showdown.js.svn-base

mooltools源码很不错的源码

					\[[^\]]*\]		// 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
				)
			)
			\]
			\(						// literal paren
			[ \t]*
			()						// no id, so leave $3 empty
			<?(.*?)>?				// href = $4
			[ \t]*
			(						// $5
				(['"])				// quote char = $6
				(.*?)				// Title = $7
				\6					// matching quote
				[ \t]*				// ignore any spaces/tabs between closing quote and )
			)?						// title is optional
			\)
		)
		/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 = "<a href=\"" + url + "\"";

	if (title != "") {
		title = title.replace(/"/g,"&quot;");
		title = escapeCharacters(title,"*_");
		result +=  " title=\"" + title + "\"";
	}

	result += ">" + link_text + "</a>";

	return result;
}


var _DoImages = function(text) {
//
// Turn Markdown image shortcuts into <img> 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
			\(					// literal paren
			[ \t]*
			()					// no id, so leave $3 empty
			<?(\S+?)>?			// src url = $4
			[ \t]*
			(					// $5
				(['"])			// quote char = $6
				(.*?)			// title = $7
				\6				// matching quote
				[ \t]*
			)?					// title is optional
		\)
		)
		/g,writeImageTag);
	*/
	text = text.replace(/(!\[(.*?)\]\s?\([ \t]*()<?(\S+?)>?[ \t]*((['"])(.*?)\6[ \t]*)?\))/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 == "") {
		if (link_id == "") {
			// lower-case and turn embedded newlines into spaces
			link_id = alt_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 {
			return whole_match;
		}
	}	

	alt_text = alt_text.replace(/"/g,"&quot;");
	url = escapeCharacters(url,"*_");
	var result = "<img src=\"" + url + "\" alt=\"" + alt_text + "\"";

	// attacklab: Markdown.pl adds empty title attributes to images.
	// Replicate this bug.

	//if (title != "") {
		title = title.replace(/"/g,"&quot;");
		title = escapeCharacters(title,"*_");
		result +=  " title=\"" + title + "\"";
	//}

	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){return hashBlock("<h1>" + _RunSpanGamut(m1) + "</h1>");});

	text = text.replace(/^(.+)[ \t]*\n-+[ \t]*\n+/gm,
		function(matchFound,m1){return hashBlock("<h2>" + _RunSpanGamut(m1) + "</h2>");});

	// 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;
			return hashBlock("<h" + h_level + ">" + _RunSpanGamut(m2) + "</h" + h_level + ">");
		});

	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 `</$list_type>`
			// up on the preceding line, to get it past the current stupid
			// HTML block parser. This is a hack to work around the terrible
			// hack that is the HTML block parser.
			result = result.replace(/\s+$/,"");
			result = "<"+list_type+">" + result + "</"+list_type+">\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 + "</"+list_type+">\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  "<li>" + item + "</li>\n";
		}
	);

	// attacklab: strip sentinel
	list_str = list_str.replace(/~0/g,"");

	g_list_level--;
	return list_str;
}


var _DoCodeBlocks = function(text) {
//
//  Process Markdown `<pre><code>` 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