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