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