) && $wp_rewrite->using_permalinks() && !is_404() && (!is_home() || ( is_home() && (get_query_var('paged') > 1) ) ) ) {
 162                 $user_ts_type = '';
 163                 if ( get_query_var('paged') > 0 ) {
 164                         $user_ts_type = 'paged';
 165                 } else {
 166                         foreach ( array('single', 'category', 'page', 'day', 'month', 'year') as $type ) {
 167                                 $func = 'is_' . $type;
 168                                 if ( call_user_func($func) ) {
 169                                         $user_ts_type = $type;
 170                                         break;
 171                                 }
 172                         }
 173                 }
 174                 $redirect['path'] = user_trailingslashit($redirect['path'], $user_ts_type);
 175         } elseif ( is_home() ) {
 176                 $redirect['path'] = trailingslashit($redirect['path']);
 177         }
 178
 179         // Always trailing slash the 'home' URL
 180         if ( $redirect['path'] == $user_home['path'] )
 181                 $redirect['path'] = trailingslashit($redirect['path']);
 182
 183         // Ignore differences in host capitalization, as this can lead to infinite redirects
 184         if ( strtolower($original['host']) == strtolower($redirect['host']) )
 185                 $redirect['host'] = $original['host'];
 186
 187         if ( array($original['host'], $original['port'], $original['path'], $original['query']) !== array($redirect['host'], $redirect['port'], $redirect['path'], $redirect['query']) ) {
 188                 $redirect_url = $redirect['scheme'] . '://' . $redirect['host'];
 189                 if ( isset($redirect['port']) )
 190                         $redirect_url .= ':' . $redirect['port'];
 191                 $redirect_url .= $redirect['path'];
 192                 if ( $redirect['query'] )
 193                         $redirect_url .= '?' . $redirect['query'];
 194         }
 195
 196         if ( !$redirect_url || $redirect_url == $requested_url )
 197                 return false;
 198
 199         // Note that you can use the "redirect_canonical" filter to cancel a canonical redirect for whatever reason by returning FALSE
 200         $redirect_url = apply_filters('redirect_canonical', $redirect_url, $requested_url);
 201
 202         if ( !$redirect_url || $redirect_url == $requested_url ) // yes, again -- in case the filter aborted the request
 203                 return false;
 204
 205         if ( $do_redirect ) {
 206                 // protect against chained redirects
 207                 if ( !redirect_canonical($redirect_url, false) ) {
 208                         wp_redirect($redirect_url, 301);
 209                         exit();
 210                 } else {
 211                         return false;
 212                 }
 213         } else {
 214                 return $redirect_url;
 215         }
 216 }
 217
 218 /**
 219  * Attempts to guess correct post based on query vars
 220  *
 221  * @since 2.3
 222  * @uses $wpdb
 223  *
 224  * @return bool|string Returns False, if it can't find post, returns correct
 225  *              location on success.
 226  */
 227 function redirect_guess_404_permalink() {
 228         global $wpdb;
 229         if ( !get_query_var('name') )
 230                 return false;
 231
 232         $where = $wpdb->prepare("post_name LIKE %s", get_query_var('name') . '%');
 233
 234         // if any of year, monthnum, or day are set, use them to refine the query
 235         if ( get_query_var('year') )
 236                 $where .= $wpdb->prepare(" AND YEAR(post_date) = %d", get_query_var('year'));
 237         if ( get_query_var('monthnum') )
 238                 $where .= $wpdb->prepare(" AND MONTH(post_date) = %d", get_query_var('monthnum'));
 239         if ( get_query_var('day') )
 240                 $where .= $wpdb->prepare(" AND DAYOFMONTH(post_date) = %d", get_query_var('day'));
 241
 242         $post_id = $wpdb->get_var("SELECT ID FROM $wpdb->posts WHERE $where AND post_status = 'publish'");
 243         if ( !$post_id )
 244                 return false;
 245         return get_permalink($post_id);
 246 }
 247
 248 add_action('template_redirect', 'redirect_canonical');
 249
 250 ?>