-style tags. This should not be anything we
# may want to use in wikisyntax
define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
# prefix for escaping, used in two functions at least
define( 'UNIQ_PREFIX', 'NaodW29');
# Constants needed for external link processing
define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
define( 'HTTP_PROTOCOLS', 'http|https' );
# Everything except bracket, space, or control characters
define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
# Including space
define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
define( 'EXT_LINK_BRACKETED', '/\[(\b('.URL_PROTOCOLS.'):'.EXT_LINK_URL_CLASS.'+) *('.EXT_LINK_TEXT_CLASS.'*?)\]/S' );
define( 'EXT_IMAGE_REGEX',
'/^('.HTTP_PROTOCOLS.':)'. # Protocol
'('.EXT_LINK_URL_CLASS.'+)\\/'. # Hostname and path
'('.EXT_IMAGE_FNAME_CLASS.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS.')$/S' # Filename
);
/**
* PHP Parser
*
* Processes wiki markup
*
*
* There are three main entry points into the Parser class:
* parse()
* produces HTML output
* preSaveTransform().
* produces altered wiki markup.
* transformMsg()
* performs brace substitution on MediaWiki messages
*
* Globals used:
* objects: $wgLang, $wgDateFormatter, $wgLinkCache
*
* NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
*
* settings:
* $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
* $wgNamespacesWithSubpages, $wgAllowExternalImages*,
* $wgLocaltimezone
*
* * only within ParserOptions
*
*
* @package MediaWiki
*/
class Parser
{
/**#@+
* @access private
*/
# Persistent:
var $mTagHooks;
# Cleared with clearState():
var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
# Temporary:
var $mOptions, $mTitle, $mOutputType,
$mTemplates, // cache of already loaded templates, avoids
// multiple SQL queries for the same string
$mTemplatePath; // stores an unsorted hash of all the templates already loaded
// in this path. Used for loop detection.
/**#@-*/
/**
* Constructor
*
* @access public
*/
function Parser() {
$this->mTemplates = array();
$this->mTemplatePath = array();
$this->mTagHooks = array();
$this->clearState();
}
/**
* Clear Parser state
*
* @access private
*/
function clearState() {
$this->mOutput = new ParserOutput;
$this->mAutonumber = 0;
$this->mLastSection = '';
$this->mDTopen = false;
$this->mVariables = false;
$this->mIncludeCount = array();
$this->mStripState = array();
$this->mArgStack = array();
$this->mInPre = false;
}
/**
* First pass--just handle sections, pass the rest off
* to internalParse() which does all the real work.
*
* @access private
* @param string $text Text we want to parse
* @param Title &$title A title object
* @param array $options
* @param boolean $linestart
* @param boolean $clearState
* @return ParserOutput a ParserOutput
*/
function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
global $wgUseTidy, $wgContLang;
$fname = 'Parser::parse';
wfProfileIn( $fname );
if ( $clearState ) {
$this->clearState();
}
$this->mOptions = $options;
$this->mTitle =& $title;
$this->mOutputType = OT_HTML;
$stripState = NULL;
global $fnord; $fnord = 1;
//$text = $this->strip( $text, $this->mStripState );
// VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
$x =& $this->mStripState;
$text = $this->strip( $text, $x );
$text = $this->internalParse( $text, $linestart );
$dashReplace = array(
'/ - /' => " – ", # N dash
'/(?<=[0-9])-(?=[0-9])/' => "–", # N dash between numbers
'/ -- /' => " — " # M dash
);
$text = preg_replace( array_keys($dashReplace), array_values($dashReplace), $text );
$text = $this->unstrip( $text, $this->mStripState );
# Clean up special characters, only run once, next-to-last before doBlockLevels
global $wgUseTidy;
if(!$wgUseTidy) {
$fixtags = array(
# french spaces, last one Guillemet-left
# only if there is something before the space
'/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1 \\2',
# french spaces, Guillemet-right
'/(\\302\\253) /' => '\\1 ',
'/
/i' => '
',
'/
/i' => '
',
'//i' => '',
'/<\\/center *>/i' => '
',
);
$text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
$text = Sanitizer::normalizeCharReferences( $text );
} else {
$fixtags = array(
# french spaces, last one Guillemet-left
'/ (\\?|:|;|!|\\302\\273)/' => ' \\1',
# french spaces, Guillemet-right
'/(\\302\\253) /' => '\\1 ',
'//i' => '',
'/<\\/center *>/i' => '
'
);
$text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
}
# only once and last
$text = $this->doBlockLevels( $text, $linestart );
$this->replaceLinkHolders( $text );
$text = $wgContLang->convert($text);
$this->mOutput->setTitleText($wgContLang->getParsedTitle());
$text = $this->unstripNoWiki( $text, $this->mStripState );
if ($wgUseTidy) {
$text = Parser::tidy($text);
}
$this->mOutput->setText( $text );
wfProfileOut( $fname );
return $this->mOutput;
}
/**
* Get a random string
*
* @access private
* @static
*/
function getRandomString() {
return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
}
/**
* Replaces all occurrences of <$tag>content$tag> in the text
* with a random marker and returns the new text. the output parameter
* $content will be an associative array filled with data on the form
* $unique_marker => content.
*
* If $content is already set, the additional entries will be appended
* If $tag is set to STRIP_COMMENTS, the function will extract
*
*
* @access private
* @static
*/
function extractTags($tag, $text, &$content, $uniq_prefix = ''){
$rnd = $uniq_prefix . '-' . $tag . Parser::getRandomString();
if ( !$content ) {
$content = array( );
}
$n = 1;
$stripped = '';
while ( '' != $text ) {
if($tag==STRIP_COMMENTS) {
$p = preg_split( '//i', $p[1], 2 );
} else {
$q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
}
$marker = $rnd . sprintf('%08X', $n++);
$content[$marker] = $q[0];
$stripped .= $marker;
$text = $q[1];
}
}
return $stripped;
}
/**
* Strips and renders nowiki, pre, math, hiero
* If $render is set, performs necessary rendering operations on plugins
* Returns the text, and fills an array with data needed in unstrip()
* If the $state is already a valid strip state, it adds to the state
*
* @param bool $stripcomments when set, HTML comments
* will be stripped in addition to other tags. This is important
* for section editing, where these comments cause confusion when
* counting the sections in the wikisource
*
* @access private
*/
function strip( $text, &$state, $stripcomments = false ) {
$render = ($this->mOutputType == OT_HTML);
$html_content = array();
$nowiki_content = array();
$math_content = array();
$pre_content = array();
$comment_content = array();
$ext_content = array();
$gallery_content = array();
# Replace any instances of the placeholders
$uniq_prefix = UNIQ_PREFIX;
#$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
# html
global $wgRawHtml, $wgWhitelistEdit;
if( $wgRawHtml && $wgWhitelistEdit ) {
$text = Parser::extractTags('html', $text, $html_content, $uniq_prefix);
foreach( $html_content as $marker => $content ) {
if ($render ) {
# Raw and unchecked for validity.
$html_content[$marker] = $content;
} else {
$html_content[$marker] = ''.$content.'';
}
}
}
# nowiki
$text = Parser::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
foreach( $nowiki_content as $marker => $content ) {
if( $render ){
$nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
} else {
$nowiki_content[$marker] = ''.$content.'';
}
}
# math
$text = Parser::extractTags('math', $text, $math_content, $uniq_prefix);
foreach( $math_content as $marker => $content ){
if( $render ) {
if( $this->mOptions->getUseTeX() ) {
$math_content[$marker] = renderMath( $content );
} else {
$math_content[$marker] = '<math>'.$content.'<math>';
}
} else {
$math_content[$marker] = '';
}
}
# pre
$text = Parser::extractTags('pre', $text, $pre_content, $uniq_prefix);
foreach( $pre_content as $marker => $content ){
if( $render ){
$pre_content[$marker] = '' . wfEscapeHTMLTagsOnly( $content ) . '
';
} else {
$pre_content[$marker] = ''.$content.'
';
}
}
# gallery
$text = Parser::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
foreach( $gallery_content as $marker => $content ) {
require_once( 'ImageGallery.php' );
if ( $render ) {
$gallery_content[$marker] = Parser::renderImageGallery( $content );
} else {
$gallery_content[$marker] = ''.$content.'';
}
}
# Comments
if($stripcomments) {
$text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
foreach( $comment_content as $marker => $content ){
$comment_content[$marker] = '';
}
}
# Extensions
foreach ( $this->mTagHooks as $tag => $callback ) {
$ext_contents[$tag] = array();
$text = Parser::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
foreach( $ext_content[$tag] as $marker => $content ) {
if ( $render ) {
$ext_content[$tag][$marker] = $callback( $content );
} else {
$ext_content[$tag][$marker] = "<$tag>$content$tag>";
}
}
}
# Merge state with the pre-existing state, if there is one
if ( $state ) {
$state['html'] = $state['html'] + $html_content;
$state['nowiki'] = $state['nowiki'] + $nowiki_content;
$state['math'] = $state['math'] + $math_content;
$state['pre'] = $state['pre'] + $pre_content;
$state['comment'] = $state['comment'] + $comment_content;
$state['gallery'] = $state['gallery'] + $gallery_content;
foreach( $ext_content as $tag => $array ) {
if ( array_key_exists( $tag, $state ) ) {
$state[$tag] = $state[$tag] + $array;
}
}
} else {
$state = array(
'html' => $html_content,
'nowiki' => $nowiki_content,
'math' => $math_content,
'pre' => $pre_content,
'comment' => $comment_content,
'gallery' => $gallery_content,
) + $ext_content;
}
return $text;
}
/**
* restores pre, math, and hiero removed by strip()
*
* always call unstripNoWiki() after this one
* @access private
*/
function unstrip( $text, &$state ) {
# Must expand in reverse order, otherwise nested tags will be corrupted
$contentDict = end( $state );
for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
if( key($state) != 'nowiki' && key($state) != 'html') {
for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
$text = str_replace( key( $contentDict ), $content, $text );
}
}
}
return $text;
}
/**
* always call this after unstrip() to preserve the order
*
* @access private
*/
function unstripNoWiki( $text, &$state ) {
# Must expand in reverse order, otherwise nested tags will be corrupted
for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
$text = str_replace( key( $state['nowiki'] ), $content, $text );
}
global $wgRawHtml;
if ($wgRawHtml) {
for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
$text = str_replace( key( $state['html'] ), $content, $text );
}
}
return $text;
}
/**
* Add an item to the strip state
* Returns the unique tag which must be inserted into the stripped text
* The tag will be replaced with the original text in unstrip()
*
* @access private
*/
function insertStripItem( $text, &$state ) {
$rnd = UNIQ_PREFIX . '-item' . Parser::getRandomString();
if ( !$state ) {
$state = array(
'html' => array(),
'nowiki' => array(),
'math' => array(),
'pre' => array()
);
}
$state['item'][$rnd] = $text;
return $rnd;
}
/**
* interface with html tidy, used if $wgUseTidy = true
*
* @access public
* @static
*/
function tidy ( $text ) {
global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
$fname = 'Parser::tidy';
wfProfileIn( $fname );
$cleansource = '';
$opts = ' -utf8';
$wrappedtext = ''.
'test'.$text.'';
$descriptorspec = array(
0 => array('pipe', 'r'),
1 => array('pipe', 'w'),
2 => array('file', '/dev/null', 'a')
);
$process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
if (is_resource($process)) {
fwrite($pipes[0], $wrappedtext);
fclose($pipes[0]);
while (!feof($pipes[1])) {
$cleansource .= fgets($pipes[1], 1024);
}
fclose($pipes[1]);
$return_value = proc_close($process);
}
wfProfileOut( $fname );
if( $cleansource == '' && $text != '') {
wfDebug( "Tidy error detected!\n" );
return $text . "\n\n";
} else {
return $cleansource;
}
}
/**
* parse the wiki syntax used to render tables
*
* @access private
*/
function doTableStuff ( $t ) {
$fname = 'Parser::doTableStuff';
wfProfileIn( $fname );
$t = explode ( "\n" , $t ) ;
$td = array () ; # Is currently a td tag open?
$ltd = array () ; # Was it TD or TH?
$tr = array () ; # Is currently a tr tag open?
$ltr = array () ; # tr attributes
$indent_level = 0; # indent level of the table
foreach ( $t AS $k => $x )
{
$x = trim ( $x ) ;
$fc = substr ( $x , 0 , 1 ) ;
if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
$indent_level = strlen( $matches[1] );
$t[$k] = str_repeat( '- ', $indent_level ) .
'
' ;
array_push ( $td , false ) ;
array_push ( $ltd , '' ) ;
array_push ( $tr , false ) ;
array_push ( $ltr , '' ) ;
}
else if ( count ( $td ) == 0 ) { } # Don't do any of the following
else if ( '|}' == substr ( $x , 0 , 2 ) ) {
$z = "
" . substr ( $x , 2);
$l = array_pop ( $ltd ) ;
if ( array_pop ( $tr ) ) $z = '' . $z ;
if ( array_pop ( $td ) ) $z = ''.$l.'>' . $z ;
array_pop ( $ltr ) ;
$t[$k] = $z . str_repeat( '
', $indent_level );
}
else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
$x = substr ( $x , 1 ) ;
while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
$z = '' ;
$l = array_pop ( $ltd ) ;
if ( array_pop ( $tr ) ) $z = '' . $z ;
if ( array_pop ( $td ) ) $z = ''.$l.'>' . $z ;
array_pop ( $ltr ) ;
$t[$k] = $z ;
array_push ( $tr , false ) ;
array_push ( $td , false ) ;
array_push ( $ltd , '' ) ;
array_push ( $ltr , Sanitizer::fixTagAttributes ( $x, 'tr' ) ) ;
}
else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) { # Caption
# $x is a table row
if ( '|+' == substr ( $x , 0 , 2 ) ) {
$fc = '+' ;
$x = substr ( $x , 1 ) ;
}
$after = substr ( $x , 1 ) ;
if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
$after = explode ( '||' , $after ) ;
$t[$k] = '' ;
# Loop through each table cell
foreach ( $after AS $theline )
{
$z = '' ;
if ( $fc != '+' )
{
$tra = array_pop ( $ltr ) ;
if ( !array_pop ( $tr ) ) $z = '\n" ;
array_push ( $tr , true ) ;
array_push ( $ltr , '' ) ;
}
$l = array_pop ( $ltd ) ;
if ( array_pop ( $td ) ) $z = ''.$l.'>' . $z ;
if ( $fc == '|' ) $l = 'td' ;
else if ( $fc == '!' ) $l = 'th' ;
else if ( $fc == '+' ) $l = 'caption' ;
else $l = '' ;
array_push ( $ltd , $l ) ;
# Cell parameters
$y = explode ( '|' , $theline , 2 ) ;
# Note that a '|' inside an invalid link should not
# be mistaken as delimiting cell parameters
if ( strpos( $y[0], '[[' ) !== false ) {
$y = array ($theline);
}
if ( count ( $y ) == 1 )
$y = "{$z}<{$l}>{$y[0]}" ;
else $y = $y = "{$z}<{$l}".Sanitizer::fixTagAttributes($y[0], $l).">{$y[1]}" ;
$t[$k] .= $y ;
array_push ( $td , true ) ;
}
}
}
# Closing open td, tr && table
while ( count ( $td ) > 0 )
{
if ( array_pop ( $td ) ) $t[] = '' ;
if ( array_pop ( $tr ) ) $t[] = '
' ;
$t[] = '' ;
}
$t = implode ( "\n" , $t ) ;
# $t = Sanitizer::removeHTMLtags( $t );
wfProfileOut( $fname );
return $t ;
}
/**
* Helper function for parse() that transforms wiki markup into
* HTML. Only called for $mOutputType == OT_HTML.
*
* @access private
*/
function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
global $wgContLang;
$fname = 'Parser::internalParse';
wfProfileIn( $fname );
$text = Sanitizer::removeHTMLtags( $text );
$text = $this->replaceVariables( $text, $args );
$text = preg_replace( '/(^|\n)-----*/', '\\1
', $text );
$text = $this->doHeadings( $text );
if($this->mOptions->getUseDynamicDates()) {
global $wgDateFormatter;
$text = $wgDateFormatter->reformat( $this->mOptions->getDateFormat(), $text );
}
$text = $this->doAllQuotes( $text );
$text = $this->replaceInternalLinks( $text );
$text = $this->replaceExternalLinks( $text );
# replaceInternalLinks may sometimes leave behind
# absolute URLs, which have to be masked to hide them from replaceExternalLinks
$text = str_replace("http-noparse://","http://",$text);
$text = $this->doMagicLinks( $text );
$text = $this->doTableStuff( $text );
$text = $this->formatHeadings( $text, $isMain );
wfProfileOut( $fname );
return $text;
}
/**
* Replace special strings like "ISBN xxx" and "RFC xxx" with
* magic external links.
*
* @access private
*/
function &doMagicLinks( &$text ) {
$text = $this->magicISBN( $text );
$text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
$text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
return $text;
}
/**
* Parse ^^ tokens and return html
*
* @access private
*/
function doExponent( $text ) {
$fname = 'Parser::doExponent';
wfProfileIn( $fname );
$text = preg_replace('/\^\^(.*)\^\^/','\\1', $text);
wfProfileOut( $fname );
return $text;
}
/**
* Parse headers and return html
*
* @access private
*/
function doHeadings( $text ) {
$fname = 'Parser::doHeadings';
wfProfileIn( $fname );
for ( $i = 6; $i >= 1; --$i ) {
$h = substr( '======', 0, $i );
$text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
"\\1\\2", $text );
}
wfProfileOut( $fname );
return $text;
}
/**
* Replace single quotes with HTML markup
* @access private
* @return string the altered text
*/
function doAllQuotes( $text ) {
$fname = 'Parser::doAllQuotes';
wfProfileIn( $fname );
$outtext = '';
$lines = explode( "\n", $text );
foreach ( $lines as $line ) {
$outtext .= $this->doQuotes ( $line ) . "\n";
}
$outtext = substr($outtext, 0,-1);
wfProfileOut( $fname );
return $outtext;
}
/**
* Helper function for doAllQuotes()
* @access private
*/
function doQuotes( $text ) {
$arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
if ( count( $arr ) == 1 )
return $text;
else
{
# First, do some preliminary work. This may shift some apostrophes from
# being mark-up to being text. It also counts the number of occurrences
# of bold and italics mark-ups.
$i = 0;
$numbold = 0;
$numitalics = 0;
foreach ( $arr as $r )
{
if ( ( $i % 2 ) == 1 )
{
# If there are ever four apostrophes, assume the first is supposed to
# be text, and the remaining three constitute mark-up for bold text.
if ( strlen( $arr[$i] ) == 4 )
{
$arr[$i-1] .= "'";
$arr[$i] = "'''";
}
# If there are more than 5 apostrophes in a row, assume they're all
# text except for the last 5.
else if ( strlen( $arr[$i] ) > 5 )
{
$arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
$arr[$i] = "'''''";
}
# Count the number of occurrences of bold and italics mark-ups.
# We are not counting sequences of five apostrophes.
if ( strlen( $arr[$i] ) == 2 ) $numitalics++; else
if ( strlen( $arr[$i] ) == 3 ) $numbold++; else
if ( strlen( $arr[$i] ) == 5 ) { $numitalics++; $numbold++; }
}
$i++;
}
# If there is an odd number of both bold and italics, it is likely
# that one of the bold ones was meant to be an apostrophe followed
# by italics. Which one we cannot know for certain, but it is more
# likely to be one that has a single-letter word before it.
if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) )
{
$i = 0;
$firstsingleletterword = -1;
$firstmultiletterword = -1;
$firstspace = -1;
foreach ( $arr as $r )
{
if ( ( $i % 2 == 1 ) and ( strlen( $r ) == 3 ) )
{
$x1 = substr ($arr[$i-1], -1);
$x2 = substr ($arr[$i-1], -2, 1);
if ($x1 == ' ') {
if ($firstspace == -1) $firstspace = $i;
} else if ($x2 == ' ') {
if ($firstsingleletterword == -1) $firstsingleletterword = $i;
} else {
if ($firstmultiletterword == -1) $firstmultiletterword = $i;
}
}
$i++;
}
# If there is a single-letter word, use it!
if ($firstsingleletterword > -1)
{
$arr [ $firstsingleletterword ] = "''";
$arr [ $firstsingleletterword-1 ] .= "'";
}
# If not, but there's a multi-letter word, use that one.
else if ($firstmultiletterword > -1)
{
$arr [ $firstmultiletterword ] = "''";
$arr [ $firstmultiletterword-1 ] .= "'";
}
# ... otherwise use the first one that has neither.
# (notice that it is possible for all three to be -1 if, for example,
# there is only one pentuple-apostrophe in the line)
else if ($firstspace > -1)
{
$arr [ $firstspace ] = "''";
$arr [ $firstspace-1 ] .= "'";
}
}
# Now let's actually convert our apostrophic mush to HTML!
$output = '';
$buffer = '';
$state = '';
$i = 0;
foreach ($arr as $r)
{
if (($i % 2) == 0)
{
if ($state == 'both')
$buffer .= $r;
else
$output .= $r;
}
else
{
if (strlen ($r) == 2)
{
if ($state == 'i')
{ $output .= ''; $state = ''; }
else if ($state == 'bi')
{ $output .= ''; $state = 'b'; }
else if ($state == 'ib')
{ $output .= ''; $state = 'b'; }
else if ($state == 'both')
{ $output .= ''.$buffer.''; $state = 'b'; }
else # $state can be 'b' or ''
{ $output .= ''; $state .= 'i'; }
}
else if (strlen ($r) == 3)
{
if ($state == 'b')
{ $output .= ''; $state = ''; }
else if ($state == 'bi')
{ $output .= ''; $state = 'i'; }
else if ($state == 'ib')
{ $output .= ''; $state = 'i'; }
else if ($state == 'both')
{ $output .= ''.$buffer.''; $state = 'i'; }
else # $state can be 'i' or ''
{ $output .= ''; $state .= 'b'; }
}
else if (strlen ($r) == 5)
{
if ($state == 'b')
{ $output .= ''; $state = 'i'; }
else if ($state == 'i')
{ $output .= ''; $state = 'b'; }
else if ($state == 'bi')
{ $output .= ''; $state = ''; }
else if ($state == 'ib')
{ $output .= ''; $state = ''; }
else if ($state == 'both')
{ $output .= ''.$buffer.''; $state = ''; }
else # ($state == '')
{ $buffer = ''; $state = 'both'; }
}
}
$i++;
}
# Now close all remaining tags. Notice that the order is important.
if ($state == 'b' || $state == 'ib')
$output .= '';
if ($state == 'i' || $state == 'bi' || $state == 'ib')
$output .= '';
if ($state == 'bi')
$output .= '';
if ($state == 'both')
$output .= ''.$buffer.'';
return $output;
}
}
/**
* Replace external links
*
* Note: this is all very hackish and the order of execution matters a lot.
* Make sure to run maintenance/parserTests.php if you change this code.
*
* @access private
*/
function replaceExternalLinks( $text ) {
$fname = 'Parser::replaceExternalLinks';
wfProfileIn( $fname );
$sk =& $this->mOptions->getSkin();
global $wgContLang;
$linktrail = $wgContLang->linkTrail();
$bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
$s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
$i = 0;
while ( $i' (which were escaped by
# removeHTMLtags()) should not be included in
# URLs, per RFC 2396.
if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
$text = substr($url, $m2[0][1]) . ' ' . $text;
$url = substr($url, 0, $m2[0][1]);
}
# If the link text is an image URL, replace it with an
tag
# This happened by accident in the original parser, but some people used it extensively
$img = $this->maybeMakeImageLink( $text );
if ( $img !== false ) {
$text = $img;
}
$dtrail = '';
# Set linktype for CSS - if URL==text, link is essentially free
$linktype = ($text == $url) ? 'free' : 'text';
# No link text, e.g. [http://domain.tld/some.link]
if ( $text == '' ) {
# Autonumber if allowed
if ( strpos( HTTP_PROTOCOLS, $protocol ) !== false ) {
$text = '[' . ++$this->mAutonumber . ']';
$linktype = 'autonumber';
} else {
# Otherwise just use the URL
$text = htmlspecialchars( $url );
$linktype = 'free';
}
} else {
# Have link text, e.g. [http://domain.tld/some.link text]s
# Check for trail
if ( preg_match( $linktrail, $trail, $m2 ) ) {
$dtrail = $m2[1];
$trail = $m2[2];
}
}
# Replace & from obsolete syntax with &.
# All HTML entities will be escaped by makeExternalLink()
# or maybeMakeImageLink()
$url = str_replace( '&', '&', $url );
# Process the trail (i.e. everything after this link up until start of the next link),
# replacing any non-bracketed links
$trail = $this->replaceFreeExternalLinks( $trail );
# Use the encoded URL
# This means that users can paste URLs directly into the text
# Funny characters like ö aren't valid in URLs anyway
# This was changed in August 2004
$s .= $sk->makeExternalLink( $url, $text, false, $linktype ) . $dtrail . $trail;
}
wfProfileOut( $fname );
return $s;
}
/**
* Replace anything that looks like a URL with a link
* @access private
*/
function replaceFreeExternalLinks( $text ) {
$fname = 'Parser::replaceFreeExternalLinks';
wfProfileIn( $fname );
$bits = preg_split( '/(\b(?:'.URL_PROTOCOLS.'):)/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
$s = array_shift( $bits );
$i = 0;
$sk =& $this->mOptions->getSkin();
while ( $i < count( $bits ) ){
$protocol = $bits[$i++];
$remainder = $bits[$i++];
if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
# Found some characters after the protocol that look promising
$url = $protocol . $m[1];
$trail = $m[2];
# The characters '<' and '>' (which were escaped by
# removeHTMLtags()) should not be included in
# URLs, per RFC 2396.
if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
$trail = substr($url, $m2[0][1]) . $trail;
$url = substr($url, 0, $m2[0][1]);
}
# Move trailing punctuation to $trail
$sep = ',;\.:!?';
# If there is no left bracket, then consider right brackets fair game too
if ( strpos( $url, '(' ) === false ) {
$sep .= ')';
}
$numSepChars = strspn( strrev( $url ), $sep );
if ( $numSepChars ) {
$trail = substr( $url, -$numSepChars ) . $trail;
$url = substr( $url, 0, -$numSepChars );
}
# Replace & from obsolete syntax with &.
# All HTML entities will be escaped by makeExternalLink()
# or maybeMakeImageLink()
$url = str_replace( '&', '&', $url );
# Is this an external image?
$text = $this->maybeMakeImageLink( $url );
if ( $text === false ) {
# Not an image, make a link
$text = $sk->makeExternalLink( $url, $url, true, 'free' );
}
$s .= $text . $trail;
} else {
$s .= $protocol . $remainder;
}
}
wfProfileOut();
return $s;
}
/**
* make an image if it's allowed
* @access private
*/
function maybeMakeImageLink( $url ) {
$sk =& $this->mOptions->getSkin();
$text = false;
if ( $this->mOptions->getAllowExternalImages() ) {
if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
# Image found
$text = $sk->makeImage( htmlspecialchars( $url ) );
}
}
return $text;
}
/**
* Process [[ ]] wikilinks
*
* @access private
*/
function replaceInternalLinks( $s ) {
global $wgLang, $wgContLang, $wgLinkCache;
global $wgDisableLangConversion;
static $fname = 'Parser::replaceInternalLinks' ;
wfProfileIn( $fname );
wfProfileIn( $fname.'-setup' );
static $tc = FALSE;
# the % is needed to support urlencoded titles as well
if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
$sk =& $this->mOptions->getSkin();
global $wgUseOldExistenceCheck;
# "Post-parse link colour check" works only on wiki text since it's now
# in Parser. Enable it, then disable it when we're done.
$saveParseColour = $sk->postParseLinkColour( !$wgUseOldExistenceCheck );
$redirect = MagicWord::get ( MAG_REDIRECT ) ;
#split the entire text string on occurences of [[
$a = explode( '[[', ' ' . $s );
#get the first element (all text up to first [[), and remove the space we added
$s = array_shift( $a );
$s = substr( $s, 1 );
# Match a link having the form [[namespace:link|alternate]]trail
static $e1 = FALSE;
if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
# Match cases where there is no "]]", which might still be images
static $e1_img = FALSE;
if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
# Match the end of a line for a word that's not followed by whitespace,
# e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
$useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
if( is_null( $this->mTitle ) ) {
wfDebugDieBacktrace( 'nooo' );
}
$nottalk = !$this->mTitle->isTalkPage();
if ( $useLinkPrefixExtension ) {
if ( preg_match( $e2, $s, $m ) ) {
$first_prefix = $m[2];
$s = $m[1];
} else {
$first_prefix = false;
}
} else {
$prefix = '';
}
$selflink = $this->mTitle->getPrefixedText();
wfProfileOut( $fname.'-setup' );
$checkVariantLink = sizeof($wgContLang->getVariants())>1;
$useSubpages = $this->areSubpagesAllowed();
# Loop for each link
for ($k = 0; isset( $a[$k] ); $k++) {
$line = $a[$k];
if ( $useLinkPrefixExtension ) {
wfProfileIn( $fname.'-prefixhandling' );
if ( preg_match( $e2, $s, $m ) ) {
$prefix = $m[2];
$s = $m[1];
} else {
$prefix='';
}
# first link
if($first_prefix) {
$prefix = $first_prefix;
$first_prefix = false;
}
wfProfileOut( $fname.'-prefixhandling' );
}
$might_be_img = false;
if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
$text = $m[2];
# fix up urlencoded title texts
if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
$trail = $m[3];
} elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
$might_be_img = true;
$text = $m[2];
if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
$trail = "";
} else { # Invalid form; output directly
$s .= $prefix . '[[' . $line ;
continue;
}
# Don't allow internal links to pages containing
# PROTO: where PROTO is a valid URL protocol; these
# should be external links.
if (preg_match('/^(\b(?:'.URL_PROTOCOLS.'):)/', $m[1])) {
$s .= $prefix . '[[' . $line ;
continue;
}
# Make subpage if necessary
if( $useSubpages ) {
$link = $this->maybeDoSubpageLink( $m[1], $text );
} else {
$link = $m[1];
}
$noforce = (substr($m[1], 0, 1) != ':');
if (!$noforce) {
# Strip off leading ':'
$link = substr($link, 1);
}
$nt =& Title::newFromText( $this->unstripNoWiki($link, $this->mStripState) );
if( !$nt ) {
$s .= $prefix . '[[' . $line;
continue;
}
#check other language variants of the link
#if the article does not exist
if( $checkVariantLink
&& $nt->getArticleID() == 0 ) {
$wgContLang->findVariantLink($link, $nt);
}
$ns = $nt->getNamespace();
$iw = $nt->getInterWiki();
if ($might_be_img) { # if this is actually an invalid link
if ($ns == NS_IMAGE && $noforce) { #but might be an image
$found = false;
while (isset ($a[$k+1]) ) {
#look at the next 'line' to see if we can close it there
$next_line = array_shift(array_splice( $a, $k + 1, 1) );
if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
# the first ]] closes the inner link, the second the image
$found = true;
$text .= '[[' . $m[1];
$trail = $m[2];
break;
} elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
#if there's exactly one ]] that's fine, we'll keep looking
$text .= '[[' . $m[0];
} else {
#if $next_line is invalid too, we need look no further
$text .= '[[' . $next_line;
break;
}
}
if ( !$found ) {
# we couldn't find the end of this imageLink, so output it raw
#but don't ignore what might be perfectly normal links in the text we've examined
$text = $this->replaceInternalLinks($text);
$s .= $prefix . '[[' . $link . '|' . $text;
# note: no $trail, because without an end, there *is* no trail
continue;
}
} else { #it's not an image, so output it raw
$s .= $prefix . '[[' . $link . '|' . $text;
# note: no $trail, because without an end, there *is* no trail
continue;
}
}
$wasblank = ( '' == $text );
if( $wasblank ) $text = $link;
# Link not escaped by : , create the various objects
if( $noforce ) {
# Interwikis
if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
array_push( $this->mOutput->mLanguageLinks, $nt->getFullText() );
$tmp = $prefix . $trail ;
$s .= (trim($tmp) == '')? '': $tmp;
continue;
}
if ( $ns == NS_IMAGE ) {
wfProfileIn( "$fname-image" );
# recursively parse links inside the image caption
# actually, this will parse them in any other parameters, too,
# but it might be hard to fix that, and it doesn't matter ATM
$text = $this->replaceExternalLinks($text);
$text = $this->replaceInternalLinks($text);
# cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
$s .= $prefix . str_replace('http://', 'http-noparse://', $sk->makeImageLinkObj( $nt, $text ) ) . $trail;
$wgLinkCache->addImageLinkObj( $nt );
wfProfileOut( "$fname-image" );
continue;
}
if ( $ns == NS_CATEGORY ) {
wfProfileIn( "$fname-category" );
$t = $nt->getText();
$wgLinkCache->suspend(); # Don't save in links/brokenlinks
$pPLC=$sk->postParseLinkColour();
$sk->postParseLinkColour( false );
$t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
$sk->postParseLinkColour( $pPLC );
$wgLinkCache->resume();
if ( $wasblank ) {
if ( $this->mTitle->getNamespace() == NS_CATEGORY ) {
$sortkey = $this->mTitle->getText();
} else {
$sortkey = $this->mTitle->getPrefixedText();
}
} else {
$sortkey = $text;
}
$wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
$this->mOutput->addCategoryLink( $t );
/**
* Strip the whitespace Category links produce, see bug 87
* @todo We might want to use trim($tmp, "\n") here.
*/
$tmp = $prefix . $trail;
$s .= trim($tmp) == '' ? '': $tmp;
wfProfileOut( "$fname-category" );
continue;
}
}
if( ( $nt->getPrefixedText() === $selflink ) &&
( $nt->getFragment() === '' ) ) {
# Self-links are handled specially; generally de-link and change to bold.
$s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
continue;
}
# Special and Media are pseudo-namespaces; no pages actually exist in them
if( $ns == NS_MEDIA ) {
$s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
$wgLinkCache->addImageLinkObj( $nt );
continue;
} elseif( $ns == NS_SPECIAL ) {
$s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
continue;
}
$s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
}
$sk->postParseLinkColour( $saveParseColour );
wfProfileOut( $fname );
return $s;
}
/**
* Return true if subpage links should be expanded on this page.
* @return bool
*/
function areSubpagesAllowed() {
# Some namespaces don't allow subpages
global $wgNamespacesWithSubpages;
return !empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()]);
}
/**
* Handle link to subpage if necessary
* @param string $target the source of the link
* @param string &$text the link text, modified as necessary
* @return string the full name of the link
* @access private
*/
function maybeDoSubpageLink($target, &$text) {
# Valid link forms:
# Foobar -- normal
# :Foobar -- override special treatment of prefix (images, language links)
# /Foobar -- convert to CurrentPage/Foobar
# /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
# ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
# ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
$fname = 'Parser::maybeDoSubpageLink';
wfProfileIn( $fname );
$ret = $target; # default return value is no change
# Some namespaces don't allow subpages,
# so only perform processing if subpages are allowed
if( $this->areSubpagesAllowed() ) {
# Look at the first character
if( $target != '' && $target{0} == '/' ) {
# / at end means we don't want the slash to be shown
if( substr( $target, -1, 1 ) == '/' ) {
$target = substr( $target, 1, -1 );
$noslash = $target;
} else {
$noslash = substr( $target, 1 );
}
$ret = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
if( '' === $text ) {
$text = $target;
} # this might be changed for ugliness reasons
} else {
# check for .. subpage backlinks
$dotdotcount = 0;
$nodotdot = $target;
while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
++$dotdotcount;
$nodotdot = substr( $nodotdot, 3 );
}
if($dotdotcount > 0) {
$exploded = explode( '/', $this->mTitle->GetPrefixedText() );
if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
$ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
# / at the end means don't show full path
if( substr( $nodotdot, -1, 1 ) == '/' ) {
$nodotdot = substr( $nodotdot, 0, -1 );
if( '' === $text ) {
$text = $nodotdot;
}
}
$nodotdot = trim( $nodotdot );
if( $nodotdot != '' ) {
$ret .= '/' . $nodotdot;
}
}
}
}
}
wfProfileOut( $fname );
return $ret;
}
/**#@+
* Used by doBlockLevels()
* @access private
*/
/* private */ function closeParagraph() {
$result = '';
if ( '' != $this->mLastSection ) {
$result = '' . $this->mLastSection . ">\n";
}
$this->mInPre = false;
$this->mLastSection = '';
return $result;
}
# getCommon() returns the length of the longest common substring
# of both arguments, starting at the beginning of both.
#
/* private */ function getCommon( $st1, $st2 ) {
$fl = strlen( $st1 );
$shorter = strlen( $st2 );
if ( $fl < $shorter ) { $shorter = $fl; }
for ( $i = 0; $i < $shorter; ++$i ) {
if ( $st1{$i} != $st2{$i} ) { break; }
}
return $i;
}
# These next three functions open, continue, and close the list
# element appropriate to the prefix character passed into them.
#
/* private */ function openList( $char ) {
$result = $this->closeParagraph();
if ( '*' == $char ) { $result .= '- '; }
else if ( '#' == $char ) { $result .= '
- '; }
else if ( ':' == $char ) { $result .= '
- '; }
else if ( ';' == $char ) {
$result .= '
- ';
$this->mDTopen = true;
}
else { $result = ''; }
return $result;
}
/* private */ function nextItem( $char ) {
if ( '*' == $char || '#' == $char ) { return '
- '; }
else if ( ':' == $char || ';' == $char ) {
$close = '';
if ( $this->mDTopen ) { $close = ''; }
if ( ';' == $char ) {
$this->mDTopen = true;
return $close . '
- ';
} else {
$this->mDTopen = false;
return $close . '
- ';
}
}
return '';
}
/* private */ function closeList( $char ) {
if ( '*' == $char ) { $text = '
'; }
else if ( '#' == $char ) { $text = ''; }
else if ( ':' == $char ) {
if ( $this->mDTopen ) {
$this->mDTopen = false;
$text = '';
} else {
$text = '';
}
}
else { return ''; }
return $text."\n";
}
/**#@-*/
/**
* Make lists from lines starting with ':', '*', '#', etc.
*
* @access private
* @return string the lists rendered as HTML
*/
function doBlockLevels( $text, $linestart ) {
$fname = 'Parser::doBlockLevels';
wfProfileIn( $fname );
# Parsing through the text line by line. The main thing
# happening here is handling of block-level elements p, pre,
# and making lists from lines starting with * # : etc.
#
$textLines = explode( "\n", $text );
$lastPrefix = $output = $lastLine = '';
$this->mDTopen = $inBlockElem = false;
$prefixLength = 0;
$paragraphStack = false;
if ( !$linestart ) {
$output .= array_shift( $textLines );
}
foreach ( $textLines as $oLine ) {
$lastPrefixLength = strlen( $lastPrefix );
$preCloseMatch = preg_match('/<\\/pre/i', $oLine );
$preOpenMatch = preg_match('/mInPre ) {
# Multiple prefixes may abut each other for nested lists.
$prefixLength = strspn( $oLine, '*#:;' );
$pref = substr( $oLine, 0, $prefixLength );
# eh?
$pref2 = str_replace( ';', ':', $pref );
$t = substr( $oLine, $prefixLength );
$this->mInPre = !empty($preOpenMatch);
} else {
# Don't interpret any other prefixes in preformatted text
$prefixLength = 0;
$pref = $pref2 = '';
$t = $oLine;
}
# List generation
if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
# Same as the last item, so no need to deal with nesting or opening stuff
$output .= $this->nextItem( substr( $pref, -1 ) );
$paragraphStack = false;
if ( substr( $pref, -1 ) == ';') {
# The one nasty exception: definition lists work like this:
# ; title : definition text
# So we check for : in the remainder text to split up the
# title and definition, without b0rking links.
if ($this->findColonNoLinks($t, $term, $t2) !== false) {
$t = $t2;
$output .= $term . $this->nextItem( ':' );
}
}
} elseif( $prefixLength || $lastPrefixLength ) {
# Either open or close a level...
$commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
$paragraphStack = false;
while( $commonPrefixLength < $lastPrefixLength ) {
$output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
--$lastPrefixLength;
}
if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
$output .= $this->nextItem( $pref{$commonPrefixLength-1} );
}
while ( $prefixLength > $commonPrefixLength ) {
$char = substr( $pref, $commonPrefixLength, 1 );
$output .= $this->openList( $char );
if ( ';' == $char ) {
# FIXME: This is dupe of code above
if ($this->findColonNoLinks($t, $term, $t2) !== false) {
$t = $t2;
$output .= $term . $this->nextItem( ':' );
}
}
++$commonPrefixLength;
}
$lastPrefix = $pref2;
}
if( 0 == $prefixLength ) {
wfProfileIn( "$fname-paragraph" );
# No prefix (not in list)--go to paragraph mode
$uniq_prefix = UNIQ_PREFIX;
// XXX: use a stack for nestable elements like span, table and div
$openmatch = preg_match('/(