/*
HTMl, Javascript, Css format for Emeditor
---------------
  $Date: 2008-08-02 23:32:18
  Written by KevinYuan whyork@21cn.com
  You are free to use this in any way you want
feature:
1> Format HTML, Javascript, CSS ,XML
2> Format HTML with inline CSS, Javascript

!Caution:
1> Javascript is the only script tag which supported
2> If you want format with perfect result, make sure your source HTMl code with XHTML protocal compatible.
*/

var indentCount = 0;
var StyleBlock = [];
var ScriptBlock = [];
var CommentBlock = [];
var style_tag = /<style[\w\W]+?<\/style>/img;
var script_tag = /<script[\w\W]+?<\/script>/mig;
var comment_tag = /(<\!--[\w\W]*?-->)|(<textarea[\w\W]+?<\/textarea>)|(<pre[\w\W]+?<\/pre>)/img;
function FormatHTML(htmlStr) {

	//匹配类似<meta  br/>的标签
	//匹配类似<div p开头
	//匹配类似/div>结尾
	var reg = /<[^\>]*\/>|<[^\>]+>|<\/[\w]+>/mig;
	var str = htmlStr.replace(reg, fw);

	//将单tag的换行及空格去掉  2008-09-01 16:51:01
	str = str.replace(/<[a-z]+[^>]*>/img, finalHtml);
	str = str.replace(/(<[a-z]+[^>]*>)[\t\r\n]*([^<]+?)[\t\r\n]*(<\/[^>]*>)/img, finalHtml);

	//将空白换行去掉  2008-09-01 16:50:49
	str = str.replace(/\n[\s \t\r\n]*\n/igm, '\n');
	return str;
}
function finalHtml(mc) {
	//将单tag的换行及空格去掉  2008-09-01 21:52:32
	mc = mc.replace(/[\t\r\n]+/img, ' ');
	//将双空格替换成单空格  2008-09-01 16:50:31
	mc = mc.replace(/[ ]+/g, ' ');
	//将空tag中的空格去掉(由于tag替换造成的多余空格)  2008-09-01 21:53:11
	mc = mc.replace(/> </g, '><');
	return mc;
}
function fw(mc) {
	//发现有 close_tag  2008-09-01 21:54:15
	if (mc.indexOf('</') > -1) {
		indentCount--;
		var str = getIndent(indentCount) + mc;
		return str;
		//发现 有单体tag  2008-09-01 21:54:36
	} else if (mc.indexOf('\/>') > 0 || mc.indexOf('<?') == 0 || mc.indexOf('<!') == 0) {
		return getIndent(indentCount) + mc;
	} else {
		//发现 open_tag 开始计算缩进  2008-09-01 21:54:49
		var str = getIndent(indentCount) + mc;
		indentCount++;
		return str;
	}
}

function formatXML(xmlstr) {
	var reg = /<[^\>]*\/>|<[\w\!\?\-]+|[^<]*<\/[\w]+>/mig;
	var indent = 0;
	function inlineFun(mc) {
		if (mc.indexOf('</') > -1) {
			indent--;
			var str = getIndent(indent) + mc;
			return str;
		} else if (mc.indexOf('\/>') > 0 || mc.indexOf('<!') >= 0 || mc.indexOf('<?') >= 0) {
			return getIndent(indent) + mc;
		} else {
			var str = getIndent(indent) + mc;
			indent++;
			return str;
		}
	}
	var str = xmlstr.replace(reg, inlineFun);
	str = str.replace(/^[\s]*\n/gm, '') //.replace(/<[^>]+>[^<]+<\/[^>]+>/igm,removeBlank);
	return str;
}

function removeBlank(mc) {
	return mc.replace(/[\t\n\r]*/igm, '');
}
function getIndent(inCount) {
	var arr = [];
	arr.push('\r\n');
	for (var n = 0; n < inCount; n++) {
		arr.push('\t');
	}
	return arr.join('');
}

function ProcessStyleBlock(mc) {
	StyleBlock.push(mc);
	return '##Style' + (StyleBlock.length - 1);
}
function ProcessScriptBlock(mc) {
	if (mc.match(/src=/i)) {
		return mc
	}
	mc = mc.replace(/<script[^>]*>([\w\W]+?)<\/script>/,'$1');
	ScriptBlock.push(mc);
	return '##Script' + (ScriptBlock.length - 1);
	
}
function ProcessCommentBlock(mc) {
	CommentBlock.push(mc);
	return '##Comment' + (CommentBlock.length - 1);
} (function() {
	var word = document.selection.Text;
	var currentLine = document.selection.GetActivePointY(eePosView);
	//如果当前选择的文字为空,则选择整个文档的内容
	if (word == '') {
		document.selection.selectAll();
		word = document.selection.Text;
	}
	word = word.replace(style_tag, ProcessStyleBlock);
	word = word.replace(script_tag, ProcessScriptBlock);
	word = word.replace(comment_tag, ProcessCommentBlock);
	var result;
	//获取匹配当前文档的配置名如 html,htm,xml,css,js,c#	
	var type = document.ConfigName.toLowerCase().replace(/ /g, '');
	if (type.indexOf('htm') > -1) {
		result = FormatHTML(word);
	} else if (type.indexOf('script') > -1) {
		result = formatJS(word);
	} else if (type.indexOf('htm') > -1) {
		result = formatXML(word);
	} else if (type.indexOf('css') > -1) {
		result = formatCss(word);
	}
	for (var i = 0; i < StyleBlock.length; i++) {
		result = result.replace('##Style' + i, formatCss(StyleBlock[i]));
	}
	for (var i = 0; i < ScriptBlock.length; i++) {
		result = result.replace('##Script' + i, '<script type="text/javascript">\r\n'+formatJS(ScriptBlock[i])+'\r\n</script>');
	}
	for (var i = 0; i < CommentBlock.length; i++) {
		result = result.replace('##Comment' + i, CommentBlock[i]);
	}

	document.selection.Text = result;
	document.selection.SetActivePoint(eePosView, 1, currentLine);

})();

function formatCss(code) {
	code = code.replace(/(\s){2,}/ig, '$1');
	code = code.replace(/(\S)\s*\{/ig, '$1 {');
	code = code.replace(/\*\/(.[^\}\{]*)}/ig, '\*\/\n$1}');
	code = code.replace(/\/\*/ig, '\n\/\*');
	code = code.replace(/;\s*(\S)/ig, ';\n\t$1');
	code = code.replace(/\}\s*(\S)/ig, '\}\n$1');
	code = code.replace(/\n\s*\}/ig, '\n\}');
	code = code.replace(/\{\s*(\S)/ig, '\{\n\t$1');
	code = code.replace(/(\S)\s*\*\//ig, '$1\*\/');
	code = code.replace(/\*\/\s*([^\}\{]\S)/ig, '\*\/\n\t$1');
	code = code.replace(/(\S)\}/ig, '$1\n\}');
	code = code.replace(/(\n){2,}/ig, '\n');
	code = code.replace(/:/ig, ': ');
	code = code.replace(/  /ig, ' ');
	return code;
}

function formatJS(code) {
	return js_beautify(code, 1, '\t');
}
/*
 JS Beautifier
---------------
  $Date: 2008-06-10 14:49:11 +0300 (Tue, 10 Jun 2008) $
  $Revision: 60 $
  Written by Einars "elfz" Lielmanis, <elfz@laacz.lv> 
      http://elfz.laacz.lv/beautify/
  Originally converted to javascript by Vital, <vital76@gmail.com> 
      http://my.opera.com/Vital/blog/2007/11/21/javascript-beautify-on-javascript-translated
  You are free to use this in any way you want, in case you find this useful or working for you.
  Usage:
    js_beautify(js_source_text);
*/
function js_beautify(js_source_text, indent_size, indent_character, indent_level) {
	var input, output, token_text, last_type, last_text, last_word, current_mode, modes, indent_string;
	var whitespace, wordchar, punct, parser_pos, line_starters, in_case;
	var prefix, token_type, do_block_just_closed, var_line, var_line_tainted;
	function trim_output() {
		while (output.length && (output[output.length - 1] === ' ' || output[output.length - 1] === indent_string)) {
			output.pop();
		}
	}
	function print_newline(ignore_repeated) {
		ignore_repeated = typeof ignore_repeated === 'undefined' ? true: ignore_repeated;

		trim_output();
		if (!output.length) {
			return; // no newline on start of file
		}
		if (output[output.length - 1] !== "\n" || !ignore_repeated) {
			output.push("\n");
		}
		for (var i = 0; i < indent_level; i++) {
			output.push(indent_string);
		}
	}
	function print_space() {
		var last_output = output.length ? output[output.length - 1] : ' ';
		if (last_output !== ' ' && last_output !== '\n' && last_output !== indent_string) { // prevent occassional duplicate space
			output.push(' ');
		}
	}
	function print_token() {
		output.push(token_text);
	}
	function indent() {
		indent_level++;
	}
	function unindent() {
		if (indent_level) {
			indent_level--;
		}
	}
	function remove_indent() {
		if (output.length && output[output.length - 1] === indent_string) {
			output.pop();
		}
	}
	function set_mode(mode) {
		modes.push(current_mode);
		current_mode = mode;
	}
	function restore_mode() {
		do_block_just_closed = current_mode === 'DO_BLOCK';
		current_mode = modes.pop();
	}
	function in_array(what, arr) {
		for (var i = 0; i < arr.length; i++) {
			if (arr[i] === what) {
				return true;
			}
		}
		return false;
	}
	function get_next_token() {
		var n_newlines = 0;
		var c = '';
		do {
			if (parser_pos >= input.length) {
				return ['', 'TK_EOF'];
			}
			c = input.charAt(parser_pos);
			parser_pos += 1;
			if (c === "\n") {
				n_newlines += 1;
			}
		} while ( in_array ( c , whitespace ));
		if (n_newlines > 1) {
			for (var i = 0; i < 2; i++) {
				print_newline(i === 0);
			}
		}
		var wanted_newline = (n_newlines === 1);
		if (in_array(c, wordchar)) {
			if (parser_pos < input.length) {
				while (in_array(input.charAt(parser_pos), wordchar)) {
					c += input.charAt(parser_pos);
					parser_pos += 1;
					if (parser_pos === input.length) {
						break;
					}
				}
			}
			// small and surprisingly unugly hack for 1E-10 representation
			if (parser_pos !== input.length && c.match(/^[0-9]+[Ee]$/) && input.charAt(parser_pos) === '-') {
				parser_pos += 1;
				var t = get_next_token(parser_pos);
				c += '-' + t[0];
				return [c, 'TK_WORD'];
			}
			if (c === 'in') { // hack for 'in' operator
				return [c, 'TK_OPERATOR'];
			}
			return [c, 'TK_WORD'];
		}

		if (c === '(' || c === '[') {
			return [c, 'TK_START_EXPR'];
		}
		if (c === ')' || c === ']') {
			return [c, 'TK_END_EXPR'];
		}
		if (c === '{') {
			return [c, 'TK_START_BLOCK'];
		}
		if (c === '}') {
			return [c, 'TK_END_BLOCK'];
		}
		if (c === ';') {
			return [c, 'TK_END_COMMAND'];
		}
		if (c === '/') {
			var comment = '';
			// peek for comment /* ... */
			if (input.charAt(parser_pos) === '*') {
				parser_pos += 1;
				if (parser_pos < input.length) {
					while (! (input.charAt(parser_pos) === '*' && input.charAt(parser_pos + 1) && input.charAt(parser_pos + 1) === '/') && parser_pos < input.length) {
						comment += input.charAt(parser_pos);
						parser_pos += 1;
						if (parser_pos >= input.length) {
							break;
						}
					}
				}
				parser_pos += 2;
				return ['/*' + comment + '*/', 'TK_BLOCK_COMMENT'];
			}
			// peek for comment // ...
			if (input.charAt(parser_pos) === '/') {
				comment = c;
				while (input.charAt(parser_pos) !== "\x0d" && input.charAt(parser_pos) !== "\x0a") {
					comment += input.charAt(parser_pos);
					parser_pos += 1;
					if (parser_pos >= input.length) {
						break;
					}
				}
				parser_pos += 1;
				if (wanted_newline) {
					print_newline();
				}
				return [comment, 'TK_COMMENT'];
			}
		}
		if (c === "'" || // string
		c === '"' || // string
		(c === '/' && ((last_type === 'TK_WORD' && last_text === 'return') || (last_type === 'TK_START_EXPR' || last_type === 'TK_END_BLOCK' || last_type === 'TK_OPERATOR' || last_type === 'TK_EOF' || last_type === 'TK_END_COMMAND')))) { // regexp
			var sep = c;
			var esc = false;
			c = '';
			if (parser_pos < input.length) {
				while (esc || input.charAt(parser_pos) !== sep) {
					c += input.charAt(parser_pos);
					if (!esc) {
						esc = input.charAt(parser_pos) === '\\';
					} else {
						esc = false;
					}
					parser_pos += 1;
					if (parser_pos >= input.length) {
						break;
					}
				}
			}
			parser_pos += 1;
			if (last_type === 'TK_END_COMMAND') {
				print_newline();
			}
			return [sep + c + sep, 'TK_STRING'];
		}
		if (in_array(c, punct)) {
			while (parser_pos < input.length && in_array(c + input.charAt(parser_pos), punct)) {
				c += input.charAt(parser_pos);
				parser_pos += 1;
				if (parser_pos >= input.length) {
					break;
				}
			}
			return [c, 'TK_OPERATOR'];
		}
		return [c, 'TK_UNKNOWN'];
	}
	//----------------------------------
	indent_character = indent_character || ' ';
	indent_size = indent_size || 4;
	indent_string = '';
	while (indent_size--) {
		indent_string += indent_character;
	}
	input = js_source_text;
	last_word = ''; // last 'TK_WORD' passed
	last_type = 'TK_START_EXPR'; // last token type
	last_text = ''; // last token text
	output = [];
	do_block_just_closed = false;
	var_line = false;
	var_line_tainted = false;
	whitespace = "\n\r\t ".split('');
	wordchar = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_$'.split('');
	punct = '+ - * / % & ++ -- = += -= *= /= %= == === != !== > < >= <= >> << >>> >>>= >>= <<= && &= | || ! !! , : ? ^ ^= |='.split(' ');
	// words which should always start on new line.
	line_starters = 'continue,try,throw,return,var,if,switch,case,default,for,while,break,function'.split(',');
	// states showing if we are currently in expression (i.e. "if" case) - 'EXPRESSION', or in usual block (like, procedure), 'BLOCK'.
	// some formatting depends on that.
	current_mode = 'BLOCK';
	modes = [current_mode];
	indent_level = indent_level || 0;
	parser_pos = 0; // parser position
	in_case = false; // flag for parser that case/default has been processed, and next colon needs special attention
	while (true) {
		var t = get_next_token(parser_pos);
		token_text = t[0];
		token_type = t[1];
		if (token_type === 'TK_EOF') {
			break;
		}
		switch (token_type) {
		case 'TK_START_EXPR':
			var_line = false;
			set_mode('EXPRESSION');
			if (last_type === 'TK_END_EXPR' || last_type === 'TK_START_EXPR') {
				// do nothing on (( and )( and ][ and ]( ..
			} else if (last_type !== 'TK_WORD' && last_type !== 'TK_OPERATOR') {
				print_space();
			} else if (in_array(last_word, line_starters) && last_word !== 'function') {
				print_space();
			}
			print_token();
			break;
		case 'TK_END_EXPR':
			print_token();
			restore_mode();
			break;
		case 'TK_START_BLOCK':

			if (last_word === 'do') {
				set_mode('DO_BLOCK');
			} else {
				set_mode('BLOCK');
			}
			if (last_type !== 'TK_OPERATOR' && last_type !== 'TK_START_EXPR') {
				if (last_type === 'TK_START_BLOCK') {
					print_newline();
				} else {
					print_space();
				}
			}
			print_token();
			indent();
			break;
		case 'TK_END_BLOCK':
			if (last_type === 'TK_START_BLOCK') {
				// nothing
				trim_output();
				unindent();
			} else {
				unindent();
				print_newline();
			}
			print_token();
			restore_mode();
			break;
		case 'TK_WORD':
			if (do_block_just_closed) {
				print_space();
				print_token();
				print_space();
				break;
			}
			if (token_text === 'case' || token_text === 'default') {
				if (last_text === ':') {
					// switch cases following one another
					remove_indent();
				} else {
					// case statement starts in the same line where switch
					unindent();
					print_newline();
					indent();
				}
				print_token();
				in_case = true;
				break;
			}
			prefix = 'NONE';
			if (last_type === 'TK_END_BLOCK') {
				if (!in_array(token_text.toLowerCase(), ['else', 'catch', 'finally'])) {
					prefix = 'NEWLINE';
				} else {
					prefix = 'SPACE';
					print_space();
				}
			} else if (last_type === 'TK_END_COMMAND' && (current_mode === 'BLOCK' || current_mode === 'DO_BLOCK')) {
				prefix = 'NEWLINE';
			} else if (last_type === 'TK_END_COMMAND' && current_mode === 'EXPRESSION') {
				prefix = 'SPACE';
			} else if (last_type === 'TK_WORD') {
				prefix = 'SPACE';
			} else if (last_type === 'TK_START_BLOCK') {
				prefix = 'NEWLINE';
			} else if (last_type === 'TK_END_EXPR') {
				print_space();
				prefix = 'NEWLINE';
			}
			if (last_type !== 'TK_END_BLOCK' && in_array(token_text.toLowerCase(), ['else', 'catch', 'finally'])) {
				print_newline();
			} else if (in_array(token_text, line_starters) || prefix === 'NEWLINE') {
				if (last_text === 'else') {
					// no need to force newline on else break
					print_space();
				} else if ((last_type === 'TK_START_EXPR' || last_text === '=') && token_text === 'function') {
					// no need to force newline on 'function': (function
					// DONOTHING
				} else if (last_type === 'TK_WORD' && (last_text === 'return' || last_text === 'throw')) {
					// no newline between 'return nnn'
					print_space();
				} else if (last_type !== 'TK_END_EXPR') {
					if ((last_type !== 'TK_START_EXPR' || token_text !== 'var') && last_text !== ':') {
						// no need to force newline on 'var': for (var x = 0...)
						if (token_text === 'if' && last_type === 'TK_WORD' && last_word === 'else') {
							// no newline for } else if {
							print_space();
						} else {
							print_newline();
						}
					}
				} else {
					if (in_array(token_text, line_starters) && last_text !== ')') {
						print_newline();
					}
				}
			} else if (prefix === 'SPACE') {
				print_space();
			}
			print_token();
			last_word = token_text;
			if (token_text === 'var') {
				var_line = true;
				var_line_tainted = false;
			}
			break;
		case 'TK_END_COMMAND':
			print_token();
			var_line = false;
			break;
		case 'TK_STRING':
			if (last_type === 'TK_START_BLOCK' || last_type === 'TK_END_BLOCK') {
				print_newline();
			} else if (last_type === 'TK_WORD') {
				print_space();
			}
			print_token();
			break;
		case 'TK_OPERATOR':
			var start_delim = true;
			var end_delim = true;
			if (var_line && token_text !== ',') {
				var_line_tainted = true;
				if (token_text === ':') {
					var_line = false;
				}
			}
			if (token_text === ':' && in_case) {
				print_token(); // colon really asks for separate treatment
				print_newline();
				break;
			}
			in_case = false;
			if (token_text === ',') {
				if (var_line) {
					if (var_line_tainted) {
						print_token();
						print_newline();
						var_line_tainted = false;
					} else {
						print_token();
						print_space();
					}
				} else if (last_type === 'TK_END_BLOCK') {
					print_token();
					print_newline();
				} else {
					if (current_mode === 'BLOCK') {
						print_token();
						print_newline();
					} else {
						// EXPR od DO_BLOCK
						print_token();
						print_space();
					}
				}
				break;
			} else if (token_text === '--' || token_text === '++') { // unary operators special case
				if (last_text === ';') {
					// space for (;; ++i)
					start_delim = true;
					end_delim = false;
				} else {
					start_delim = false;
					end_delim = false;
				}
			} else if (token_text === '!' && last_type === 'TK_START_EXPR') {
				// special case handling: if (!a)
				start_delim = false;
				end_delim = false;
			} else if (last_type === 'TK_OPERATOR') {
				start_delim = false;
				end_delim = false;
			} else if (last_type === 'TK_END_EXPR') {
				start_delim = true;
				end_delim = true;
			} else if (token_text === '.') {
				// decimal digits or object.property
				start_delim = false;
				end_delim = false;
			} else if (token_text === ':') {
				// zz: xx
				// can't differentiate ternary op, so for now it's a ? b: c; without space before colon
				if (last_text.match(/^\d+$/)) {
					// a little help for ternary a ? 1 : 0;
					start_delim = true;
				} else {
					start_delim = false;
				}
			}
			if (start_delim) {
				print_space();
			}
			print_token();
			if (end_delim) {
				print_space();
			}
			break;
		case 'TK_BLOCK_COMMENT':
			print_newline();
			print_token();
			print_newline();
			break;
		case 'TK_COMMENT':
			// print_newline();
			print_space();
			print_token();
			print_newline();
			break;
		case 'TK_UNKNOWN':
			print_token();
			break;
		}
		last_type = token_type;
		last_text = token_text;
	}
	return output.join('');
}