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