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
  35         /**
  36          * tokenize a stream
  37          * @return {Array} of tokens
  38          *
  39          * ts = new TextStream(File.read(str));
  40          * tr = TokenReader({ keepComments : true, keepWhite : true });
  41          * tr.tokenize(ts)
  42          *
  43          */
  44         tokenize : function(/**JSDOC.TextStream*/stream) {
  45             this.line =1;
  46             var tokens = [];
  47             /**@ignore*/
  48             tokens.last    = function() { return tokens[tokens.length-1]; }
  49             /**@ignore*/
  50             tokens.lastSym = function() {
  51                 for (var i = tokens.length-1; i >= 0; i--) {
  52                     if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
  53                 }
  54             }
  55
  56             while (!stream.look().eof) {
  57                 if (this.read_mlcomment(stream, tokens)) continue;
  58                 if (this.read_slcomment(stream, tokens)) continue;
  59                 if (this.read_dbquote(stream, tokens))   continue;
  60                 if (this.read_snquote(stream, tokens))   continue;
  61                 if (this.read_regx(stream, tokens))      continue;
  62                 if (this.read_numb(stream, tokens))      continue;
  63                 if (this.read_punc(stream, tokens))      continue;
  64                 if (this.read_newline(stream, tokens))   continue;
  65                 if (this.read_space(stream, tokens))     continue;
  66                 if (this.read_word(stream, tokens))      continue;
  67
  68                 // if execution reaches here then an error has happened
  69                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
  70             }
  71
  72
  73
  74             return tokens;
  75         },
  76
  77         /**
  78          * findPuncToken - find the id of a token (previous to current)
  79          * need to back check syntax..
  80          *
  81          * @arg {Array} tokens the array of tokens.
  82          * @arg {String} token data (eg. '(')
  83          * @arg {Number} offset where to start reading from
  84          * @return {Number} position of token
  85          */
  86         findPuncToken : function(tokens, data, n) {
  87             n = n || tokens.length -1;
  88             var stack = 0;
  89             while (n > -1) {
  90
  91                 if (!stack && tokens[n].data == data) {
  92                     return n;
  93                 }
  94
  95                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
  96                     stack++;
  97                     n--;
  98                     continue;
  99                 }
 100                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 101                     stack--;
 102                     n--;
 103                     continue;
 104                 }
 105
 106
 107                 n--;
 108             }
 109             return -1;
 110         },
 111         /**
 112          * lastSym - find the last token symbol
 113          * need to back check syntax..
 114          *
 115          * @arg {Array} tokens the array of tokens.
 116          * @arg {Number} offset where to start..
 117          * @return {Token} the token
 118          */
 119         lastSym : function(tokens, n) {
 120             for (var i = n-1; i >= 0; i--) {
 121                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 122             }
 123         },
 124
 125
 126
 127         /**
 128             @returns {Boolean} Was the token found?
 129          */
 130         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 131             var found = "";
 132             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 133                 found += stream.next();
 134             }
 135
 136             if (found === "") {
 137                 return false;
 138             }
 139
 140             var name;
 141             if ((name = Lang.keyword(found))) {
 142                 if (found == 'return' && tokens.lastSym().data == ')') {
 143                     //Seed.print('@' + tokens.length);
 144                     var n = this.findPuncToken(tokens, ')');
 145                     //Seed.print(')@' + n);
 146                     n = this.findPuncToken(tokens, '(', n-1);
 147                     //Seed.print('(@' + n);
 148
 149                     var lt = this.lastSym(tokens, n);
 150                    Seed.print(JSON.stringify(lt));
 151                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 152                         throw {
 153                             name : "ArgumentError",
 154                             message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 155                         }
 156                     }
 157
 158
 159
 160                 }
 161
 162                 tokens.push(new Token(found, "KEYW", name, this.line));
 163                 return true;
 164             }
 165             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 166                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 167                 return true;
 168             }
 169             var n = found.split('.');
 170             var p = false;
 171             var _this = this;
 172             n.forEach(function(nm) {
 173                 if (p) {
 174                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 175                 }
 176                 p=true;
 177                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 178             });
 179             return true;
 180
 181
 182         },
 183
 184         /**
 185             @returns {Boolean} Was the token found?
 186          */
 187         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 188             var found = "";
 189             var name;
 190             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 191                 found += stream.next();
 192             }
 193
 194
 195             if (found === "") {
 196                 return false;
 197             }
 198
 199             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 200                 //print("Error - comma found before " + found);
 201                 //print(JSON.stringify(tokens.lastSym(), null,4));
 202                 throw {
 203                     name : "ArgumentError",
 204                     message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 205                 }
 206             }
 207
 208             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 209             return true;
 210
 211         },
 212
 213         /**
 214             @returns {Boolean} Was the token found?
 215          */
 216         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 217             var found = "";
 218
 219             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 220                 found += stream.next();
 221             }
 222
 223             if (found === "") {
 224                 return false;
 225             }
 226             //print("WHITE = " + JSON.stringify(found));
 227             if (this.collapseWhite) found = " ";
 228             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 229             return true;
 230
 231         },
 232
 233         /**
 234             @returns {Boolean} Was the token found?
 235          */
 236         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 237             var found = "";
 238             var line = this.line;
 239             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 240                 this.line++;
 241                 found += stream.next();
 242             }
 243
 244             if (found === "") {
 245                 return false;
 246             }
 247             //this.line++;
 248             if (this.collapseWhite) {
 249                 found = "\n";
 250             }
 251             if (this.keepWhite) {
 252                 var last = tokens.pop();
 253                 if (last && last.name != "WHIT") {
 254                     tokens.push(last);
 255                 }
 256
 257                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 258             }
 259             return true;
 260         },
 261
 262         /**
 263             @returns {Boolean} Was the token found?
 264          */
 265         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 266             if (stream.look() == "/" && stream.look(1) == "*") {
 267                 var found = stream.next(2);
 268                 var c = '';
 269                 var line = this.line;
 270                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 271                     c = stream.next();
 272                     if (c == "\n") this.line++;
 273                     found += c;
 274                 }
 275
 276                 // to start doclet we allow /** or /*** but not /**/ or /****
 277                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 278                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 279                 return true;
 280             }
 281             return false;
 282         },
 283
 284         /**
 285             @returns {Boolean} Was the token found?
 286          */
 287         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 288             var found;
 289             if (
 290                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 291                 ||
 292                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 293             ) {
 294                 var line = this.line;
 295                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 296                     found += stream.next();
 297                 }
 298                 if (!stream.look().eof) {
 299                     found += stream.next();
 300                 }
 301                 if (this.keepComments) {
 302                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 303                 }
 304                 this.line++;
 305                 return true;
 306             }
 307             return false;
 308         },
 309
 310         /**
 311             @returns {Boolean} Was the token found?
 312          */
 313         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 314             if (stream.look() == "\"") {
 315                 // find terminator
 316                 var string = stream.next();
 317
 318                 while (!stream.look().eof) {
 319                     if (stream.look() == "\\") {
 320                         if (Lang.isNewline(stream.look(1))) {
 321                             do {
 322                                 stream.next();
 323                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 324                             string += "\\\n";
 325                         }
 326                         else {
 327                             string += stream.next(2);
 328                         }
 329                     }
 330                     else if (stream.look() == "\"") {
 331                         string += stream.next();
 332                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 333                         return true;
 334                     }
 335                     else {
 336                         string += stream.next();
 337                     }
 338                 }
 339             }
 340             return false; // error! unterminated string
 341         },
 342
 343         /**
 344             @returns {Boolean} Was the token found?
 345          */
 346         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 347             if (stream.look() == "'") {
 348                 // find terminator
 349                 var string = stream.next();
 350
 351                 while (!stream.look().eof) {
 352                     if (stream.look() == "\\") { // escape sequence
 353                         string += stream.next(2);
 354                     }
 355                     else if (stream.look() == "'") {
 356                         string += stream.next();
 357                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 358                         return true;
 359                     }
 360                     else {
 361                         string += stream.next();
 362                     }
 363                 }
 364             }
 365             return false; // error! unterminated string
 366         },
 367
 368         /**
 369             @returns {Boolean} Was the token found?
 370          */
 371         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 372             if (stream.look() === "0" && stream.look(1) == "x") {
 373                 return this.read_hex(stream, tokens);
 374             }
 375
 376             var found = "";
 377
 378             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 379                 found += stream.next();
 380             }
 381
 382             if (found === "") {
 383                 return false;
 384             }
 385             else {
 386                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 387                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 388                 return true;
 389             }
 390         },
 391         /*t:
 392             requires("../lib/JSDOC/TextStream.js");
 393             requires("../lib/JSDOC/Token.js");
 394             requires("../lib/JSDOC/Lang.js");
 395
 396             plan(3, "testing read_numb");
 397
 398             //// setup
 399             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 400             var tr = new TokenReader();
 401             var tokens = tr.tokenize(new TextStream(src));
 402
 403             var hexToken, octToken, decToken;
 404             for (var i = 0; i < tokens.length; i++) {
 405                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 406                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 407                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 408             }
 409             ////
 410
 411             is(decToken.data, "8.0", "decimal number is found in source.");
 412             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 413             is(octToken.data, "0777", "octal number is found in source.");
 414         */
 415
 416         /**
 417             @returns {Boolean} Was the token found?
 418          */
 419         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 420             var found = stream.next(2);
 421
 422             while (!stream.look().eof) {
 423                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 424                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 425                     return true;
 426                 }
 427                 else {
 428                     found += stream.next();
 429                 }
 430             }
 431             return false;
 432         },
 433
 434         /**
 435             @returns {Boolean} Was the token found?
 436          */
 437         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 438             var last;
 439             if (
 440                 stream.look() == "/"
 441                 &&
 442                 (
 443
 444                     (
 445                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 446                         ||
 447                         (
 448                                !last.is("NUMB")
 449                             && !last.is("NAME")
 450                             && !last.is("RIGHT_PAREN")
 451                             && !last.is("RIGHT_BRACKET")
 452                         )
 453                     )
 454                 )
 455             ) {
 456                 var regex = stream.next();
 457
 458                 while (!stream.look().eof) {
 459                     if (stream.look() == "\\") { // escape sequence
 460                         regex += stream.next(2);
 461                     }
 462                     else if (stream.look() == "/") {
 463                         regex += stream.next();
 464
 465                         while (/[gmi]/.test(stream.look())) {
 466                             regex += stream.next();
 467                         }
 468
 469                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 470                         return true;
 471                     }
 472                     else {
 473                         regex += stream.next();
 474                     }
 475                 }
 476                 // error: unterminated regex
 477             }
 478             return false;
 479         }
 480 });