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

   1 //<script type="text/javascript">
   2
   3
   4
   5
   6 //const Token   = imports.Token.Token;
   7 //const Lang    = imports.Lang.Lang;
   8
   9 /**
  10         @class Search a {@link JSDOC.TextStream} for language tokens.
  11 */
  12
  13 namespace JSDOC {
  14
  15     public class TokenArray: Object {
  16
  17         Gee.ArrayList<Token> tokens;
  18
  19         public TokenArray()
  20         {
  21             this.items = new Gee.ArrayList<Token>();
  22         }
  23
  24         public Token? last() {
  25             if (this.tokens > 0) {
  26                 return this.tokens[this.tokens.length-1];
  27             }
  28             return null;
  29         }
  30         public Token? lastSym () {
  31             for (var i = this.tokens.length-1; i >= 0; i--) {
  32                 if (!(this.tokens.get(i).is("WHIT") || this.tokens.get(i).is("COMM")))  {
  33                     return this.tokens.get(i);
  34                 }
  35             }
  36             return null;
  37         }
  38         public void push (Token t) {
  39             this.tokens.add(t);
  40         }
  41     }
  42
  43
  44     public class TokenReader : Object
  45     {
  46
  47
  48
  49         /*
  50          *
  51          * I wonder if this will accept the prop: value, prop2 :value construxtor if we do not define one...
  52          */
  53
  54         /** @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token **/
  55         public bool collapseWhite = false, // only reduces white space...
  56         /** @cfg {Boolean} keepDocs keep JSDOC comments **/
  57         public bool keepDocs = true,
  58         /** @cfg {Boolean} keepWhite keep White space **/
  59         public bool keepWhite = false,
  60         /** @cfg {Boolean} keepComments  keep all comments **/
  61         public bool keepComments = false,
  62         /** @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] ) **/
  63         public bool sepIdents = false,
  64         /** @cfg {String} filename name of file being parsed. **/
  65         public string filename = "";
  66         /** @config {Boolean} ignoreBadGrammer do not throw errors if we find stuff that might break compression **/
  67         public bool ignoreBadGrammer = false,
  68
  69
  70         int line = 0;
  71
  72         /**
  73          * tokenize a stream
  74          * @return {Array} of tokens
  75          *
  76          * ts = new TextStream(File.read(str));
  77          * tr = TokenReader({ keepComments : true, keepWhite : true });
  78          * tr.tokenize(ts)
  79          *
  80          */
  81         public TokenArray tokenize(TextStream stream)
  82         {
  83             this.line =1;
  84             var tokens = new TokenArray();
  85
  86             bool eof;
  87             while (true) {
  88
  89                 stream.look(0, out eof)
  90                 if (eof) {
  91                     break;
  92                 }
  93                 if (this.read_mlcomment(stream, tokens)) continue;
  94                 if (this.read_slcomment(stream, tokens)) continue;
  95                 if (this.read_dbquote(stream, tokens))   continue;
  96                 if (this.read_snquote(stream, tokens))   continue;
  97                 if (this.read_regx(stream, tokens))      continue;
  98                 if (this.read_numb(stream, tokens))      continue;
  99                 if (this.read_punc(stream, tokens))      continue;
 100                 if (this.read_newline(stream, tokens))   continue;
 101                 if (this.read_space(stream, tokens))     continue;
 102                 if (this.read_word(stream, tokens))      continue;
 103
 104                 // if execution reaches here then an error has happened
 105                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
 106             }
 107
 108
 109
 110             return tokens;
 111         },
 112
 113         /**
 114          * findPuncToken - find the id of a token (previous to current)
 115          * need to back check syntax..
 116          *
 117          * @arg {Array} tokens the array of tokens.
 118          * @arg {String} token data (eg. '(')
 119          * @arg {Number} offset where to start reading from
 120          * @return {Number} position of token
 121          */
 122         findPuncToken : function(tokens, data, n) {
 123             n = n || tokens.length -1;
 124             var stack = 0;
 125             while (n > -1) {
 126
 127                 if (!stack && tokens[n].data == data) {
 128                     return n;
 129                 }
 130
 131                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
 132                     stack++;
 133                     n--;
 134                     continue;
 135                 }
 136                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 137                     stack--;
 138                     n--;
 139                     continue;
 140                 }
 141
 142
 143                 n--;
 144             }
 145             return -1;
 146         },
 147         /**
 148          * lastSym - find the last token symbol
 149          * need to back check syntax..
 150          *
 151          * @arg {Array} tokens the array of tokens.
 152          * @arg {Number} offset where to start..
 153          * @return {Token} the token
 154          */
 155         lastSym : function(tokens, n) {
 156             for (var i = n-1; i >= 0; i--) {
 157                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 158             }
 159             return null;
 160         },
 161
 162
 163
 164         /**
 165             @returns {Boolean} Was the token found?
 166          */
 167         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 168             var found = "";
 169             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 170                 found += stream.next();
 171             }
 172
 173             if (found === "") {
 174                 return false;
 175             }
 176
 177             var name;
 178             if ((name = Lang.keyword(found))) {
 179                 if (found == 'return' && tokens.lastSym().data == ')') {
 180                     //Seed.print('@' + tokens.length);
 181                     var n = this.findPuncToken(tokens, ')');
 182                     //Seed.print(')@' + n);
 183                     n = this.findPuncToken(tokens, '(', n-1);
 184                     //Seed.print('(@' + n);
 185
 186                     var lt = this.lastSym(tokens, n);
 187                     print(JSON.stringify(lt));
 188                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 189                         if (!this.ignoreBadGrammer) {
 190                             throw {
 191                                 name : "ArgumentError",
 192                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 193                             }
 194                         }
 195                     }
 196
 197
 198
 199                 }
 200
 201                 tokens.push(new Token(found, "KEYW", name, this.line));
 202                 return true;
 203             }
 204             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 205                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 206                 return true;
 207             }
 208             var n = found.split('.');
 209             var p = false;
 210             var _this = this;
 211             n.forEach(function(nm) {
 212                 if (p) {
 213                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 214                 }
 215                 p=true;
 216                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 217             });
 218             return true;
 219
 220
 221         },
 222
 223         /**
 224             @returns {Boolean} Was the token found?
 225          */
 226         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 227             var found = "";
 228             var name;
 229             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 230                 found += stream.next();
 231             }
 232
 233
 234             if (found === "") {
 235                 return false;
 236             }
 237
 238             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 239                 //print("Error - comma found before " + found);
 240                 //print(JSON.stringify(tokens.lastSym(), null,4));
 241                 if (this.ignoreBadGrammer) {
 242                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 243                 } else {
 244
 245                     throw {
 246                         name : "ArgumentError",
 247                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 248                     }
 249                 }
 250             }
 251
 252             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 253             return true;
 254
 255         },
 256
 257         /**
 258             @returns {Boolean} Was the token found?
 259          */
 260         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 261             var found = "";
 262
 263             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 264                 found += stream.next();
 265             }
 266
 267             if (found === "") {
 268                 return false;
 269             }
 270             //print("WHITE = " + JSON.stringify(found));
 271             if (this.collapseWhite) found = " ";
 272             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 273             return true;
 274
 275         },
 276
 277         /**
 278             @returns {Boolean} Was the token found?
 279          */
 280         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 281             var found = "";
 282             var line = this.line;
 283             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 284                 this.line++;
 285                 found += stream.next();
 286             }
 287
 288             if (found === "") {
 289                 return false;
 290             }
 291             //this.line++;
 292             if (this.collapseWhite) {
 293                 found = "\n";
 294             }
 295              if (this.keepWhite) {
 296                 var last = tokens ? tokens.pop() : false;
 297                 if (last && last.name != "WHIT") {
 298                     tokens.push(last);
 299                 }
 300
 301                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 302             }
 303             return true;
 304         },
 305
 306         /**
 307             @returns {Boolean} Was the token found?
 308          */
 309         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 310             if (stream.look() == "/" && stream.look(1) == "*") {
 311                 var found = stream.next(2);
 312                 var c = '';
 313                 var line = this.line;
 314                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 315                     c = stream.next();
 316                     if (c == "\n") this.line++;
 317                     found += c;
 318                 }
 319
 320                 // to start doclet we allow /** or /*** but not /**/ or /****
 321                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 322                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 323                 return true;
 324             }
 325             return false;
 326         },
 327
 328         /**
 329             @returns {Boolean} Was the token found?
 330          */
 331         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 332             var found;
 333             if (
 334                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 335                 ||
 336                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 337             ) {
 338                 var line = this.line;
 339                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 340                     found += stream.next();
 341                 }
 342                 if (!stream.look().eof) {
 343                     found += stream.next();
 344                 }
 345                 if (this.keepComments) {
 346                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 347                 }
 348                 this.line++;
 349                 return true;
 350             }
 351             return false;
 352         },
 353
 354         /**
 355             @returns {Boolean} Was the token found?
 356          */
 357         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 358             if (stream.look() == "\"") {
 359                 // find terminator
 360                 var string = stream.next();
 361
 362                 while (!stream.look().eof) {
 363                     if (stream.look() == "\\") {
 364                         if (Lang.isNewline(stream.look(1))) {
 365                             do {
 366                                 stream.next();
 367                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 368                             string += "\\\n";
 369                         }
 370                         else {
 371                             string += stream.next(2);
 372                         }
 373                     }
 374                     else if (stream.look() == "\"") {
 375                         string += stream.next();
 376                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 377                         return true;
 378                     }
 379                     else {
 380                         string += stream.next();
 381                     }
 382                 }
 383             }
 384             return false; // error! unterminated string
 385         },
 386
 387         /**
 388             @returns {Boolean} Was the token found?
 389          */
 390         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 391             if (stream.look() == "'") {
 392                 // find terminator
 393                 var string = stream.next();
 394
 395                 while (!stream.look().eof) {
 396                     if (stream.look() == "\\") { // escape sequence
 397                         string += stream.next(2);
 398                     }
 399                     else if (stream.look() == "'") {
 400                         string += stream.next();
 401                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 402                         return true;
 403                     }
 404                     else {
 405                         string += stream.next();
 406                     }
 407                 }
 408             }
 409             return false; // error! unterminated string
 410         },
 411
 412         /**
 413             @returns {Boolean} Was the token found?
 414          */
 415         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 416             if (stream.look() === "0" && stream.look(1) == "x") {
 417                 return this.read_hex(stream, tokens);
 418             }
 419
 420             var found = "";
 421
 422             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 423                 found += stream.next();
 424             }
 425
 426             if (found === "") {
 427                 return false;
 428             }
 429             else {
 430                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 431                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 432                 return true;
 433             }
 434         },
 435         /*t:
 436             requires("../lib/JSDOC/TextStream.js");
 437             requires("../lib/JSDOC/Token.js");
 438             requires("../lib/JSDOC/Lang.js");
 439
 440             plan(3, "testing read_numb");
 441
 442             //// setup
 443             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 444             var tr = new TokenReader();
 445             var tokens = tr.tokenize(new TextStream(src));
 446
 447             var hexToken, octToken, decToken;
 448             for (var i = 0; i < tokens.length; i++) {
 449                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 450                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 451                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 452             }
 453             ////
 454
 455             is(decToken.data, "8.0", "decimal number is found in source.");
 456             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 457             is(octToken.data, "0777", "octal number is found in source.");
 458         */
 459
 460         /**
 461             @returns {Boolean} Was the token found?
 462          */
 463         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 464             var found = stream.next(2);
 465
 466             while (!stream.look().eof) {
 467                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 468                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 469                     return true;
 470                 }
 471                 else {
 472                     found += stream.next();
 473                 }
 474             }
 475             return false;
 476         },
 477
 478         /**
 479             @returns {Boolean} Was the token found?
 480          */
 481         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 482             var last;
 483             if (
 484                 stream.look() == "/"
 485                 &&
 486                 (
 487
 488                     (
 489                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 490                         ||
 491                         (
 492                                !last.is("NUMB")
 493                             && !last.is("NAME")
 494                             && !last.is("RIGHT_PAREN")
 495                             && !last.is("RIGHT_BRACKET")
 496                         )
 497                     )
 498                 )
 499             ) {
 500                 var regex = stream.next();
 501
 502                 while (!stream.look().eof) {
 503                     if (stream.look() == "\\") { // escape sequence
 504                         regex += stream.next(2);
 505                     }
 506                     else if (stream.look() == "/") {
 507                         regex += stream.next();
 508
 509                         while (/[gmi]/.test(stream.look())) {
 510                             regex += stream.next();
 511                         }
 512
 513                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 514                         return true;
 515                     }
 516                     else {
 517                         regex += stream.next();
 518                     }
 519                 }
 520                 // error: unterminated regex
 521             }
 522             return false;
 523         }
 524 });