// // showdown.js -- A javascript port of Markdown. // // Modifications // Copyright (c) 2011 Chris Spencer // // PHP Markdown Extra // Copyright (c) 2009 Michel Fortin // // node-markdown // Copyright (c) 2010 Andris Reinman // // Original showdown.js // Copyright (c) 2007 John Fraser. // // Original Markdown Copyright (c) 2004-2005 John Gruber // // // Redistributable under a BSD-style open source license. // See license.txt for more information. // // The full source distribution is at: // // A A L // T C A // T K B // // // // // Wherever possible, Showdown is a straight, line-by-line port // of the Perl version of Markdown. // // This is not a normal parser design; it's basically just a // series of string substitutions. It's hard to read and // maintain this way, but keeping Showdown close to the original // design makes it easier to port new features. // // More importantly, Showdown behaves like markdown.pl in most // edge cases. So web applications can do client-side preview // in Javascript, and then build identical HTML on the server. // // This port needs the new RegExp functionality of ECMA 262, // 3rd Edition (i.e. Javascript 1.5). Most modern web browsers // should do fine. Even with the new regular expression features, // We do a lot of work to emulate Perl's regex functionality. // The tricky changes in this file mostly have the "attacklab:" // label. Major or self-explanatory changes don't. // // Smart diff tools like Araxis Merge will be able to match up // this file with markdown.pl in a useful way. A little tweaking // helps: in a copy of markdown.pl, replace "#" with "//" and // replace "$text" with "text". Be sure to ignore whitespace // and line endings. // // // Showdown usage: // // var text = "Markdown *rocks*."; // // var converter = new Showdown.converter(); // var html = converter.makeHtml(text); // // alert(html); // // Note: move the sample code to the bottom of this // file before uncommenting it. // // // Showdown namespace // var Showdown = {}; // // converter // // Wraps all "globals" so that the only thing // exposed is makeHtml(). // // accepts an array of config options that will merge with the config array below Showdown.converter = function() { // // Globals: // // Global hashes, used by various utility routines var g_urls; var g_titles; var g_html_blocks; var g_print_refs; var g_print_refs_count; // Used to track when we're inside an ordered or unordered list // (see _ProcessListItems() for details): var g_list_level = 0; var config = this.config = { stripHTML: false, headerautoid: false, tables: false, math: false, figures: false, refprint: false, github_flavouring: false, additional_parser: false } // loop through each argument passed and set it as a config option if (arguments[0]) { for (x in arguments[0]) { this.config[x] = arguments[0][x]; } } this.makeHtml = function(text) { // // Main function. The order in which other subs are called here is // essential. Link and image substitutions need to happen before // _EscapeSpecialCharsWithinTagAttributes(), so that any *'s or _'s in the // and tags get encoded. // // Clear the global hashes. If we don't clear these, you get conflicts // from other articles when generating a page which contains more than // one article (e.g. an index page that shows the N most recent // articles): g_urls = new Array(); g_titles = new Array(); g_html_blocks = new Array(); g_print_refs = {}; g_print_refs_count = 0; // attacklab: Replace ~ with ~T // This lets us use tilde as an escape char to avoid md5 hashes // The choice of character is arbitray; anything that isn't // magic in Markdown will work. text = text.replace(/~/g,"~T"); // attacklab: Replace $ with ~D // RegExp interprets $ as a special character // when it's in a replacement string text = text.replace(/\$/g,"~D"); // Standardize line endings text = text.replace(/\r\n/g,"\n"); // DOS to Unix text = text.replace(/\r/g,"\n"); // Mac to Unix // Make sure text begins and ends with a couple of newlines: text = "\n\n" + text + "\n\n"; // Convert all tabs to spaces. text = _Detab(text); // Strip any lines consisting only of spaces and tabs. // This makes subsequent regexen easier to write, because we can // match consecutive blank lines with /\n+/ instead of something // contorted like /[ \t]*\n+/ . text = text.replace(/^[ \t]+$/mg,""); // Turn block-level HTML blocks into hash entries text = _HashHTMLBlocks(text); // Strip link definitions, store in hashes. text = _StripLinkDefinitions(text); text = _RunBlockGamut(text); text = _UnescapeSpecialChars(text); // attacklab: Restore dollar signs text = text.replace(/~D/g,"$$"); // attacklab: Restore tildes text = text.replace(/~T/g,"~"); if (config.stripHTML) { text = stripUnwantedHTML(text); } if (config.refprint && g_print_refs_count) { var link_table = '
    '; for (i in g_print_refs) { link_table += '
  • [' + g_print_refs[i] + ']: ' + i + '
  • ' } link_table += '
'; text += link_table; } if (typeof config.additional_parser == 'function') { text = config.additional_parser(text); } return text; } var _StripLinkDefinitions = function(text) { // // Strips link definitions from text, stores the URLs and titles in // hash references. // // Link defs are in the form: ^[id]: url "optional title" /* var text = text.replace(/ ^[ ]{0,3}\[(.+)\]: // id = $1 attacklab: g_tab_width - 1 [ \t]* \n? // maybe *one* newline [ \t]* ? // url = $2 [ \t]* \n? // maybe one newline [ \t]* (?: (\n*) // any lines skipped = $3 attacklab: lookbehind removed ["(] (.+?) // title = $4 [")] [ \t]* )? // title is optional (?:\n+|$) /gm, function(){...}); */ var text = text.replace(/^[ ]{0,3}\[(.+)\]:[ \t]*\n?[ \t]*?[ \t]*\n?[ \t]*(?:(\n*)["(](.+?)[")][ \t]*)?(?:\n+|\Z)/gm, function (wholeMatch,m1,m2,m3,m4) { m1 = m1.toLowerCase(); g_urls[m1] = _EncodeAmpsAndAngles(m2); // Link IDs are case-insensitive if (m3) { // Oops, found blank lines, so it's not a title. // Put back the parenthetical statement we stole. return m3+m4; } else if (m4) { g_titles[m1] = m4.replace(/"/g,"""); } // Completely remove the definition from the text return ""; } ); return text; } var _HashHTMLBlocks = function(text) { // attacklab: Double up blank lines to reduce lookaround text = text.replace(/\n/g,"\n\n"); // Hashify HTML blocks: // We only want to do this for block-level HTML tags, such as headers, // lists, and tables. That's because we still want to wrap

s around // "paragraphs" that are wrapped in non-block-level tags, such as anchors, // phrase emphasis, and spans. The list of tags we're looking for is // hard-coded: var block_tags_a = "p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math|ins|del" var block_tags_b = "p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math" // First, look for nested blocks, e.g.: //

//
// tags for inner block must be indented. //
//
// // The outermost tags must start at the left margin for this to match, and // the inner nested divs must be indented. // We need to do this before the next, more liberal match, because the next // match will start at the first `
` and stop at the first `
`. // attacklab: This regex can be expensive when it fails. /* var text = text.replace(/ ( // save in $1 ^ // start of line (with /m) <($block_tags_a) // start tag = $2 \b // word break // attacklab: hack around khtml/pcre bug... [^\r]*?\n // any number of lines, minimally matching // the matching end tag [ \t]* // trailing spaces/tabs (?=\n+) // followed by a newline ) // attacklab: there are sentinel newlines at end of document /gm,function(){...}}; */ text = text.replace(/^(<(p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math|ins|del)\b[^\r]*?\n<\/\2>[ \t]*(?=\n+))/gm,hashElement); // // Now match more liberally, simply from `\n` to `\n` // /* var text = text.replace(/ ( // save in $1 ^ // start of line (with /m) <($block_tags_b) // start tag = $2 \b // word break // attacklab: hack around khtml/pcre bug... [^\r]*? // any number of lines, minimally matching .* // the matching end tag [ \t]* // trailing spaces/tabs (?=\n+) // followed by a newline ) // attacklab: there are sentinel newlines at end of document /gm,function(){...}}; */ text = text.replace(/^(<(p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math)\b[^\r]*?.*<\/\2>[ \t]*(?=\n+)\n)/gm,hashElement); // Special case just for
. It was easier to make a special case than // to make the other regex more complicated. /* text = text.replace(/ ( // save in $1 \n\n // Starting after a blank line [ ]{0,3} (<(hr) // start tag = $2 \b // word break ([^<>])*? // \/?>) // the matching end tag [ \t]* (?=\n{2,}) // followed by a blank line ) /g,hashElement); */ text = text.replace(/(\n[ ]{0,3}(<(hr)\b([^<>])*?\/?>)[ \t]*(?=\n{2,}))/g,hashElement); // Special case for standalone HTML comments: /* text = text.replace(/ ( // save in $1 \n\n // Starting after a blank line [ ]{0,3} // attacklab: g_tab_width - 1 [ \t]* (?=\n{2,}) // followed by a blank line ) /g,hashElement); */ text = text.replace(/(\n\n[ ]{0,3}[ \t]*(?=\n{2,}))/g,hashElement); // PHP and ASP-style processor instructions ( and <%...%>) /* text = text.replace(/ (?: \n\n // Starting after a blank line ) ( // save in $1 [ ]{0,3} // attacklab: g_tab_width - 1 (?: <([?%]) // $2 [^\r]*? \2> ) [ \t]* (?=\n{2,}) // followed by a blank line ) /g,hashElement); */ text = text.replace(/(?:\n\n)([ ]{0,3}(?:<([?%])[^\r]*?\2>)[ \t]*(?=\n{2,}))/g,hashElement); // attacklab: Undo double lines (see comment at top of this function) text = text.replace(/\n\n/g,"\n"); return text; } var hashElement = function(wholeMatch,m1) { var blockText = m1; // Undo double lines blockText = blockText.replace(/\n\n/g,"\n"); blockText = blockText.replace(/^\n/,""); // strip trailing blank lines blockText = blockText.replace(/\n+$/g,""); // Replace the element text with a marker ("~KxK" where x is its key) blockText = "\n\n~K" + (g_html_blocks.push(blockText)-1) + "K\n\n"; return blockText; }; var _RunBlockGamut = function(text) { // // These are all the transformations that form block-level // tags like paragraphs, headers, and list items. // text = _DoHeaders(text); // Escape pipes early for tables. text = text.replace(/\\([\|])/g,escapeCharacters_callback); if (config.tables) { text = _DoTables(text); } // Do Horizontal Rules: var key = hashBlock("
"); text = text.replace(/^[ ]{0,2}([ ]?\*[ ]?){3,}[ \t]*$/gm,key); text = text.replace(/^[ ]{0,2}([ ]?\-[ ]?){3,}[ \t]*$/gm,key); text = text.replace(/^[ ]{0,2}([ ]?\_[ ]?){3,}[ \t]*$/gm,key); text = _DoLists(text); text = _DoCodeBlocks(text); text = _DoBlockQuotes(text); // We already ran _HashHTMLBlocks() before, in Markdown(), but that // was to escape raw HTML in the original Markdown source. This time, // we're escaping the markup we've just created, so that we don't wrap //

tags around block-level tags. text = _HashHTMLBlocks(text); text = _FormParagraphs(text); return text; } var _RunSpanGamut = function(text) { // // These are all the transformations that occur *within* block-level // tags like paragraphs, headers, and list items. // text = _DoCodeSpans(text); text = _EscapeSpecialCharsWithinTagAttributes(text); text = _EncodeBackslashEscapes(text); // Process anchor and image tags. Images must come first, // because ![foo][f] looks like an anchor. text = _DoImages(text); text = _DoAnchors(text); // Make links out of things like `` // Must come after _DoAnchors(), because you can use < and > // delimiters in inline links like [this](). text = _DoAutoLinks(text); text = _EncodeAmpsAndAngles(text); text = _DoTextStyle(text); // Do hard breaks: text = text.replace(/ +\n/g,"
\n"); return text; } var _EscapeSpecialCharsWithinTagAttributes = function(text) { // // Within tags -- meaning between < and > -- encode [\ ` * _] so they // don't conflict with their use in Markdown for code, italics and strong. // // Build a regex to find HTML tags and comments. See Friedl's // "Mastering Regular Expressions", 2nd Ed., pp. 200-201. var regex = /(<[a-z\/!$]("[^"]*"|'[^']*'|[^'">])*>|)/gi; text = text.replace(regex, function(wholeMatch) { var tag = wholeMatch.replace(/(.)<\/?code>(?=.)/g,"$1`"); tag = escapeCharacters(tag,"\\`*_"); return tag; }); return text; } var _DoAnchors = function(text) { // // Turn Markdown link shortcuts into XHTML
tags. // // // First, handle reference-style links: [link text] [id] // /* text = text.replace(/ ( // wrap whole match in $1 \[ ( (?: \[[^\]]*\] // allow brackets nested one level | [^\[] // or anything else )* ) \] [ ]? // one optional space (?:\n[ ]*)? // one optional newline followed by spaces \[ (.*?) // id = $3 \] )()()()() // pad remaining backreferences /g,_DoAnchors_callback); */ text = text.replace(/(\[((?:\[[^\]]*\]|[^\[\]])*)\][ ]?(?:\n[ ]*)?\[(.*?)\])()()()()/g,writeAnchorTag); // // Next, inline-style links: [link text](url "optional title") // /* text = text.replace(/ ( // wrap whole match in $1 \[ ( (?: \[[^\]]*\] // allow brackets nested one level | [^\[\]] // or anything else ) ) \] \( // literal paren [ \t]* () // no id, so leave $3 empty (?:[^\(]*?\([^\)]*?\).*?) ? [ \t]* ( // $5 (['"]) // quote char = $6 (.*?) // Title = $7 \6 // matching quote [ \t]* // ignore any spaces/tabs between closing quote and ) )? // title is optional \) ) /g,writeAnchorTag); */ text = text.replace(/(\[((?:\[[^\]]*\]|[^\[\]])*)\]\([ \t]*()?[ \t]*((['"])(.*?)\6[ \t]*)?\))/g,writeAnchorTag); // // Last, handle reference-style shortcuts: [link text] // These must come last in case you've also got [link test][1] // or [link test](/foo) // /* text = text.replace(/ ( // wrap whole match in $1 \[ ([^\[\]]+) // link text = $2; can't contain '[' or ']' \] )()()()()() // pad rest of backreferences /g, writeAnchorTag); */ text = text.replace(/(\[([^\[\]]+)\])()()()()()/g, writeAnchorTag); return text; } var writeAnchorTag = function(wholeMatch,m1,m2,m3,m4,m5,m6,m7) { if (m7 == undefined) m7 = ""; var whole_match = m1; var link_text = m2; var link_id = m3.toLowerCase(); var url = m4; var title = m7; if (url == "") { if (link_id == "") { // lower-case and turn embedded newlines into spaces link_id = link_text.toLowerCase().replace(/ ?\n/g," "); } url = "#"+link_id; if (g_urls[link_id] != undefined) { url = g_urls[link_id]; if (g_titles[link_id] != undefined) { title = g_titles[link_id]; } } else { if (whole_match.search(/\(\s*\)$/m)>-1) { // Special case for explicit empty url url = ""; } else { return whole_match; } } } url = escapeCharacters(url,"*_"); var result = ""; if (config.refprint) { if (!g_print_refs[url]) { g_print_refs[url] = ++g_print_refs_count; } result += '~E91E' + g_print_refs[url] + '~E93E'; } return result; } var _DoImages = function(text) { // // Turn Markdown image shortcuts into tags. // // // First, handle reference-style labeled images: ![alt text][id] // /* text = text.replace(/ ( // wrap whole match in $1 !\[ (.*?) // alt text = $2 \] [ ]? // one optional space (?:\n[ ]*)? // one optional newline followed by spaces \[ (.*?) // id = $3 \] )()()()() // pad rest of backreferences /g,writeImageTag); */ text = text.replace(/(![<>]?\[(.*?)\][ ]?(?:\n[ ]*)?\[(.*?)\])()()()()/g,writeImageTag); // // Next, handle inline images: ![alt text](url "optional title") // Don't forget: encode * and _ /* text = text.replace(/ ( // wrap whole match in $1 !\[ (.*?) // alt text = $2 \] \s? // One optional whitespace character \( // literal paren [ \t]* () // no id, so leave $3 empty ? [ \t]* ( // $5 (['"]) // quote char = $6 (.*?) // title = $7 \6 // matching quote [ \t]* )? // title is optional \) ) /g,writeImageTag); */ text = text.replace(/(![<>]?\[(.*?)\]\s?\([ \t]*()?[ \t]*((['"])(.*?)\6[ \t]*)?\))/g,writeImageTag); return text; } var writeImageTag = function(wholeMatch,m1,m2,m3,m4,m5,m6,m7) { var whole_match = m1; var alt_text = m2; var link_id = m3.toLowerCase(); var url = m4; var title = m7; if (!title) title = ""; if (url == "" && link_id !== "") { // lower-case and turn embedded newlines into spaces if (g_urls[link_id] != undefined) { url = g_urls[link_id]; if (g_titles[link_id] != undefined) { title = g_titles[link_id]; } else { title = undefined; } } } var figure = false, match; if (match = whole_match.match(/^!([<>])/)) { if (match[1] === ">") { figure = "right"; } else if (match[1] === "<") { figure = "left"; } } alt_text = alt_text.replace(/"/g,"""); url = escapeCharacters(url,"*_"); var result = "\""\n ' + result + '\n'; } else { result = '

'; } } return result; } var _DoHeaders = function(text) { // Setext-style headers: // Header 1 // ======== // // Header 2 // -------- // text = text.replace(/^(.+)[ \t]*\n=+[ \t]*\n+/gm, function(wholeMatch,m1) { if (config.headerautoid) { return hashBlock('

' + _RunSpanGamut(m1) + "

"); } else { return hashBlock('

' + _RunSpanGamut(m1) + "

"); } } ); text = text.replace(/^(.+)[ \t]*\n-+[ \t]*\n+/gm, function(matchFound,m1) { if (config.headerautoid) { return hashBlock('

' + _RunSpanGamut(m1) + "

"); } else { return hashBlock('

' + _RunSpanGamut(m1) + "

"); } } ); // atx-style headers: // # Header 1 // ## Header 2 // ## Header 2 with closing hashes ## // ... // ###### Header 6 // /* text = text.replace(/ ^(\#{1,6}) // $1 = string of #'s [ \t]* (.+?) // $2 = Header text [ \t]* \#* // optional closing #'s (not counted) \n+ /gm, function() {...}); */ text = text.replace(/^(\#{1,6})[ \t]*(.+?)[ \t]*\#*\n+/gm, function(wholeMatch,m1,m2) { var h_level = m1.length; if (config.headerautoid) { return hashBlock("' + _RunSpanGamut(m2) + ""); } else { return hashBlock("' + _RunSpanGamut(m2) + ""); } }); function headerId(m) { return m.replace(/[^\w]/g, '').toLowerCase(); } return text; } // This declaration keeps Dojo compressor from outputting garbage: var _ProcessListItems; var _DoLists = function(text) { // // Form HTML ordered (numbered) and unordered (bulleted) lists. // // attacklab: add sentinel to hack around khtml/safari bug: // http://bugs.webkit.org/show_bug.cgi?id=11231 text += "~0"; // Re-usable pattern to match any entirel ul or ol list: /* var whole_list = / ( // $1 = whole list ( // $2 [ ]{0,3} // attacklab: g_tab_width - 1 ([*+-]|\d+[.]) // $3 = first list item marker [ \t]+ ) [^\r]+? ( // $4 ~0 // sentinel for workaround; should be $ | \n{2,} (?=\S) (?! // Negative lookahead for another list item marker [ \t]* (?:[*+-]|\d+[.])[ \t]+ ) ) )/g */ var whole_list = /^(([ ]{0,3}([*+-]|\d+[.])[ \t]+)[^\r]+?(~0|\n{2,}(?=\S)(?![ \t]*(?:[*+-]|\d+[.])[ \t]+)))/gm; if (g_list_level) { text = text.replace(whole_list,function(wholeMatch,m1,m2) { var list = m1; var list_type = (m2.search(/[*+-]/g)>-1) ? "ul" : "ol"; // Turn double returns into triple returns, so that we can make a // paragraph for the last item in a list, if necessary: list = list.replace(/\n{2,}/g,"\n\n\n");; var result = _ProcessListItems(list); // Trim any trailing whitespace, to put the closing `` // up on the preceding line, to get it past the current stupid // HTML block parser. This is a hack to work around the terrible // hack that is the HTML block parser. result = result.replace(/\s+$/,""); result = "<"+list_type+">\n" + result + "\n"; return result; }); } else { whole_list = /(\n\n|^\n?)(([ ]{0,3}([*+-]|\d+[.])[ \t]+)[^\r]+?(~0|\n{2,}(?=\S)(?![ \t]*(?:[*+-]|\d+[.])[ \t]+)))/g; text = text.replace(whole_list,function(wholeMatch,m1,m2,m3) { var runup = m1; var list = m2; var list_type = (m3.search(/[*+-]/g)>-1) ? "ul" : "ol"; // Turn double returns into triple returns, so that we can make a // paragraph for the last item in a list, if necessary: var list = list.replace(/\n{2,}/g,"\n\n\n");; var result = _ProcessListItems(list); result = runup + "<"+list_type+">\n" + result + "\n"; return result; }); } // attacklab: strip sentinel text = text.replace(/~0/,""); return text; } _ProcessListItems = function(list_str) { // // Process the contents of a single ordered or unordered list, splitting it // into individual list items. // // The $g_list_level global keeps track of when we're inside a list. // Each time we enter a list, we increment it; when we leave a list, // we decrement. If it's zero, we're not in a list anymore. // // We do this because when we're not inside a list, we want to treat // something like this: // // I recommend upgrading to version // 8. Oops, now this line is treated // as a sub-list. // // As a single paragraph, despite the fact that the second line starts // with a digit-period-space sequence. // // Whereas when we're inside a list (or sub-list), that line will be // treated as the start of a sub-list. What a kludge, huh? This is // an aspect of Markdown's syntax that's hard to parse perfectly // without resorting to mind-reading. Perhaps the solution is to // change the syntax rules such that sub-lists must start with a // starting cardinal number; e.g. "1." or "a.". g_list_level++; // trim trailing blank lines: list_str = list_str.replace(/\n{2,}$/,"\n"); // attacklab: add sentinel to emulate \z list_str += "~0"; /* list_str = list_str.replace(/ (\n)? // leading line = $1 (^[ \t]*) // leading whitespace = $2 ([*+-]|\d+[.]) [ \t]+ // list marker = $3 ([^\r]+? // list item text = $4 (\n{1,2})) (?= \n* (~0 | \2 ([*+-]|\d+[.]) [ \t]+)) /gm, function(){...}); */ list_str = list_str.replace(/(\n)?(^[ \t]*)([*+-]|\d+[.])[ \t]+([^\r]+?(\n{1,2}))(?=\n*(~0|\2([*+-]|\d+[.])[ \t]+))/gm, function(wholeMatch,m1,m2,m3,m4){ var item = m4; var leading_line = m1; var leading_space = m2; if (leading_line || (item.search(/\n{2,}/)>-1)) { item = _RunBlockGamut(_Outdent(item)); } else { // Recursion for sub-lists: item = _DoLists(_Outdent(item)); item = item.replace(/\n$/,""); // chomp(item) item = _RunSpanGamut(item); } return "
  • " + item + "
  • \n"; } ); // attacklab: strip sentinel list_str = list_str.replace(/<\/li>\n<"); list_str = list_str.replace(/~0/g,""); g_list_level--; return list_str; } var _DoCodeBlocks = function(text) { // // Process Markdown `
    ` blocks.
    //  
    
      /*
        text = text.replace(text,
          /(?:\n\n|^)
          (                // $1 = the code block -- one or more lines, starting with a space/tab
            (?:
              (?:[ ]{4}|\t)      // Lines must start with a tab or a tab-width of spaces - attacklab: g_tab_width
              .*\n+
            )+
          )
          (\n*[ ]{0,3}[^ \t\n]|(?=~0))  // attacklab: g_tab_width
        /g,function(){...});
      */
    
      // attacklab: sentinel workarounds for lack of \A and \Z, safari\khtml bug
      text += "~0";
      
      if (config.github_flavouring) {  
        text = text.replace(/\n```([a-zA-Z]+)?\s*\n((?:.*\n+)+?)(\n*```|(?=~0))/g,
          function (wholeMatch, m1, m2) {
            var codeblock = _EncodeCode(m2);
            codeblock = _Detab(codeblock);
            codeblock = codeblock.replace(/^\n+/g,""); // trim leading newlines
            codeblock = codeblock.replace(/\n+$/g,""); // trim trailing whitespace
          
            if (m1) {
              codeblock = "
    " + codeblock + "\n
    "; } else { codeblock = "
    " + codeblock + "\n
    "; } return hashBlock(codeblock); }); } text = text.replace(/(?:\n\n|^)((?:(?:[ ]{4}|\t).*\n+)+)(\n*[ ]{0,3}[^ \t\n]|(?=~0))/g, function(wholeMatch,m1,m2) { var codeblock = m1; var nextChar = m2; codeblock = _EncodeCode( _Outdent(codeblock)); codeblock = _Detab(codeblock); codeblock = codeblock.replace(/^\n+/g,""); // trim leading newlines codeblock = codeblock.replace(/\n+$/g,""); // trim trailing whitespace codeblock = "
    " + codeblock + "\n
    "; return hashBlock(codeblock) + nextChar; } ); // attacklab: strip sentinel text = text.replace(/~0/,""); return text; } var hashBlock = function(text) { text = text.replace(/(^\n+|\n+$)/g,""); return "\n\n~K" + (g_html_blocks.push(text)-1) + "K\n\n"; } var _DoCodeSpans = function(text) { // // * Backtick quotes are used for spans. // // * You can use multiple backticks as the delimiters if you want to // include literal backticks in the code span. So, this input: // // Just type ``foo `bar` baz`` at the prompt. // // Will translate to: // //

    Just type foo `bar` baz at the prompt.

    // // There's no arbitrary limit to the number of backticks you // can use as delimters. If you need three consecutive backticks // in your code, use four for delimiters, etc. // // * You can use spaces to get literal backticks at the edges: // // ... type `` `bar` `` ... // // Turns to: // // ... type `bar` ... // /* text = text.replace(/ (^|[^\\]) // Character before opening ` can't be a backslash (`+) // $2 = Opening run of ` ( // $3 = The code block [^\r]*? [^`] // attacklab: work around lack of lookbehind ) \2 // Matching closer (?!`) /gm, function(){...}); */ text = text.replace(/(^|[^\\])(`+)([^\r]*?[^`])\2(?!`)/gm, function(wholeMatch,m1,m2,m3,m4) { var c = m3; c = c.replace(/^([ \t]*)/g,""); // leading whitespace c = c.replace(/[ \t]*$/g,""); // trailing whitespace c = _EncodeCode(c); return m1+""+c+""; }); if (config.math) { text = text.replace(/(^|[^\\])(%%)([^\r]*?[^%])\2(?!%)/gm, function(wholeMatch,m1,m2,m3,m4) { var c = m3; c = c.replace(/^([ \t]*)/g,""); // leading whitespace c = c.replace(/[ \t]*$/g,""); // trailing whitespace c = _EncodeCode(c); return m1+'%%'+c+"%%"; }); text = text.replace(/(^|[^\\])(~D~D)([^\r]*?[^~])\2(?!~D)/gm, function(wholeMatch,m1,m2,m3,m4) { var c = m3; c = c.replace(/^([ \t]*)/g,""); // leading whitespace c = c.replace(/[ \t]*$/g,""); // trailing whitespace c = _EncodeCode(c); return m1+'~D~D'+c+"~D~D"; }); } return text; } var _EncodeCode = function(text) { // // Encode/escape certain characters inside Markdown code runs. // The point is that in code, these characters are literals, // and lose their special Markdown meanings. // // Encode all ampersands; HTML entities are not // entities within a Markdown code span. text = text.replace(/&/g,"&"); // Do the angle bracket song and dance: text = text.replace(//g,">"); // Pipes are escaped early, unescape them into escaped pipes. // Need to find better solution. text = text.replace(/~E124E/g, "\\|"); // Now, escape characters that are magic in Markdown: text = escapeCharacters(text,"\*_{}[]\\",false); // jj the line above breaks this: //--- //* Item // 1. Subitem // special char: * //--- return text; } var _DoTextStyle = function(text) { // must go first: text = text.replace(/(\*\*|__)(?=\S)([^\r]*?\S[*_]*)\1/g, "$2"); if (config.github_flavouring) { text = text.replace(/(\w)_(\w)/g, "$1~E95E$2") // ** GFM ** "~E95E" == escaped "_" } text = text.replace(/(~T~T)(?=\S)([^\r]*?\S[*_]*)\1/g, "$2"); text = text.replace(/(\*|_)(?=\S)([^\r]*?\S)\1/g, "$2"); return text; } var _DoBlockQuotes = function(text) { /* text = text.replace(/ ( // Wrap whole match in $1 ( ^[ \t]*>[ \t]? // '>' at the start of a line .+\n // rest of the first line (.+\n)* // subsequent consecutive lines \n* // blanks )+ ) /gm, function(){...}); */ text = text.replace(/((^[ \t]*>[ \t]?.+\n(.+\n)*\n*)+)/gm, function(wholeMatch,m1) { var bq = m1; // attacklab: hack around Konqueror 3.5.4 bug: // "----------bug".replace(/^-/g,"") == "bug" bq = bq.replace(/^[ \t]*>[ \t]?/gm,"~0"); // trim one level of quoting // attacklab: clean up hack bq = bq.replace(/~0/g,""); bq = bq.replace(/^[ \t]+$/gm,""); // trim whitespace-only lines bq = _RunBlockGamut(bq); // recurse bq = bq.replace(/(^|\n)/g,"$1 "); // These leading spaces screw with
     content, so we need to fix that:
          bq = bq.replace(
              /(\s*
    [^\r]+?<\/pre>)/gm,
            function(wholeMatch,m1) {
              var pre = m1;
              // attacklab: hack around Konqueror 3.5.4 bug:
              pre = pre.replace(/^  /mg,"~0");
              pre = pre.replace(/~0/g,"");
              return pre;
            });
          
          return hashBlock("
    \n" + bq + "\n
    "); }); return text; } var _FormParagraphs = function(text) { // // Params: // $text - string to process with html

    tags // // Strip leading and trailing lines: text = text.replace(/^\n+/g,""); text = text.replace(/\n+$/g,""); var grafs = text.split(/\n{2,}/g); var grafsOut = new Array(); // // Wrap

    tags. // var end = grafs.length; for (var i=0; i= 0) { grafsOut.push(str); } else if (str.search(/\S/) >= 0) { str = _RunSpanGamut(str); if (config.github_flavouring) { str = str.replace(/\n/g,"
    "); // ** GFM ** } str = str.replace(/^([ \t]*)/g,"

    "); str += "

    " grafsOut.push(str); } } // // Unhashify HTML blocks // end = grafsOut.length; for (var i=0; i= 0) { var blockText = g_html_blocks[RegExp.$1]; blockText = blockText.replace(/\$/g,"$$$$"); // Escape any dollar signs grafsOut[i] = grafsOut[i].replace(/~K\d+K/,blockText); } } return grafsOut.join("\n\n"); } var _EncodeAmpsAndAngles = function(text) { // Smart processing for ampersands and angle brackets that need to be encoded. // Ampersand-encoding based entirely on Nat Irons's Amputator MT plugin: // http://bumppo.net/projects/amputator/ text = text.replace(/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/g,"&"); // Encode naked <'s text = text.replace(/<(?![a-z\/?\$!])/gi,"<"); return text; } var _EncodeBackslashEscapes = function(text) { // // Parameter: String. // Returns: The string, with after processing the following backslash // escape sequences. // // attacklab: The polite way to do this is with the new // escapeCharacters() function: // // text = escapeCharacters(text,"\\",true); // text = escapeCharacters(text,"`*_{}[]()>#+-.!",true); // // ...but we're sidestepping its use of the (slow) RegExp constructor // as an optimization for Firefox. This function gets called a LOT. text = text.replace(/\\(\\)/g,escapeCharacters_callback); text = text.replace(/\\([`*_{}\[\]()>#+-.!])/g,escapeCharacters_callback); return text; } var _DoAutoLinks = function(text) { text = text.replace(/<((https?|ftp|dict):[^'">\s]+)>/gi,"
    $1"); // Email addresses: /* text = text.replace(/ < (?:mailto:)? ( [-.\w]+ \@ [-a-z0-9]+(\.[-a-z0-9]+)*\.[a-z]+ ) > /gi, _DoAutoLinks_callback()); */ text = text.replace(/<(?:mailto:)?([-.\w]+\@[-a-z0-9]+(\.[-a-z0-9]+)*\.[a-z]+)>/gi, function(wholeMatch,m1) { return _EncodeEmailAddress( _UnescapeSpecialChars(m1) ); } ); return text; } var _EncodeEmailAddress = function(addr) { // // Input: an email address, e.g. "foo@example.com" // // Output: the email address as a mailto link, with each character // of the address encoded as either a decimal or hex entity, in // the hopes of foiling most address harvesting spam bots. E.g.: // // foo // @example.com // // Based on a filter by Matthew Wickline, posted to the BBEdit-Talk // mailing list: // // attacklab: why can't javascript speak hex? function char2hex(ch) { var hexDigits = '0123456789ABCDEF'; var dec = ch.charCodeAt(0); return(hexDigits.charAt(dec>>4) + hexDigits.charAt(dec&15)); } var encode = [ function(ch){return "&#"+ch.charCodeAt(0)+";";}, function(ch){return "&#x"+char2hex(ch)+";";}, function(ch){return ch;} ]; addr = "mailto:" + addr; addr = addr.replace(/./g, function(ch) { if (ch == "@") { // this *must* be encoded. I insist. ch = encode[Math.floor(Math.random()*2)](ch); } else if (ch !=":") { // leave ':' alone (to spot mailto: later) var r = Math.random(); // roughly 10% raw, 45% hex, 45% dec ch = ( r > .9 ? encode[2](ch) : r > .45 ? encode[1](ch) : encode[0](ch) ); } return ch; }); addr = "" + addr + ""; addr = addr.replace(/">.+:/g,"\">"); // strip the mailto: from the visible part return addr; } var _UnescapeSpecialChars = function(text) { // // Swap back in all the special characters we've hidden. // text = text.replace(/~E(\d+)E/g, function(wholeMatch,m1) { var charCodeToReplace = parseInt(m1); return String.fromCharCode(charCodeToReplace); } ); return text; } var _Outdent = function(text) { // // Remove one level of line-leading tabs or spaces // // attacklab: hack around Konqueror 3.5.4 bug: // "----------bug".replace(/^-/g,"") == "bug" text = text.replace(/^(\t|[ ]{1,4})/gm,"~0"); // attacklab: g_tab_width // attacklab: clean up hack text = text.replace(/~0/g,"") return text; } var _Detab = function(text) { // attacklab: Detab's completely rewritten for speed. // In perl we could fix it by anchoring the regexp with \G. // In javascript we're less fortunate. // expand first n-1 tabs text = text.replace(/\t(?=\t)/g," "); // attacklab: g_tab_width // replace the nth with two sentinels text = text.replace(/\t/g,"~A~B"); // use the sentinel to anchor our regex so it doesn't explode text = text.replace(/~B(.+?)~A/g, function(wholeMatch,m1,m2) { var leadingText = m1; var numSpaces = 4 - leadingText.length % 4; // attacklab: g_tab_width // there *must* be a better way to do this: for (var i=0; i [ ]* # Allowed whitespace. [|] .* \n # Row content. )* ) (?=\n|\Z) # Stop at final double newline. }xm', array(&$this, '_doTable_leadingPipe_callback'), $text);*/ text = text.replace(/^[ ]{0,3}[|](.+)\n[ ]{0,3}[|]([ ]*[-:]+[-| :]*)\n(((?=([ ]*[|].*\n))(?:[ ]*[|].*\n))*)(?=\n)/gm, _doTable_leadingPipe_callback); // // Find tables without leading pipe. // // Header 1 | Header 2 // -------- | -------- // Cell 1 | Cell 2 // Cell 3 | Cell 4 // /* $text = preg_replace_callback(' { ^ # Start of a line [ ]{0,'.$less_than_tab.'} # Allowed whitespace. (\S.*[|].*) \n # $1: Header row (at least one pipe) [ ]{0,'.$less_than_tab.'} # Allowed whitespace. ([-:]+[ ]*[|][-| :]*) \n # $2: Header underline ( # $3: Cells (?> .* [|] .* \n # Row content )* ) (?=\n|\Z) # Stop at final double newline. }xm', array(&$this, '_DoTable_callback'), $text);*/ text = text.replace(/^[ ]{0,3}(\S.*[|].*)\n[ ]{0,3}([ ]*[-:]+[-| :]*)\n(((?=(.*[|].*\n))(?:.*[|].*\n))*)(?=\n)/gm, _doTable_callback); return text; } var _doTable_leadingPipe_callback = function () { var content = arguments[3]; // Remove leading pipe for each row. content = content.replace(/^[ ]*[|]/gm, ''); return _doTable_callback(arguments[0], arguments[1], arguments[2], content); } var _doTable_callback = function () { var head = arguments[1]; var underline = arguments[2]; var content = arguments[3]; // Remove any tailing pipes for each line. head = head.replace(/[|][ ]*$/gm, '\n'); underline = underline.replace(/[|][ ]*\n/gm, '\n'); content = doTrim(content.replace(/[|][ ]*\n/gm, '\n')); // Reading alignement from header underline. var separators = underline.split(/[ ]*[|][ ]*/); var attr = {}; for (var i = 0, len = separators.length; i < len; ++i) { if (separators[i].match(/^[ ]*[-]+:[ ]*$/)) attr[i] = ' align="right"'; else if (separators[i].match(/^[ ]*:[-]+:[ ]*$/)) attr[i] = ' align="center"'; else if (separators[i].match(/^[ ]*:[-]+[ ]*$/)) attr[i] = ' align="left"'; else attr[i] = ''; } // Parsing span elements, including code spans, character escapes, // and inline HTML tags, so that pipes inside those gets ignored. // head = $this->parseSpan($head); // TODO var headers = head.split(/[ ]*[|][ ]*/); var col_count = headers.length; // Write column headers. var text = "\n"; text += "\n"; text += "\n"; for (var i = 0; i < col_count; ++i) { text += " " + _RunSpanGamut(doTrim(headers[i])) + "\n"; } text += "\n"; text += "\n"; // Split content by row. var rows = content.split(/\n/); text += "\n"; for (var i = 0, len = rows.length; i < len; ++i) { // Parsing span elements, including code spans, character escapes, // and inline HTML tags, so that pipes inside those gets ignored. // $row = $this->parseSpan($row); // TODO // Split row by cell. var row_cells = rows[i].split(/[ ]*[|][ ]*/); // row_cells = array_pad($row_cells, $col_count, ''); // TODO text += "\n"; for (var j = 0, len2 = row_cells.length; j < len2; ++j) { text += " " + _RunSpanGamut(doTrim(row_cells[j])) + "\n"; } text += "\n"; } text += "\n"; text += "
    "; return text; } } // end of Showdown.converter // Sourced from https://github.com/andris9/node-markdown var stripUnwantedHTML = function (html /*, allowedTags, allowedAttributes, forceProtocol */){ var allowedTags = arguments[1] || 'a|b|blockquote|code|del|dd|dl|dt|em|h1|h2|h3|h4|h5|h6|'+ 'i|img|li|ol|p|pre|sup|sub|strong|strike|ul|br|hr|span|'+ 'table|th|tr|td|tbody|thead|tfoot|div', allowedAttributes = arguments[2] || { 'img': 'src|width|height|alt', 'a': 'href', '*': 'title', 'span': 'class', 'tr': 'rowspan', 'td': 'colspan|align', 'th': 'rowspan|align', 'div': 'class', 'code': 'class' }, forceProtocol = arguments[3] || true; testAllowed = new RegExp('^('+allowedTags.toLowerCase()+')$'), findTags = /<(\/?)\s*([\w:\-]+)([^>]*)>/g, findAttribs = /(\s*)([\w:-]+)\s*=\s*(?:(?:(["'])([^\3]+?)(?:\3))|([^\s]+))/g; // convert all strings patterns into regexp objects (if not already converted) for(var i in allowedAttributes){ if(allowedAttributes.hasOwnProperty(i) && typeof allowedAttributes[i] === 'string'){ allowedAttributes[i] = new RegExp('^('+ allowedAttributes[i].toLowerCase()+')$'); } } // find and match html tags return html.replace(findTags, function(original, lslash, tag, params){ var tagAttr, wildcardAttr, rslash = params.substr(-1)=="/" && "/" || ""; tag = tag.toLowerCase(); // tag is not allowed, return empty string if(!tag.match(testAllowed)) return ""; // tag is allowed else{ // regexp objects for a particular tag tagAttr = tag in allowedAttributes && allowedAttributes[tag]; wildcardAttr = "*" in allowedAttributes && allowedAttributes["*"]; // if no attribs are allowed if(!tagAttr && !wildcardAttr) return "<"+lslash+tag+rslash+">"; // remove trailing slash if any params = params.trim(); if(rslash){ params = params.substr(0, params.length-1); } // find and remove unwanted attributes params = params.replace(findAttribs, function(original, space, name, quot, value){ name = name.toLowerCase(); if (!value && !quot) { value = ""; quot = '"'; } else if (!value) { value = quot; quot = '"'; } // force data: and javascript: links and images to # if((name=="href" || name=="src") && (value.trim().substr(0, "javascript:".length)=="javascript:" || (name == "href" && value.trim().substr(0, "data:".length)=="data:"))) { value = "#"; } // scope links and sources to http protocol if (forceProtocol && (name=="href" || name=="src") && !/^[a-zA-Z]{3,5}:\/\//.test(value) && (value.length < 8 && value.trim().substr(0, "mailto:".length)=="mailto:")) { value = "http://" + value; } if((wildcardAttr && name.match(wildcardAttr)) || (tagAttr && name.match(tagAttr))){ return space+name+"="+quot+value+quot; }else return ""; }); return "<"+lslash+tag+(params?" "+params:"")+rslash+">"; } }); } // export if (typeof exports != 'undefined') exports.Showdown = Showdown;