git.roojs.org Git - app.jsdoc/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         @scope JSDOC
  14
  15 */
  16 TokenReader = XObject.define(
  17     function(o) {
  18
  19         XObject.extend(this, o || {});
  20        //print("READER: " + JSON.stringify(o, null,4));
  21     },
  22     Object,
  23     {
  24         /**
  25          * @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token
  26          */
  27         collapseWhite : false, // only reduces white space...
  28         /**
  29          * @cfg {Boolean} keepDocs keep JSDOC comments
  30          */
  31         keepDocs : true,
  32         /**
  33          * @cfg {Boolean} keepWhite keep White space
  34          */
  35         keepWhite : false,
  36         /**
  37          * @cfg {Boolean} keepComments  keep all comments
  38          */
  39         keepComments : false,
  40         /**
  41          * @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] )
  42          */
  43         sepIdents : false,
  44         /**
  45          * @cfg {String} filename name of file being parsed.
  46          */
  47         filename : '',
  48         /**
  49          * @cfg {Boolean} ignoreBadGrammer do not throw errors if we find stuff that might break compression
  50          */
  51         ignoreBadGrammer : false,
  52         /**
  53          * @cfg {String} lang what language to parse..
  54          */
  55         lang : 'js',
  56         /**
  57          * tokenize a stream
  58          * @param {JSDOC.TextStream} stream The text stream to tokenize
  59          * @return {Array} of tokens
  60          *
  61          * ts = new TextStream(File.read(str));
  62          * tr = TokenReader({ keepComments : true, keepWhite : true });
  63          * tr.tokenize(ts)
  64          *
  65          */
  66         tokenize : function(stream) {
  67             this.line =1;
  68             var tokens = [];
  69             /**@ignore*/
  70             tokens.last    = function() { return tokens[tokens.length-1]; }
  71             /**@ignore*/
  72             tokens.lastSym = function() {
  73                 for (var i = tokens.length-1; i >= 0; i--) {
  74                     if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
  75                 }
  76                                 return null;
  77             }
  78
  79             var incode = this.lang == 'php' ?  0 : -1;
  80
  81
  82
  83
  84             while (!stream.look().eof) {
  85
  86
  87                 if (incode == 0) {
  88                     if (this.read_codestart(stream, tokens)) {
  89                         incode = 1;
  90                     }
  91                 }
  92                 if (incode == 1) {
  93                     if (this.read_codeend(stream, tokens)) {
  94                         incode = 0;
  95                         continue;
  96                     }
  97                 }
  98                 if (this.read_shebang(stream, tokens))      continue;
  99
 100                 if (this.read_mlcomment(stream, tokens)) continue;
 101                 if (this.read_slcomment(stream, tokens)) continue;
 102                 if (this.read_dbquote(stream, tokens))   continue;
 103                 if (this.read_snquote(stream, tokens))   continue;
 104                 if (this.read_regx(stream, tokens))      continue;
 105                 if (this.read_numb(stream, tokens))      continue;
 106                 if (this.read_punc(stream, tokens))      continue;
 107                 if (this.read_newline(stream, tokens))   continue;
 108                 if (this.read_space(stream, tokens))     continue;
 109                 if (this.read_word(stream, tokens))      continue;
 110                 // if execution reaches here then an error has happened
 111                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
 112             }
 113
 114
 115
 116             return tokens;
 117         },
 118
 119         /**
 120          * findPuncToken - find the id of a token (previous to current)
 121          * need to back check syntax..
 122          *
 123          * @arg {Array} tokens the array of tokens.
 124          * @arg {String} token data (eg. '(')
 125          * @arg {Number} offset where to start reading from
 126          * @return {Number} position of token
 127          */
 128         findPuncToken : function(tokens, data, n)
 129                 {
 130             n = n || tokens.length -1;
 131             var stack = 0;
 132             while (n > -1) {
 133
 134                 if (!stack && tokens[n].data == data) {
 135                     return n;
 136                 }
 137
 138                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
 139                     stack++;
 140                     n--;
 141                     continue;
 142                 }
 143                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 144                     stack--;
 145                     n--;
 146                     continue;
 147                 }
 148
 149
 150                 n--;
 151             }
 152             return -1;
 153         },
 154         /**
 155          * lastSym - find the last token symbol
 156          * need to back check syntax..
 157          *
 158          * @arg {Array} tokens the array of tokens.
 159          * @arg {Number} offset where to start..
 160          * @return {Token} the token
 161          */
 162         lastSym : function(tokens, n) {
 163             for (var i = n-1; i >= 0; i--) {
 164                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 165             }
 166             return null;
 167         },
 168
 169
 170
 171         /**
 172             @returns {Boolean} Was the token found?
 173          */
 174         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 175             var found = "";
 176             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 177                 found += stream.next();
 178             }
 179
 180             if (found === "") {
 181                 return false;
 182             }
 183
 184             var name;
 185             if ((name = Lang.keyword(found))) {
 186                 if (found == 'return' && tokens.lastSym().data == ')') {
 187                     //Seed.print('@' + tokens.length);
 188                     var n = this.findPuncToken(tokens, ')');
 189                     //Seed.print(')@' + n);
 190                     n = this.findPuncToken(tokens, '(', n-1);
 191                     //Seed.print('(@' + n);
 192
 193                     var lt = this.lastSym(tokens, n);
 194                    //Seed.print(JSON.stringify(lt));
 195                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 196                         if (!this.ignoreBadGrammer) {
 197                             throw {
 198                                 name : "ArgumentError",
 199                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 200                             }
 201                         }
 202                     }
 203
 204
 205
 206                 }
 207
 208                 tokens.push(new Token(found, "KEYW", name, this.line));
 209                 return true;
 210             }
 211             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 212                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 213                 return true;
 214             }
 215             var n = found.split('.');
 216             var p = false;
 217             var _this = this;
 218             n.forEach(function(nm) {
 219                 if (p) {
 220                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 221                 }
 222                 p=true;
 223                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 224             });
 225             return true;
 226
 227
 228         },
 229
 230         /**
 231             @returns {Boolean} Was the token found?
 232          */
 233         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 234             var found = "";
 235             var name;
 236             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 237                 found += stream.next();
 238             }
 239
 240
 241             if (found === "") {
 242                 return false;
 243             }
 244
 245             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 246                 //print("Error - comma found before " + found);
 247                 //print(JSON.stringify(tokens.lastSym(), null,4));
 248                 if (this.ignoreBadGrammer) {
 249                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 250                 } else {
 251
 252                     throw {
 253                         name : "ArgumentError",
 254                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 255                     }
 256                 }
 257             }
 258
 259             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 260             return true;
 261
 262         },
 263
 264         /**
 265             @returns {Boolean} Was the token found?
 266          */
 267         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 268             var found = "";
 269
 270             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 271                 found += stream.next();
 272             }
 273
 274             if (found === "") {
 275                 return false;
 276             }
 277             //print("WHITE = " + JSON.stringify(found));
 278             if (this.collapseWhite) found = " ";
 279             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 280             return true;
 281
 282         },
 283
 284         /**
 285             @returns {Boolean} Was the token found?
 286          */
 287         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 288             var found = "";
 289             var line = this.line;
 290             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 291                 this.line++;
 292                 found += stream.next();
 293             }
 294
 295             if (found === "") {
 296                 return false;
 297             }
 298             //this.line++;
 299             if (this.collapseWhite) {
 300                 found = "\n";
 301             }
 302             if (this.keepWhite) {
 303                 var last = tokens.pop();
 304                 if (last && last.name != "WHIT") {
 305                     tokens.push(last);
 306                 }
 307
 308                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 309             }
 310             return true;
 311         },
 312
 313         /**
 314             @returns {Boolean} Was the token found?
 315          */
 316         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 317             if (stream.look() == "/" && stream.look(1) == "*") {
 318                 var found = stream.next(2);
 319                 var c = '';
 320                 var line = this.line;
 321                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 322                     c = stream.next();
 323                     if (c == "\n") this.line++;
 324                     found += c;
 325                 }
 326
 327                 // to start doclet we allow /** or /*** but not /**/ or /****
 328                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 329                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 330                 return true;
 331             }
 332             return false;
 333         },
 334
 335         /**
 336             @returns {Boolean} Was the token found?
 337          */
 338         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 339             var found;
 340             if (
 341                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 342                 ||
 343                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 344             ) {
 345                 var line = this.line;
 346                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 347                     found += stream.next();
 348                 }
 349                 if (!stream.look().eof) {
 350                     found += stream.next();
 351                 }
 352                 if (this.keepComments) {
 353                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 354                 }
 355                 this.line++;
 356                 return true;
 357             }
 358             return false;
 359         },
 360
 361         /**
 362             @returns {Boolean} Was the token found?
 363          */
 364         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 365             if (stream.look() == "\"") {
 366                 // find terminator
 367                 var string = stream.next();
 368
 369                 while (!stream.look().eof) {
 370                     if (stream.look() == "\\") {
 371                         if (Lang.isNewline(stream.look(1))) {
 372                             do {
 373                                 stream.next();
 374                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 375                             string += "\\\n";
 376                         }
 377                         else {
 378                             string += stream.next(2);
 379                         }
 380                     }
 381                     else if (stream.look() == "\"") {
 382                         string += stream.next();
 383                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 384                         return true;
 385                     }
 386                     else {
 387                         string += stream.next();
 388                     }
 389                 }
 390             }
 391             return false; // error! unterminated string
 392         },
 393
 394         /**
 395             @returns {Boolean} Was the token found?
 396          */
 397         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 398             if (stream.look() == "'") {
 399                 // find terminator
 400                 var string = stream.next();
 401
 402                 while (!stream.look().eof) {
 403                     if (stream.look() == "\\") { // escape sequence
 404                         string += stream.next(2);
 405                     }
 406                     else if (stream.look() == "'") {
 407                         string += stream.next();
 408                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 409                         return true;
 410                     }
 411                     else {
 412                         string += stream.next();
 413                     }
 414                 }
 415             }
 416             return false; // error! unterminated string
 417         },
 418
 419         /**
 420             @returns {Boolean} Was the token found?
 421          */
 422         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 423             if (stream.look() === "0" && stream.look(1) == "x") {
 424                 return this.read_hex(stream, tokens);
 425             }
 426
 427             var found = "";
 428
 429             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 430                 found += stream.next();
 431             }
 432
 433             if (found === "") {
 434                 return false;
 435             }
 436             else {
 437                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 438                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 439                 return true;
 440             }
 441         },
 442         /*t:
 443             requires("../lib/JSDOC/TextStream.js");
 444             requires("../lib/JSDOC/Token.js");
 445             requires("../lib/JSDOC/Lang.js");
 446
 447             plan(3, "testing read_numb");
 448
 449             //// setup
 450             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 451             var tr = new TokenReader();
 452             var tokens = tr.tokenize(new TextStream(src));
 453
 454             var hexToken, octToken, decToken;
 455             for (var i = 0; i < tokens.length; i++) {
 456                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 457                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 458                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 459             }
 460             ////
 461
 462             is(decToken.data, "8.0", "decimal number is found in source.");
 463             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 464             is(octToken.data, "0777", "octal number is found in source.");
 465         */
 466
 467         /**
 468             @returns {Boolean} Was the token found?
 469          */
 470         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 471             var found = stream.next(2);
 472
 473             while (!stream.look().eof) {
 474                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 475                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 476                     return true;
 477                 }
 478                 else {
 479                     found += stream.next();
 480                 }
 481             }
 482             return false;
 483         },
 484        /**
 485          * read_shebang ( #!/usr/bin/seed
 486          * @param {JSDOC.TokenStream} stream the text stream
 487          * @param {Array} tokens
 488          *  @returns {Boolean} Was the token found?
 489          */
 490         read_shebang : function(/**JSDOC.TokenStream*/stream, tokens)
 491         {
 492             if (stream.cursor != 0) {
 493                 return false;
 494             }
 495
 496             if (stream.look() == '#' && stream.look(1) == '!') {
 497                 while (stream.next() != "\n") { }; //eof?
 498                 return true;
 499             }
 500             return false;
 501         },
 502
 503         /**
 504          * read_codeend (currently on PHP supported?)
 505          *  @returns {Boolean} Was the token found?
 506          */
 507         read_codeend : function(/**JSDOC.TokenStream*/stream, tokens)
 508         {
 509             var found = '';
 510             if (stream.look() == '?' && stream.look(1) == '>') {
 511                 stream.next(2);
 512                 return true;
 513             }
 514             return false;
 515         },
 516
 517         /**
 518          * read_codestart (currently on PHP supported?)
 519          *  @returns {Boolean} Was the token found?
 520          */
 521         read_codestart : function(/**JSDOC.TokenStream*/stream, tokens) {
 522             var found = '';
 523
 524             while (!stream.look().eof) {
 525                 var c = stream.next(1);
 526
 527                 if (c != '<') {
 528                     found += c;;
 529                     continue;
 530                 }
 531                 // got '<'
 532                 found += c;;
 533                 c = stream.next(1);
 534                 if (c != '?') {
 535                     found += c;;
 536                     continue;
 537                 }
 538                 found += c;
 539                 var php = stream.next(3).toLowerCase();
 540                 print(php);
 541                 if (php != 'php') {
 542                     found += c;
 543                     continue;
 544                 }
 545                 // got <?php
 546                 return true;
 547             }
 548             return false;
 549         },
 550
 551         /**
 552             @returns {Boolean} Was the token found?
 553          */
 554         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 555             var last;
 556             if (
 557                 stream.look() == "/"
 558                 &&
 559                 (
 560
 561                     (
 562                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 563                         ||
 564                         (
 565                                !last.is("NUMB")
 566                             && !last.is("NAME")
 567                             && !last.is("RIGHT_PAREN")
 568                             && !last.is("RIGHT_BRACKET")
 569                         )
 570                     )
 571                 )
 572             ) {
 573                 var regex = stream.next();
 574
 575                 while (!stream.look().eof) {
 576                     if (stream.look() == "\\") { // escape sequence
 577                         regex += stream.next(2);
 578                     }
 579                     else if (stream.look() == "/") {
 580                         regex += stream.next();
 581
 582                         while (/[gmi]/.test(stream.look())) {
 583                             regex += stream.next();
 584                         }
 585
 586                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 587                         return true;
 588                     }
 589                     else {
 590                         regex += stream.next();
 591                     }
 592                 }
 593                 // error: unterminated regex
 594             }
 595             return false;
 596         }
 597 });