includes/Tidy.php

   1 <?php
   2 /**
   3  * Interface with html tidy, used if $wgUseTidy = true from Parser::parse.
   4  * If tidy isn't able to correct the markup, the original will be
   5  * returned in all its glory with a warning comment appended.
   6  *
   7 */
   8 class Tidy {
   9          /*
  10          * Either the external tidy program or the in-process tidy extension
  11          * will be used depending on availability. Override the default
  12          * $wgTidyInternal setting to disable the internal if it's not working.
  13          * @param string $text Hideous HTML input
  14          * @return string Corrected HTML output
  15          * @public
  16          * @static
  17          */
  18         public static function RunOn( $text ) {
  19                 global $wgTidyInternal;
  20                 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
  21 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
  22 '<head><title>test</title></head><body>'.$text.'</body></html>';
  23                 if( $wgTidyInternal ) {
  24                         $correctedtext = Tidy::internal( $wrappedtext );
  25                 } else {
  26                         $correctedtext = Tidy::external( $wrappedtext );
  27                 }
  28                 if( is_null( $correctedtext ) ) {
  29                         wfDebug( "Tidy error detected!\n" );
  30                         return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
  31                 }
  32                 return $correctedtext;
  33         }
  34
  35         /**
  36          * Spawn an external HTML tidy process and get corrected markup back from it.
  37          */
  38         private static function external( $text ) {
  39                 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
  40                 $fname = 'Parser::externalTidy';
  41                 wfProfileIn( $fname );
  42
  43                 $cleansource = '';
  44                 $opts = ' -utf8';
  45
  46                 $descriptorspec = array(
  47                         0 => array('pipe', 'r'),
  48                         1 => array('pipe', 'w'),
  49                         2 => array('file', '/dev/null', 'a')  // FIXME: this line in UNIX-specific, it generates a warning on Windows, because /dev/null is not a valid Windows file.
  50                 );
  51                 $pipes = array();
  52                 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
  53                 if (is_resource($process)) {
  54                         // Theoretically, this style of communication could cause a deadlock
  55                         // here. If the stdout buffer fills up, then writes to stdin could
  56                         // block. This doesn't appear to happen with tidy, because tidy only
  57                         // writes to stdout after it's finished reading from stdin. Search
  58                         // for tidyParseStdin and tidySaveStdout in console/tidy.c
  59                         fwrite($pipes[0], $text);
  60                         fclose($pipes[0]);
  61                         while (!feof($pipes[1])) {
  62                                 $cleansource .= fgets($pipes[1], 1024);
  63                         }
  64                         fclose($pipes[1]);
  65                         proc_close($process);
  66                 }
  67
  68                 wfProfileOut( $fname );
  69
  70                 if( $cleansource == '' && $text != '') {
  71                         // Some kind of error happened, so we couldn't get the corrected text.
  72                         // Just give up; we'll use the source text and append a warning.
  73                         return null;
  74                 } else {
  75                         return $cleansource;
  76                 }
  77         }
  78
  79         /**
  80          * Use the HTML tidy PECL extension to use the tidy library in-process,
  81          * saving the overhead of spawning a new process. Currently written to
  82          * the PHP 4.3.x version of the extension, may not work on PHP 5.
  83          *
  84          * 'pear install tidy' should be able to compile the extension module.
  85          */
  86         private static function internal( $text ) {
  87                 global $wgTidyConf;
  88                 $fname = 'Parser::internalTidy';
  89                 wfProfileIn( $fname );
  90
  91                 tidy_load_config( $wgTidyConf );
  92                 tidy_set_encoding( 'utf8' );
  93                 tidy_parse_string( $text );
  94                 tidy_clean_repair();
  95                 if( tidy_get_status() == 2 ) {
  96                         // 2 is magic number for fatal error
  97                         // http://www.php.net/manual/en/function.tidy-get-status.php
  98                         $cleansource = null;
  99                 } else {
 100                         $cleansource = tidy_get_output();
 101                 }
 102                 wfProfileOut( $fname );
 103                 return $cleansource;
 104         }
 105 }
 106 ?>