最开始本打算做代码格式化的，等熬过了所谓的词法分析，却发现还有座大山压着，格式化的话，缩进和空格是关键，可是缩进层级的确是个大问题...必须结和上下语境才能得出正确的缩进层级，貌似这已经不是词法分析了，改算语法分析了吧...
所以，还是先记录下简单一点的高亮吧，解决了我上一篇文章说的那几个难点，或者说能解决大部分难点，基本高亮就够了...http://www.cnblogs.com/hongru/archive/2011/04/12/2000049.html
没有用正则，因为我正则不好，另外个原因就是如果要接着做格式化的话，正则是办不到的...所以借鉴了beautify.js 的方式。
/* * javascript format * ------------------ * writen by hongru.chen [hongru.chenhr[at]gmail.com] * * Options: * { * indentSize: (default 4), --> indentation size * indentChar: (default space), --> character to indentation with * isBlankLinesRetain: (default true), --> whether to preserve blank new lines * maxBlankLines: (default unlimited), --> max blank lines to preserve * isSpaceAfterAnonFunc: (default true), --> 'function()' or 'function ()' * braceStyle: (default 'collapse'), --> 'collapse' || 'expand' || 'end-expand' * isKeepArrayIndent: (default false) --> [[], []] or [ [], [] ] * } * * e.g * new jsFormat(sourceText, options) * */ (function () { var extend = function (target, source, overwrite) { if (overwrite === undefined) overwrite = true; for (var p in source) { if (overwrite || !(p in target)) target[p] = source[p]; } return target; }, log = function (s) { !!window.console && window.console.log(s); }; var jsFormat = function (source, options) { this.setOptions(options); this.setVars(source); //this.setMode(); this.initialize(); }; extend(jsFormat.prototype, { setOptions: function (o) { this.opt = { indentSize : 4, indentChar : ' ', isBlankLinesRetain : true, maxBlankLines : false, isSpaceAfterAnonFunc: true, isKeepArrayIndent : false, braceStyle : 'collapse', isParseHtml : true }; extend(this.opt, o || {}); }, setVars: function (s) { this.indentStr = ''; this.source = s; this.lastWord = ''; this.lastType = 'TK_START_EXPR'; // the default value of lastType this.lastText = ''; this.preLastText = ''; this.output = []; this.whitespace = ' \n\r\t'.split(''); this.wordchar = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_$'.split(''); this.numbers = '0123456789'.split(''); this.operator = '+ - * % & ++ -- = += -= *= /= %= == === != !== > < >= <= >> << >>> >>>= >>= <<= && &= | || ! !! , : ? ^ ^= |= ::'.split(' '); this.lineKeywords = 'continue,try,catch,finaly,throw,return,var,if,else,switch,case,default,for,while,do,break,function,this,new'.split(','); this.jsKeywords = 'getElementById,getElementsByTagName,getElementsByName,length,offsetLeft,offsetRight,offsetWidth,offsetHeight,scrollLeft,scrollTop,setInterval,clearInterval,setTimeout,clearTimeout,clientWidth,clientHeight,document,window,Date,Number,String,Object,Array,Boolean,Function,null,true,false'.split(','); this.tokenStore = []; this.parsePos = 0; this.sourceLength = s.length; this.token = {}; this.initIndentLv = 0; this.toHtml = ''; }, encodeHtml: function (str) { if (this.opt.isParseHtml) { var el = document.createElement('pre'), t = document.createTextNode(str); el.appendChild(t); return el.innerHTML; } return str; }, isInArray: function (c, arr) { for (var i = 0; i < arr.length; i ++) { if (c === arr[i]) return true; } return false; }, isArrayMode: function (mode) { return (mode === '[EXPRESSION]' || mode === '[INDENTED_EXPRESSION]'); }, // trim whitespace at the end of output // if isEatLines == true, also trim the newline space such as '\n' '\r' trimOutput: function (isEatLines) { var o = this.output, l = o.length; while(!!l && (o[l-1] === ' ' || o[l-1] === this.indentStr || (!!isEatLines && (o[l-1] === '\n' || o[l-1] === '\r')))) { this.output.pop(); } }, getToken: function () { if (this.parsePos > this.sourceLength) { return ['', 'TK_EOF']; } var c = this.source.charAt(this.parsePos); this.parsePos ++; this.blanklinesN = 0; // word if (this.isInArray(c, this.wordchar)) { return this._wordToken(c); } // expression '(' or '[' if (c === '(' || c === '[') { return [c, 'TK_START_EXPR']; } // expression ')' or ']' if (c === ')' || c === ']') { return [c, 'TK_END_EXPR']; } // brace '{' if (c === '{') { return [c, 'TK_START_BRACE']; } // brace '}' if (c === '}') { return [c, 'TK_END_BRACE']; } // semicolon ';' if (c === ';') { return [c, 'TK_SEMICOLON']; } /* !! modify at 2010 04 12 */ if (c === '/') { if (this.source.charAt(this.parsePos) === '/' || this.source.charAt(this.parsePos) === '*') { return this._commentToken(c); } else if (this.lastType === 'TK_REGEXP' || this.lastType === 'TK_STRING' //|| this.lastType === 'TK_END_BRACE' || this.lastType === 'TK_END_EXPR' || (this.lastType === 'TK_BLOCK_COMMENT' && this.preLastType === 'TK_WORD' && this.preLastText !== 'return' && this.preLastText !== 'do') || (this.lastType === 'TK_WORD' && this.lastText !== 'return' && this.lastText !== 'do')) { return [c, 'TK_OPERATOR'] } else { return this._regexpToken(c); } } // operator without '/' if (this.isInArray(c, this.operator)) { return this._operatorToken(c); } // string such as '"' or "'" if (c === '"' || c === "'") { return this._stringToken(c); } // whitespace to fit the situation of keep the array-indent of source code if (this.opt.isKeepArrayIndent && this.isArrayMode(this.tkConfig.mode) && this.isInArray(c, this.whitespace)) { var ret = this._keepindentWhitespace(c); if (ret !== undefined && ret instanceof Array) {return ret}; } else if (this.isInArray(c, this.whitespace)) { // normal whitespace var ret = this._normalWhitespace(c); if (ret !== undefined && ret instanceof Array) {return ret}; } return [c, 'TK_UNKNOWN'] }, newLine: function (isIgonreRepeat) { this.output.push('\n'); this.addLine = true; for (var i=0; i < this.tkConfig.indentLv; i++) { this.output.push(this.indentStr) } }, _operatorToken: function (c) { while(this.parsePos < this.sourceLength && this.isInArray(c + this.source.charAt(this.parsePos), this.operator)) { c += this.source.charAt(this.parsePos); this.parsePos ++; } return [c, 'TK_OPERATOR']; }, _normalWhitespace: function (c) { while(this.isInArray(this.source.charAt(this.parsePos), this.whitespace)) { /*if (c === '\n') { this.blanklinesN += (this.opt.maxBlankLines ? (this.blanklinesN <= this.opt.maxBlankLines ? 1 : 0) : 1); }*/ c += this.source.charAt(this.parsePos); if (this.parsePos > this.sourceLength) { return ['', 'TK_EOF']; } this.parsePos ++; } return [c, 'TK_WHITESPACE']; /*if (this.opt.isBlankLinesRetain) { if (this.blanklinesN > 1) { for (var i=0; i< this.blanklinesN; i++) { this.newLine(i === 0); this.addLine = true; } } } this.wantNewline = (this.blanklinesN > 0);*/ }, _keepindentWhitespace: function (c) { var spaceCnt = 0; while(this.isInArray(c, this.whitespace)) { if (c === '\n') { this.trimOutput(); // before push '\n' this.output.push('\n'); this.addLine = true; spaceCnt = 0; } else if (c === '\t') { spaceCnt += 4; // plus spaceCnt to keep indent the source has } else if (c === '\r') { // todo } else { spaceCnt ++; } if (this.parsePos > this.sourceLength) { return ['', 'TK_EOF'] } // get next char c = this.source.charAt(this.parsePos); this.parsePos ++; } // set every newline's indentation if (this.tkConfig.baseIndentation === undefined) this.tkConfig.baseIndentation = spaceCnt; if (this.addline === true) { var i; for (i = 0; i < this.tkConfig.indentLv + 1; i++) { this.output.push(this.indentStr) } if (this.tkConfig.baseIndentation !== undefined) { for (i = 0; i < spaceCnt - this.tiConfig.baseIndentation; i++) { this.output.push(' '); } } } }, _wordToken: function (c) { if (this.parsePos < this.sourceLength) { while(this.isInArray(this.source.charAt(this.parsePos), this.wordchar)) { c += this.source.charAt(this.parsePos); this.parsePos ++; if (this.parsePos === this.sourceLength) break; } } return [c, 'TK_WORD']; }, _commentToken: function (c) { var comment = '/*'; // match inline comments /* ... */ var isInline = true, curChar = this.source.charAt(this.parsePos); if (curChar === '*') { this.parsePos ++; if (this.parsePos < this.sourceLength) { while(!(this.source.charAt(this.parsePos) === '*' && !!this.source.charAt(this.parsePos + 1) && this.source.charAt(this.parsePos + 1) === '/') && this.parsePos < this.sourceLength) { // exclude /**/, get content from /*, end as */ //comment content start after /* c = this.source.charAt(this.parsePos); if (c === '\x0d' || c === '\x0a') { isInline = false; } comment += c; this.parsePos ++; if (this.parsePos > this.sourceLength) { break; } } } this.parsePos += 2; comment += '*/'; if (isInline) { return [comment, 'TK_INLINE_COMMENT']; } else { return [comment, 'TK_BLOCK_COMMENT']; } } // match comment like //... if (curChar === '/') { comment = c; while(this.source.charAt(this.parsePos) !== '\r' && this.source.charAt(this.parsePos) !== '\n') { comment += this.source.charAt(this.parsePos); this.parsePos ++; if (this.parsePos > this.sourceLength) { break; } } this.parsePos ++; return [comment+'\n', 'TK_COMMENT'] } }, _regexpToken: function (c) { var escape = false, ret = c; if (this.parsePos < this.sourceLength) { var retainExp = false; while((escape || retainExp || this.source.charAt(this.parsePos) !== c) && this.source.charAt(this.parsePos) !== '\n') { var curChar = this.source.charAt(this.parsePos); ret += curChar; if (!escape) { if (curChar === '\\') escape = true; // '/\\/'.charAt(1) === '\\' if (curChar === '[') retainExp = true; // '/[...' exp continue else if (curChar === ']') retainExp = false; // '/]....' exp search break } else {escape = false} this.parsePos ++; // if regexp continue to the end of file, return if (this.parsePos >= this.sourceLength) { return [ret, 'TK_REGEXP'] } } } this.parsePos ++; ret += c; // regexp may have modifier like /regexp/ig // here we relax the rules of the modifiers, more than i,g,m... while(this.parsePos < this.sourceLength && this.isInArray(this.source.charAt(this.parsePos), this.wordchar)) { ret += this.source.charAt(this.parsePos); this.parsePos ++; } return [ret, 'TK_REGEXP']; }, _stringToken: function (c) { var escape = false, ret = c; if (this.parsePos < this.sourceLength) { while(escape || this.source.charAt(this.parsePos) !== c) { var curChar = this.source.charAt(this.parsePos); ret += curChar; if (!escape) { if (curChar === '\\') escape = true; } else escape = false; this.parsePos ++; if (this.parsePos >= this.sourceLength) { return [ret, 'TK_STRING'] } } } this.parsePos ++; ret += c; ret = this.encodeHtml(ret); return [ret, 'TK_STRING']; }, setMode: function (mode) { // push previous token config if (!this.tkConfig) { this.tokenStore.push(this.tkConfig); } // define current token config this.tkConfig = { mode: mode, preMode: !!this.tkConfig ? this.tkConfig.mode : 'BLOCK', eatSpace: false, varLine: false, ifLine: false, varLineReindent: false, indentLv: (!!this.tkConfig ? this.tkConfig.indentLv + ((this.tkConfig.varLine && this.tkConfig.varLineReindent) ? 1 : 0) : 0) } }, initialize: function () { //console.log(this.source); this.setMode('BLOCK'); while(true) { var tk = this.getToken(); if (tk[1] !== 'TK_WHITESPACE') { // filter the whitespace this.token.text = tk[0]; this.token.type = tk[1]; //log(tk) if (tk[1] === 'TK_EOF') { //log('eof'); break; } this.preLastText = this.lastText; this.preLastType = this.lastType; this.lastText = this.token.text; this.lastType = this.token.type; } this.highLight(tk); } return this.toHtml; }, highLight: function (tk) { var cn = tk[1].toLowerCase(); if (tk[1] === 'TK_WORD' && this.isInArray(tk[0], this.lineKeywords)) { this.toHtml += ''+tk[0]+'' } else if (tk[1] === 'TK_WORD' && this.isInArray(tk[0], this.jsKeywords)){ this.toHtml += ''+tk[0]+'' } else { this.toHtml += ''+tk[0]+'' } } }) this.jsFormat = jsFormat; })();
此版本bug还不少，另外对html的转义目前只做了在字符串中的部分，其他地方的包括注释里的，正则里的等等都没做转义，如果这些地方有html标签的话，在pre里就直接渲染了...
可以自定义style，自定义关键词...
接下来，所谓的格式化，会是个痛苦的旅程。