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 namespace MediaWiki;
  24
  25 use MWTidy;
  26 use Html;
  27
  28 /**
  29  * @since 1.31
  30  */
  31 class OutputHandler {
  32         /**
  33          * Standard output handler for use with ob_start.
  34          *
  35          * @param string $s Web response output
  36          * @return string
  37          */
  38         public static function handle( $s ) {
  39                 global $wgDisableOutputCompression, $wgValidateAllHtml, $wgMangleFlashPolicy;
  40                 if ( $wgMangleFlashPolicy ) {
  41                         $s = self::mangleFlashPolicy( $s );
  42                 }
  43                 if ( $wgValidateAllHtml ) {
  44                         $headers = headers_list();
  45                         $isHTML = false;
  46                         foreach ( $headers as $header ) {
  47                                 $parts = explode( ':', $header, 2 );
  48                                 if ( count( $parts ) !== 2 ) {
  49                                         continue;
  50                                 }
  51                                 $name = strtolower( trim( $parts[0] ) );
  52                                 $value = trim( $parts[1] );
  53                                 if ( $name == 'content-type' && ( strpos( $value, 'text/html' ) === 0
  54                                         || strpos( $value, 'application/xhtml+xml' ) === 0 )
  55                                 ) {
  56                                         $isHTML = true;
  57                                         break;
  58                                 }
  59                         }
  60                         if ( $isHTML ) {
  61                                 $s = self::validateAllHtml( $s );
  62                         }
  63                 }
  64                 if ( !$wgDisableOutputCompression && !ini_get( 'zlib.output_compression' ) ) {
  65                         if ( !defined( 'MW_NO_OUTPUT_COMPRESSION' ) ) {
  66                                 $s = self::handleGzip( $s );
  67                         }
  68                         if ( !ini_get( 'output_handler' ) ) {
  69                                 self::emitContentLength( strlen( $s ) );
  70                         }
  71                 }
  72                 return $s;
  73         }
  74
  75         /**
  76          * Get the "file extension" that some client apps will estimate from
  77          * the currently-requested URL.
  78          *
  79          * This isn't a WebRequest method, because we need it before the class loads.
  80          * @todo As of 2018, this actually runs after autoloader in Setup.php, so
  81          * WebRequest seems like a good place for this.
  82          *
  83          * @return string
  84          */
  85         private static function findUriExtension() {
  86                 /// @todo FIXME: this sort of dupes some code in WebRequest::getRequestUrl()
  87                 if ( isset( $_SERVER['REQUEST_URI'] ) ) {
  88                         // Strip the query string...
  89                         list( $path ) = explode( '?', $_SERVER['REQUEST_URI'], 2 );
  90                 } elseif ( isset( $_SERVER['SCRIPT_NAME'] ) ) {
  91                         // Probably IIS. QUERY_STRING appears separately.
  92                         $path = $_SERVER['SCRIPT_NAME'];
  93                 } else {
  94                         // Can't get the path from the server? :(
  95                         return '';
  96                 }
  97
  98                 $period = strrpos( $path, '.' );
  99                 if ( $period !== false ) {
 100                         return strtolower( substr( $path, $period ) );
 101                 }
 102                 return '';
 103         }
 104
 105         /**
 106          * Handler that compresses data with gzip if allowed by the Accept header.
 107          *
 108          * Unlike ob_gzhandler, it works for HEAD requests too.
 109          *
 110          * @param string $s Web response output
 111          * @return string
 112          */
 113         private static function handleGzip( $s ) {
 114                 if ( !function_exists( 'gzencode' ) ) {
 115                         wfDebug( __METHOD__ . "() skipping compression (gzencode unavailable)\n" );
 116                         return $s;
 117                 }
 118                 if ( headers_sent() ) {
 119                         wfDebug( __METHOD__ . "() skipping compression (headers already sent)\n" );
 120                         return $s;
 121                 }
 122
 123                 $ext = self::findUriExtension();
 124                 if ( $ext == '.gz' || $ext == '.tgz' ) {
 125                         // Don't do gzip compression if the URL path ends in .gz or .tgz
 126                         // This confuses Safari and triggers a download of the page,
 127                         // even though it's pretty clearly labeled as viewable HTML.
 128                         // Bad Safari! Bad!
 129                         return $s;
 130                 }
 131
 132                 if ( wfClientAcceptsGzip() ) {
 133                         wfDebug( __METHOD__ . "() is compressing output\n" );
 134                         header( 'Content-Encoding: gzip' );
 135                         $s = gzencode( $s, 6 );
 136                 }
 137
 138                 // Set vary header if it hasn't been set already
 139                 $headers = headers_list();
 140                 $foundVary = false;
 141                 foreach ( $headers as $header ) {
 142                         $headerName = strtolower( substr( $header, 0, 5 ) );
 143                         if ( $headerName == 'vary:' ) {
 144                                 $foundVary = true;
 145                                 break;
 146                         }
 147                 }
 148                 if ( !$foundVary ) {
 149                         header( 'Vary: Accept-Encoding' );
 150                         global $wgUseKeyHeader;
 151                         if ( $wgUseKeyHeader ) {
 152                                 header( 'Key: Accept-Encoding;match=gzip' );
 153                         }
 154                 }
 155                 return $s;
 156         }
 157
 158         /**
 159          * Mangle flash policy tags which open up the site to XSS attacks.
 160          *
 161          * @param string $s Web response output
 162          * @return string
 163          */
 164         private static function mangleFlashPolicy( $s ) {
 165                 # Avoid weird excessive memory usage in PCRE on big articles
 166                 if ( preg_match( '/\<\s*cross-domain-policy(?=\s|\>)/i', $s ) ) {
 167                         return preg_replace( '/\<(\s*)(cross-domain-policy(?=\s|\>))/i', '<$1NOT-$2', $s );
 168                 } else {
 169                         return $s;
 170                 }
 171         }
 172
 173         /**
 174          * Add a Content-Length header if possible. This makes it cooperate with CDN better.
 175          *
 176          * @param int $length
 177          */
 178         private static function emitContentLength( $length ) {
 179                 if ( !headers_sent()
 180                         && isset( $_SERVER['SERVER_PROTOCOL'] )
 181                         && $_SERVER['SERVER_PROTOCOL'] == 'HTTP/1.0'
 182                 ) {
 183                         header( "Content-Length: $length" );
 184                 }
 185         }
 186
 187         /**
 188          * Replace the output with an error if the HTML is not valid.
 189          *
 190          * @param string $s
 191          * @return string
 192          */
 193         private static function validateAllHtml( $s ) {
 194                 $errors = '';
 195                 if ( MWTidy::checkErrors( $s, $errors ) ) {
 196                         return $s;
 197                 }
 198
 199                 header( 'Cache-Control: no-cache' );
 200
 201                 $out = Html::element( 'h1', null, 'HTML validation error' );
 202                 $out .= Html::openElement( 'ul' );
 203
 204                 $error = strtok( $errors, "\n" );
 205                 $badLines = [];
 206                 while ( $error !== false ) {
 207                         if ( preg_match( '/^line (\d+)/', $error, $m ) ) {
 208                                 $lineNum = intval( $m[1] );
 209                                 $badLines[$lineNum] = true;
 210                                 $out .= Html::rawElement( 'li', null,
 211                                         Html::element( 'a', [ 'href' => "#line-{$lineNum}" ], $error ) ) . "\n";
 212                         }
 213                         $error = strtok( "\n" );
 214                 }
 215
 216                 $out .= Html::closeElement( 'ul' );
 217                 $out .= Html::element( 'pre', null, $errors );
 218                 $out .= Html::openElement( 'ol' ) . "\n";
 219                 $line = strtok( $s, "\n" );
 220                 $i = 1;
 221                 while ( $line !== false ) {
 222                         $attrs = [];
 223                         if ( isset( $badLines[$i] ) ) {
 224                                 $attrs['class'] = 'highlight';
 225                                 $attrs['id'] = "line-$i";
 226                         }
 227                         $out .= Html::element( 'li', $attrs, $line ) . "\n";
 228                         $line = strtok( "\n" );
 229                         $i++;
 230                 }
 231                 $out .= Html::closeElement( 'ol' );
 232
 233                 $style = <<<CSS
 234         .highlight { background-color: #ffc }
 235         li { white-space: pre }
 236 CSS;
 237
 238                 $out = Html::htmlHeader( [ 'lang' => 'en', 'dir' => 'ltr' ] ) .
 239                         Html::rawElement( 'head', null,
 240                                 Html::element( 'title', null, 'HTML validation error' ) .
 241                                 Html::inlineStyle( $style ) ) .
 242                         Html::rawElement( 'body', null, $out ) .
 243                         Html::closeElement( 'html' );
 244
 245                 return $out;
 246         }
 247 }