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