git.roojs.org Git - gnome.introspection-doc-generator/blob - JSDOC/TokenReader.js

   1 //<script type="text/javascript">
   2
   3
   4 XObject = imports.XObject.XObject;
   5 console = imports.console.console;
   6
   7
   8 Token   = imports.Token.Token;
   9 Lang    = imports.Lang.Lang;
  10
  11 /**
  12         @class Search a {@link JSDOC.TextStream} for language tokens.
  13 */
  14 TokenReader = XObject.define(
  15     function(o) {
  16
  17         XObject.extend(this, o || {});
  18
  19     },
  20     Object,
  21     {
  22         /** @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token **/
  23         collapseWhite : false, // only reduces white space...
  24         /** @cfg {Boolean} keepDocs keep JSDOC comments **/
  25         keepDocs : true,
  26         /** @cfg {Boolean} keepWhite keep White space **/
  27         keepWhite : false,
  28         /** @cfg {Boolean} keepComments  keep all comments **/
  29         keepComments : false,
  30         /** @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] ) **/
  31         sepIdents : false,
  32         /** @cfg {String} filename name of file being parsed. **/
  33         filename : '',
  34         /** @config {Boolean} ignoreBadGrammer do not throw errors if we find stuff that might break compression **/
  35         ignoreBadGrammer : false,
  36         /**
  37          * tokenize a stream
  38          * @return {Array} of tokens
  39          *
  40          * ts = new TextStream(File.read(str));
  41          * tr = TokenReader({ keepComments : true, keepWhite : true });
  42          * tr.tokenize(ts)
  43          *
  44          */
  45         tokenize : function(/**JSDOC.TextStream*/stream) {
  46             this.line =1;
  47             var tokens = [];
  48             /**@ignore*/
  49             tokens.last    = function() { return tokens[tokens.length-1]; }
  50             /**@ignore*/
  51             tokens.lastSym = function() {
  52                 for (var i = tokens.length-1; i >= 0; i--) {
  53                     if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
  54                 }
  55             }
  56
  57             while (!stream.look().eof) {
  58                 if (this.read_mlcomment(stream, tokens)) continue;
  59                 if (this.read_slcomment(stream, tokens)) continue;
  60                 if (this.read_dbquote(stream, tokens))   continue;
  61                 if (this.read_snquote(stream, tokens))   continue;
  62                 if (this.read_regx(stream, tokens))      continue;
  63                 if (this.read_numb(stream, tokens))      continue;
  64                 if (this.read_punc(stream, tokens))      continue;
  65                 if (this.read_newline(stream, tokens))   continue;
  66                 if (this.read_space(stream, tokens))     continue;
  67                 if (this.read_word(stream, tokens))      continue;
  68
  69                 // if execution reaches here then an error has happened
  70                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
  71             }
  72
  73
  74
  75             return tokens;
  76         },
  77
  78         /**
  79          * findPuncToken - find the id of a token (previous to current)
  80          * need to back check syntax..
  81          *
  82          * @arg {Array} tokens the array of tokens.
  83          * @arg {String} token data (eg. '(')
  84          * @arg {Number} offset where to start reading from
  85          * @return {Number} position of token
  86          */
  87         findPuncToken : function(tokens, data, n) {
  88             n = n || tokens.length -1;
  89             var stack = 0;
  90             while (n > -1) {
  91
  92                 if (!stack && tokens[n].data == data) {
  93                     return n;
  94                 }
  95
  96                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
  97                     stack++;
  98                     n--;
  99                     continue;
 100                 }
 101                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 102                     stack--;
 103                     n--;
 104                     continue;
 105                 }
 106
 107
 108                 n--;
 109             }
 110             return -1;
 111         },
 112         /**
 113          * lastSym - find the last token symbol
 114          * need to back check syntax..
 115          *
 116          * @arg {Array} tokens the array of tokens.
 117          * @arg {Number} offset where to start..
 118          * @return {Token} the token
 119          */
 120         lastSym : function(tokens, n) {
 121             for (var i = n-1; i >= 0; i--) {
 122                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 123             }
 124         },
 125
 126
 127
 128         /**
 129             @returns {Boolean} Was the token found?
 130          */
 131         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 132             var found = "";
 133             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 134                 found += stream.next();
 135             }
 136
 137             if (found === "") {
 138                 return false;
 139             }
 140
 141             var name;
 142             if ((name = Lang.keyword(found))) {
 143                 if (found == 'return' && tokens.lastSym().data == ')') {
 144                     //Seed.print('@' + tokens.length);
 145                     var n = this.findPuncToken(tokens, ')');
 146                     //Seed.print(')@' + n);
 147                     n = this.findPuncToken(tokens, '(', n-1);
 148                     //Seed.print('(@' + n);
 149
 150                     var lt = this.lastSym(tokens, n);
 151                    Seed.print(JSON.stringify(lt));
 152                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 153                         if (!this.ignoreBadGrammer) {
 154                             throw {
 155                                 name : "ArgumentError",
 156                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 157                             }
 158                         }
 159                     }
 160
 161
 162
 163                 }
 164
 165                 tokens.push(new Token(found, "KEYW", name, this.line));
 166                 return true;
 167             }
 168             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 169                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 170                 return true;
 171             }
 172             var n = found.split('.');
 173             var p = false;
 174             var _this = this;
 175             n.forEach(function(nm) {
 176                 if (p) {
 177                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 178                 }
 179                 p=true;
 180                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 181             });
 182             return true;
 183
 184
 185         },
 186
 187         /**
 188             @returns {Boolean} Was the token found?
 189          */
 190         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 191             var found = "";
 192             var name;
 193             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 194                 found += stream.next();
 195             }
 196
 197
 198             if (found === "") {
 199                 return false;
 200             }
 201
 202             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 203                 //print("Error - comma found before " + found);
 204                 //print(JSON.stringify(tokens.lastSym(), null,4));
 205                 throw {
 206                     name : "ArgumentError",
 207                     message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 208                 }
 209             }
 210
 211             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 212             return true;
 213
 214         },
 215
 216         /**
 217             @returns {Boolean} Was the token found?
 218          */
 219         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 220             var found = "";
 221
 222             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 223                 found += stream.next();
 224             }
 225
 226             if (found === "") {
 227                 return false;
 228             }
 229             //print("WHITE = " + JSON.stringify(found));
 230             if (this.collapseWhite) found = " ";
 231             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 232             return true;
 233
 234         },
 235
 236         /**
 237             @returns {Boolean} Was the token found?
 238          */
 239         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 240             var found = "";
 241             var line = this.line;
 242             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 243                 this.line++;
 244                 found += stream.next();
 245             }
 246
 247             if (found === "") {
 248                 return false;
 249             }
 250             //this.line++;
 251             if (this.collapseWhite) {
 252                 found = "\n";
 253             }
 254             if (this.keepWhite) {
 255                 var last = tokens.pop();
 256                 if (last && last.name != "WHIT") {
 257                     tokens.push(last);
 258                 }
 259
 260                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 261             }
 262             return true;
 263         },
 264
 265         /**
 266             @returns {Boolean} Was the token found?
 267          */
 268         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 269             if (stream.look() == "/" && stream.look(1) == "*") {
 270                 var found = stream.next(2);
 271                 var c = '';
 272                 var line = this.line;
 273                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 274                     c = stream.next();
 275                     if (c == "\n") this.line++;
 276                     found += c;
 277                 }
 278
 279                 // to start doclet we allow /** or /*** but not /**/ or /****
 280                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 281                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 282                 return true;
 283             }
 284             return false;
 285         },
 286
 287         /**
 288             @returns {Boolean} Was the token found?
 289          */
 290         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 291             var found;
 292             if (
 293                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 294                 ||
 295                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 296             ) {
 297                 var line = this.line;
 298                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 299                     found += stream.next();
 300                 }
 301                 if (!stream.look().eof) {
 302                     found += stream.next();
 303                 }
 304                 if (this.keepComments) {
 305                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 306                 }
 307                 this.line++;
 308                 return true;
 309             }
 310             return false;
 311         },
 312
 313         /**
 314             @returns {Boolean} Was the token found?
 315          */
 316         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 317             if (stream.look() == "\"") {
 318                 // find terminator
 319                 var string = stream.next();
 320
 321                 while (!stream.look().eof) {
 322                     if (stream.look() == "\\") {
 323                         if (Lang.isNewline(stream.look(1))) {
 324                             do {
 325                                 stream.next();
 326                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 327                             string += "\\\n";
 328                         }
 329                         else {
 330                             string += stream.next(2);
 331                         }
 332                     }
 333                     else if (stream.look() == "\"") {
 334                         string += stream.next();
 335                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 336                         return true;
 337                     }
 338                     else {
 339                         string += stream.next();
 340                     }
 341                 }
 342             }
 343             return false; // error! unterminated string
 344         },
 345
 346         /**
 347             @returns {Boolean} Was the token found?
 348          */
 349         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 350             if (stream.look() == "'") {
 351                 // find terminator
 352                 var string = stream.next();
 353
 354                 while (!stream.look().eof) {
 355                     if (stream.look() == "\\") { // escape sequence
 356                         string += stream.next(2);
 357                     }
 358                     else if (stream.look() == "'") {
 359                         string += stream.next();
 360                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 361                         return true;
 362                     }
 363                     else {
 364                         string += stream.next();
 365                     }
 366                 }
 367             }
 368             return false; // error! unterminated string
 369         },
 370
 371         /**
 372             @returns {Boolean} Was the token found?
 373          */
 374         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 375             if (stream.look() === "0" && stream.look(1) == "x") {
 376                 return this.read_hex(stream, tokens);
 377             }
 378
 379             var found = "";
 380
 381             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 382                 found += stream.next();
 383             }
 384
 385             if (found === "") {
 386                 return false;
 387             }
 388             else {
 389                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 390                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 391                 return true;
 392             }
 393         },
 394         /*t:
 395             requires("../lib/JSDOC/TextStream.js");
 396             requires("../lib/JSDOC/Token.js");
 397             requires("../lib/JSDOC/Lang.js");
 398
 399             plan(3, "testing read_numb");
 400
 401             //// setup
 402             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 403             var tr = new TokenReader();
 404             var tokens = tr.tokenize(new TextStream(src));
 405
 406             var hexToken, octToken, decToken;
 407             for (var i = 0; i < tokens.length; i++) {
 408                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 409                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 410                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 411             }
 412             ////
 413
 414             is(decToken.data, "8.0", "decimal number is found in source.");
 415             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 416             is(octToken.data, "0777", "octal number is found in source.");
 417         */
 418
 419         /**
 420             @returns {Boolean} Was the token found?
 421          */
 422         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 423             var found = stream.next(2);
 424
 425             while (!stream.look().eof) {
 426                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 427                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 428                     return true;
 429                 }
 430                 else {
 431                     found += stream.next();
 432                 }
 433             }
 434             return false;
 435         },
 436
 437         /**
 438             @returns {Boolean} Was the token found?
 439          */
 440         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 441             var last;
 442             if (
 443                 stream.look() == "/"
 444                 &&
 445                 (
 446
 447                     (
 448                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 449                         ||
 450                         (
 451                                !last.is("NUMB")
 452                             && !last.is("NAME")
 453                             && !last.is("RIGHT_PAREN")
 454                             && !last.is("RIGHT_BRACKET")
 455                         )
 456                     )
 457                 )
 458             ) {
 459                 var regex = stream.next();
 460
 461                 while (!stream.look().eof) {
 462                     if (stream.look() == "\\") { // escape sequence
 463                         regex += stream.next(2);
 464                     }
 465                     else if (stream.look() == "/") {
 466                         regex += stream.next();
 467
 468                         while (/[gmi]/.test(stream.look())) {
 469                             regex += stream.next();
 470                         }
 471
 472                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 473                         return true;
 474                     }
 475                     else {
 476                         regex += stream.next();
 477                     }
 478                 }
 479                 // error: unterminated regex
 480             }
 481             return false;
 482         }
 483 });