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

   1 //<script type="text/javascript">
   2
   3
   4 const XObject = imports.XObject.XObject;
   5 const console = imports.console.console;
   6
   7
   8 const Token   = imports.Token.Token;
   9 const Lang    = imports.Lang.Lang;
  10
  11 /**
  12         @class Search a {@link JSDOC.TextStream} for language tokens.
  13 */
  14 const 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                 return true;
  56             }
  57
  58             while (!stream.look().eof) {
  59                 if (this.read_mlcomment(stream, tokens)) continue;
  60                 if (this.read_slcomment(stream, tokens)) continue;
  61                 if (this.read_dbquote(stream, tokens))   continue;
  62                 if (this.read_snquote(stream, tokens))   continue;
  63                 if (this.read_regx(stream, tokens))      continue;
  64                 if (this.read_numb(stream, tokens))      continue;
  65                 if (this.read_punc(stream, tokens))      continue;
  66                 if (this.read_newline(stream, tokens))   continue;
  67                 if (this.read_space(stream, tokens))     continue;
  68                 if (this.read_word(stream, tokens))      continue;
  69
  70                 // if execution reaches here then an error has happened
  71                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
  72             }
  73
  74
  75
  76             return tokens;
  77         },
  78
  79         /**
  80          * findPuncToken - find the id of a token (previous to current)
  81          * need to back check syntax..
  82          *
  83          * @arg {Array} tokens the array of tokens.
  84          * @arg {String} token data (eg. '(')
  85          * @arg {Number} offset where to start reading from
  86          * @return {Number} position of token
  87          */
  88         findPuncToken : function(tokens, data, n) {
  89             n = n || tokens.length -1;
  90             var stack = 0;
  91             while (n > -1) {
  92
  93                 if (!stack && tokens[n].data == data) {
  94                     return n;
  95                 }
  96
  97                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
  98                     stack++;
  99                     n--;
 100                     continue;
 101                 }
 102                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 103                     stack--;
 104                     n--;
 105                     continue;
 106                 }
 107
 108
 109                 n--;
 110             }
 111             return -1;
 112         },
 113         /**
 114          * lastSym - find the last token symbol
 115          * need to back check syntax..
 116          *
 117          * @arg {Array} tokens the array of tokens.
 118          * @arg {Number} offset where to start..
 119          * @return {Token} the token
 120          */
 121         lastSym : function(tokens, n) {
 122             for (var i = n-1; i >= 0; i--) {
 123                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 124             }
 125             return null;
 126         },
 127
 128
 129
 130         /**
 131             @returns {Boolean} Was the token found?
 132          */
 133         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 134             var found = "";
 135             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 136                 found += stream.next();
 137             }
 138
 139             if (found === "") {
 140                 return false;
 141             }
 142
 143             var name;
 144             if ((name = Lang.keyword(found))) {
 145                 if (found == 'return' && tokens.lastSym().data == ')') {
 146                     //Seed.print('@' + tokens.length);
 147                     var n = this.findPuncToken(tokens, ')');
 148                     //Seed.print(')@' + n);
 149                     n = this.findPuncToken(tokens, '(', n-1);
 150                     //Seed.print('(@' + n);
 151
 152                     var lt = this.lastSym(tokens, n);
 153                     print(JSON.stringify(lt));
 154                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 155                         if (!this.ignoreBadGrammer) {
 156                             throw {
 157                                 name : "ArgumentError",
 158                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 159                             }
 160                         }
 161                     }
 162
 163
 164
 165                 }
 166
 167                 tokens.push(new Token(found, "KEYW", name, this.line));
 168                 return true;
 169             }
 170             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 171                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 172                 return true;
 173             }
 174             var n = found.split('.');
 175             var p = false;
 176             var _this = this;
 177             n.forEach(function(nm) {
 178                 if (p) {
 179                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 180                 }
 181                 p=true;
 182                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 183             });
 184             return true;
 185
 186
 187         },
 188
 189         /**
 190             @returns {Boolean} Was the token found?
 191          */
 192         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 193             var found = "";
 194             var name;
 195             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 196                 found += stream.next();
 197             }
 198
 199
 200             if (found === "") {
 201                 return false;
 202             }
 203
 204             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 205                 //print("Error - comma found before " + found);
 206                 //print(JSON.stringify(tokens.lastSym(), null,4));
 207                 if (this.ignoreBadGrammer) {
 208                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 209                 } else {
 210
 211                     throw {
 212                         name : "ArgumentError",
 213                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 214                     }
 215                 }
 216             }
 217
 218             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 219             return true;
 220
 221         },
 222
 223         /**
 224             @returns {Boolean} Was the token found?
 225          */
 226         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 227             var found = "";
 228
 229             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 230                 found += stream.next();
 231             }
 232
 233             if (found === "") {
 234                 return false;
 235             }
 236             //print("WHITE = " + JSON.stringify(found));
 237             if (this.collapseWhite) found = " ";
 238             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 239             return true;
 240
 241         },
 242
 243         /**
 244             @returns {Boolean} Was the token found?
 245          */
 246         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 247             var found = "";
 248             var line = this.line;
 249             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 250                 this.line++;
 251                 found += stream.next();
 252             }
 253
 254             if (found === "") {
 255                 return false;
 256             }
 257             //this.line++;
 258             if (this.collapseWhite) {
 259                 found = "\n";
 260             }
 261              if (this.keepWhite) {
 262                 var last = tokens ? tokens.pop() : false;
 263                 if (last && last.name != "WHIT") {
 264                     tokens.push(last);
 265                 }
 266
 267                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 268             }
 269             return true;
 270         },
 271
 272         /**
 273             @returns {Boolean} Was the token found?
 274          */
 275         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 276             if (stream.look() == "/" && stream.look(1) == "*") {
 277                 var found = stream.next(2);
 278                 var c = '';
 279                 var line = this.line;
 280                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 281                     c = stream.next();
 282                     if (c == "\n") this.line++;
 283                     found += c;
 284                 }
 285
 286                 // to start doclet we allow /** or /*** but not /**/ or /****
 287                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 288                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 289                 return true;
 290             }
 291             return false;
 292         },
 293
 294         /**
 295             @returns {Boolean} Was the token found?
 296          */
 297         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 298             var found;
 299             if (
 300                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 301                 ||
 302                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 303             ) {
 304                 var line = this.line;
 305                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 306                     found += stream.next();
 307                 }
 308                 if (!stream.look().eof) {
 309                     found += stream.next();
 310                 }
 311                 if (this.keepComments) {
 312                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 313                 }
 314                 this.line++;
 315                 return true;
 316             }
 317             return false;
 318         },
 319
 320         /**
 321             @returns {Boolean} Was the token found?
 322          */
 323         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 324             if (stream.look() == "\"") {
 325                 // find terminator
 326                 var string = stream.next();
 327
 328                 while (!stream.look().eof) {
 329                     if (stream.look() == "\\") {
 330                         if (Lang.isNewline(stream.look(1))) {
 331                             do {
 332                                 stream.next();
 333                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 334                             string += "\\\n";
 335                         }
 336                         else {
 337                             string += stream.next(2);
 338                         }
 339                     }
 340                     else if (stream.look() == "\"") {
 341                         string += stream.next();
 342                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 343                         return true;
 344                     }
 345                     else {
 346                         string += stream.next();
 347                     }
 348                 }
 349             }
 350             return false; // error! unterminated string
 351         },
 352
 353         /**
 354             @returns {Boolean} Was the token found?
 355          */
 356         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 357             if (stream.look() == "'") {
 358                 // find terminator
 359                 var string = stream.next();
 360
 361                 while (!stream.look().eof) {
 362                     if (stream.look() == "\\") { // escape sequence
 363                         string += stream.next(2);
 364                     }
 365                     else if (stream.look() == "'") {
 366                         string += stream.next();
 367                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 368                         return true;
 369                     }
 370                     else {
 371                         string += stream.next();
 372                     }
 373                 }
 374             }
 375             return false; // error! unterminated string
 376         },
 377
 378         /**
 379             @returns {Boolean} Was the token found?
 380          */
 381         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 382             if (stream.look() === "0" && stream.look(1) == "x") {
 383                 return this.read_hex(stream, tokens);
 384             }
 385
 386             var found = "";
 387
 388             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 389                 found += stream.next();
 390             }
 391
 392             if (found === "") {
 393                 return false;
 394             }
 395             else {
 396                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 397                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 398                 return true;
 399             }
 400         },
 401         /*t:
 402             requires("../lib/JSDOC/TextStream.js");
 403             requires("../lib/JSDOC/Token.js");
 404             requires("../lib/JSDOC/Lang.js");
 405
 406             plan(3, "testing read_numb");
 407
 408             //// setup
 409             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 410             var tr = new TokenReader();
 411             var tokens = tr.tokenize(new TextStream(src));
 412
 413             var hexToken, octToken, decToken;
 414             for (var i = 0; i < tokens.length; i++) {
 415                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 416                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 417                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 418             }
 419             ////
 420
 421             is(decToken.data, "8.0", "decimal number is found in source.");
 422             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 423             is(octToken.data, "0777", "octal number is found in source.");
 424         */
 425
 426         /**
 427             @returns {Boolean} Was the token found?
 428          */
 429         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 430             var found = stream.next(2);
 431
 432             while (!stream.look().eof) {
 433                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 434                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 435                     return true;
 436                 }
 437                 else {
 438                     found += stream.next();
 439                 }
 440             }
 441             return false;
 442         },
 443
 444         /**
 445             @returns {Boolean} Was the token found?
 446          */
 447         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 448             var last;
 449             if (
 450                 stream.look() == "/"
 451                 &&
 452                 (
 453
 454                     (
 455                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 456                         ||
 457                         (
 458                                !last.is("NUMB")
 459                             && !last.is("NAME")
 460                             && !last.is("RIGHT_PAREN")
 461                             && !last.is("RIGHT_BRACKET")
 462                         )
 463                     )
 464                 )
 465             ) {
 466                 var regex = stream.next();
 467                 var in_brace = false;
 468                 while (!stream.look().eof) {
 469
 470                     if (stream.look() == "[") { // escape sequence
 471                         in_brace = true;
 472                         continue;
 473                     }
 474
 475                     if (stream.look() == "]") { // escape sequence
 476                         in_brace = false;
 477                         continue;
 478                     }
 479
 480                     if (stream.look() == "\\") { // escape sequence
 481                         regex += stream.next(2);
 482                         continue;
 483                     }
 484
 485                     if (!in_brace && stream.look() == "/") {
 486                         regex += stream.next();
 487
 488                         while (/[gmi]/.test(stream.look())) {
 489                             regex += stream.next();
 490                         }
 491
 492                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 493                         return true;
 494                     }
 495
 496                     regex += stream.next();
 497
 498                 }
 499                 // error: unterminated regex
 500             }
 501             return false;
 502         }
 503 });