includes/OutputHandler.php

   1 <?php
   2 /**
   3  * Functions to be used with PHP's output buffer.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  */
  22
  23 /**
  24  * Standard output handler for use with ob_start
  25  *
  26  * @param string $s
  27  *
  28  * @return string
  29  */
  30 function wfOutputHandler( $s ) {
  31         global $wgDisableOutputCompression, $wgValidateAllHtml, $wgMangleFlashPolicy;
  32         if ( $wgMangleFlashPolicy ) {
  33                 $s = wfMangleFlashPolicy( $s );
  34         }
  35         if ( $wgValidateAllHtml ) {
  36                 $headers = headers_list();
  37                 $isHTML = false;
  38                 foreach ( $headers as $header ) {
  39                         $parts = explode( ':', $header, 2 );
  40                         if ( count( $parts ) !== 2 ) {
  41                                 continue;
  42                         }
  43                         $name = strtolower( trim( $parts[0] ) );
  44                         $value = trim( $parts[1] );
  45                         if ( $name == 'content-type' && ( strpos( $value, 'text/html' ) === 0
  46                                 || strpos( $value, 'application/xhtml+xml' ) === 0 )
  47                         ) {
  48                                 $isHTML = true;
  49                                 break;
  50                         }
  51                 }
  52                 if ( $isHTML ) {
  53                         $s = wfHtmlValidationHandler( $s );
  54                 }
  55         }
  56         if ( !$wgDisableOutputCompression && !ini_get( 'zlib.output_compression' ) ) {
  57                 if ( !defined( 'MW_NO_OUTPUT_COMPRESSION' ) ) {
  58                         $s = wfGzipHandler( $s );
  59                 }
  60                 if ( !ini_get( 'output_handler' ) ) {
  61                         wfDoContentLength( strlen( $s ) );
  62                 }
  63         }
  64         return $s;
  65 }
  66
  67 /**
  68  * Get the "file extension" that some client apps will estimate from
  69  * the currently-requested URL.
  70  * This isn't on WebRequest because we need it when things aren't initialized
  71  * @private
  72  *
  73  * @return string
  74  */
  75 function wfRequestExtension() {
  76         /// @todo FIXME: this sort of dupes some code in WebRequest::getRequestUrl()
  77         if ( isset( $_SERVER['REQUEST_URI'] ) ) {
  78                 // Strip the query string...
  79                 list( $path ) = explode( '?', $_SERVER['REQUEST_URI'], 2 );
  80         } elseif ( isset( $_SERVER['SCRIPT_NAME'] ) ) {
  81                 // Probably IIS. QUERY_STRING appears separately.
  82                 $path = $_SERVER['SCRIPT_NAME'];
  83         } else {
  84                 // Can't get the path from the server? :(
  85                 return '';
  86         }
  87
  88         $period = strrpos( $path, '.' );
  89         if ( $period !== false ) {
  90                 return strtolower( substr( $path, $period ) );
  91         }
  92         return '';
  93 }
  94
  95 /**
  96  * Handler that compresses data with gzip if allowed by the Accept header.
  97  * Unlike ob_gzhandler, it works for HEAD requests too.
  98  *
  99  * @param string $s
 100  *
 101  * @return string
 102  */
 103 function wfGzipHandler( $s ) {
 104         if ( !function_exists( 'gzencode' ) ) {
 105                 wfDebug( __FUNCTION__ . "() skipping compression (gzencode unavailable)\n" );
 106                 return $s;
 107         }
 108         if ( headers_sent() ) {
 109                 wfDebug( __FUNCTION__ . "() skipping compression (headers already sent)\n" );
 110                 return $s;
 111         }
 112
 113         $ext = wfRequestExtension();
 114         if ( $ext == '.gz' || $ext == '.tgz' ) {
 115                 // Don't do gzip compression if the URL path ends in .gz or .tgz
 116                 // This confuses Safari and triggers a download of the page,
 117                 // even though it's pretty clearly labeled as viewable HTML.
 118                 // Bad Safari! Bad!
 119                 return $s;
 120         }
 121
 122         if ( wfClientAcceptsGzip() ) {
 123                 wfDebug( __FUNCTION__ . "() is compressing output\n" );
 124                 header( 'Content-Encoding: gzip' );
 125                 $s = gzencode( $s, 6 );
 126         }
 127
 128         // Set vary header if it hasn't been set already
 129         $headers = headers_list();
 130         $foundVary = false;
 131         foreach ( $headers as $header ) {
 132                 if ( substr( $header, 0, 5 ) == 'Vary:' ) {
 133                         $foundVary = true;
 134                         break;
 135                 }
 136         }
 137         if ( !$foundVary ) {
 138                 header( 'Vary: Accept-Encoding' );
 139                 global $wgUseXVO;
 140                 if ( $wgUseXVO ) {
 141                         header( 'X-Vary-Options: Accept-Encoding;list-contains=gzip' );
 142                 }
 143         }
 144         return $s;
 145 }
 146
 147 /**
 148  * Mangle flash policy tags which open up the site to XSS attacks.
 149  *
 150  * @param string $s
 151  *
 152  * @return string
 153  */
 154 function wfMangleFlashPolicy( $s ) {
 155         # Avoid weird excessive memory usage in PCRE on big articles
 156         if ( preg_match( '/\<\s*cross-domain-policy\s*\>/i', $s ) ) {
 157                 return preg_replace( '/\<\s*cross-domain-policy\s*\>/i', '<NOT-cross-domain-policy>', $s );
 158         } else {
 159                 return $s;
 160         }
 161 }
 162
 163 /**
 164  * Add a Content-Length header if possible. This makes it cooperate with squid better.
 165  *
 166  * @param int $length
 167  */
 168 function wfDoContentLength( $length ) {
 169         if ( !headers_sent()
 170                 && isset( $_SERVER['SERVER_PROTOCOL'] )
 171                 && $_SERVER['SERVER_PROTOCOL'] == 'HTTP/1.0'
 172         ) {
 173                 header( "Content-Length: $length" );
 174         }
 175 }
 176
 177 /**
 178  * Replace the output with an error if the HTML is not valid
 179  *
 180  * @param string $s
 181  *
 182  * @return string
 183  */
 184 function wfHtmlValidationHandler( $s ) {
 185
 186         $errors = '';
 187         if ( MWTidy::checkErrors( $s, $errors ) ) {
 188                 return $s;
 189         }
 190
 191         header( 'Cache-Control: no-cache' );
 192
 193         $out = Html::element( 'h1', null, 'HTML validation error' );
 194         $out .= Html::openElement( 'ul' );
 195
 196         $error = strtok( $errors, "\n" );
 197         $badLines = array();
 198         while ( $error !== false ) {
 199                 if ( preg_match( '/^line (\d+)/', $error, $m ) ) {
 200                         $lineNum = intval( $m[1] );
 201                         $badLines[$lineNum] = true;
 202                         $out .= Html::rawElement( 'li', null,
 203                                 Html::element( 'a', array( 'href' => "#line-{$lineNum}" ), $error ) ) . "\n";
 204                 }
 205                 $error = strtok( "\n" );
 206         }
 207
 208         $out .= Html::closeElement( 'ul' );
 209         $out .= Html::element( 'pre', null, $errors );
 210         $out .= Html::openElement( 'ol' ) . "\n";
 211         $line = strtok( $s, "\n" );
 212         $i = 1;
 213         while ( $line !== false ) {
 214                 $attrs = array();
 215                 if ( isset( $badLines[$i] ) ) {
 216                         $attrs['class'] = 'highlight';
 217                         $attrs['id'] = "line-$i";
 218                 }
 219                 $out .= Html::element( 'li', $attrs, $line ) . "\n";
 220                 $line = strtok( "\n" );
 221                 $i++;
 222         }
 223         $out .= Html::closeElement( 'ol' );
 224
 225         $style = <<<CSS
 226 .highlight { background-color: #ffc }
 227 li { white-space: pre }
 228 CSS;
 229
 230         $out = Html::htmlHeader( array( 'lang' => 'en', 'dir' => 'ltr' ) ) .
 231                 Html::rawElement( 'head', null,
 232                         Html::element( 'title', null, 'HTML validation error' ) .
 233                         Html::inlineStyle( $style ) ) .
 234                 Html::rawElement( 'body', null, $out ) .
 235                 Html::closeElement( 'html' );
 236
 237         return $out;
 238 }