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(
 106                         new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line)
 107                 );
 108             }
 109
 110
 111
 112             return tokens;
 113         },
 114
 115         /**
 116          * findPuncToken - find the id of a token (previous to current)
 117          * need to back check syntax..
 118          *
 119          * @arg {Array} tokens the array of tokens.
 120          * @arg {String} token data (eg. '(')
 121          * @arg {Number} offset where to start reading from
 122          * @return {Number} position of token
 123          */
 124         findPuncToken : function(tokens, data, n) {
 125             n = n || tokens.length -1;
 126             var stack = 0;
 127             while (n > -1) {
 128
 129                 if (!stack && tokens[n].data == data) {
 130                     return n;
 131                 }
 132
 133                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
 134                     stack++;
 135                     n--;
 136                     continue;
 137                 }
 138                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 139                     stack--;
 140                     n--;
 141                     continue;
 142                 }
 143
 144
 145                 n--;
 146             }
 147             return -1;
 148         },
 149         /**
 150          * lastSym - find the last token symbol
 151          * need to back check syntax..
 152          *
 153          * @arg {Array} tokens the array of tokens.
 154          * @arg {Number} offset where to start..
 155          * @return {Token} the token
 156          */
 157         lastSym : function(tokens, n) {
 158             for (var i = n-1; i >= 0; i--) {
 159                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 160             }
 161             return null;
 162         },
 163
 164
 165
 166         /**
 167             @returns {Boolean} Was the token found?
 168          */
 169         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 170             var found = "";
 171             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 172                 found += stream.next();
 173             }
 174
 175             if (found === "") {
 176                 return false;
 177             }
 178
 179             var name;
 180             if ((name = Lang.keyword(found))) {
 181                 if (found == 'return' && tokens.lastSym().data == ')') {
 182                     //Seed.print('@' + tokens.length);
 183                     var n = this.findPuncToken(tokens, ')');
 184                     //Seed.print(')@' + n);
 185                     n = this.findPuncToken(tokens, '(', n-1);
 186                     //Seed.print('(@' + n);
 187
 188                     var lt = this.lastSym(tokens, n);
 189                     print(JSON.stringify(lt));
 190                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 191                         if (!this.ignoreBadGrammer) {
 192                             throw {
 193                                 name : "ArgumentError",
 194                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 195                             }
 196                         }
 197                     }
 198
 199
 200
 201                 }
 202
 203                 tokens.push(new Token(found, "KEYW", name, this.line));
 204                 return true;
 205             }
 206             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 207                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 208                 return true;
 209             }
 210             var n = found.split('.');
 211             var p = false;
 212             var _this = this;
 213             n.forEach(function(nm) {
 214                 if (p) {
 215                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 216                 }
 217                 p=true;
 218                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 219             });
 220             return true;
 221
 222
 223         },
 224
 225         /**
 226             @returns {Boolean} Was the token found?
 227          */
 228         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 229             var found = "";
 230             var name;
 231             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 232                 found += stream.next();
 233             }
 234
 235
 236             if (found === "") {
 237                 return false;
 238             }
 239
 240             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 241                 //print("Error - comma found before " + found);
 242                 //print(JSON.stringify(tokens.lastSym(), null,4));
 243                 if (this.ignoreBadGrammer) {
 244                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 245                 } else {
 246
 247                     throw {
 248                         name : "ArgumentError",
 249                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 250                     }
 251                 }
 252             }
 253
 254             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 255             return true;
 256
 257         },
 258
 259         /**
 260             @returns {Boolean} Was the token found?
 261          */
 262         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 263             var found = "";
 264
 265             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 266                 found += stream.next();
 267             }
 268
 269             if (found === "") {
 270                 return false;
 271             }
 272             //print("WHITE = " + JSON.stringify(found));
 273             if (this.collapseWhite) found = " ";
 274             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 275             return true;
 276
 277         },
 278
 279         /**
 280             @returns {Boolean} Was the token found?
 281          */
 282         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 283             var found = "";
 284             var line = this.line;
 285             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 286                 this.line++;
 287                 found += stream.next();
 288             }
 289
 290             if (found === "") {
 291                 return false;
 292             }
 293             //this.line++;
 294             if (this.collapseWhite) {
 295                 found = "\n";
 296             }
 297              if (this.keepWhite) {
 298                 var last = tokens ? tokens.pop() : false;
 299                 if (last && last.name != "WHIT") {
 300                     tokens.push(last);
 301                 }
 302
 303                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 304             }
 305             return true;
 306         },
 307
 308         /**
 309             @returns {Boolean} Was the token found?
 310          */
 311         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 312             if (stream.look() == "/" && stream.look(1) == "*") {
 313                 var found = stream.next(2);
 314                 var c = '';
 315                 var line = this.line;
 316                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 317                     c = stream.next();
 318                     if (c == "\n") this.line++;
 319                     found += c;
 320                 }
 321
 322                 // to start doclet we allow /** or /*** but not /**/ or /****
 323                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 324                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 325                 return true;
 326             }
 327             return false;
 328         },
 329
 330         /**
 331             @returns {Boolean} Was the token found?
 332          */
 333         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 334             var found;
 335             if (
 336                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 337                 ||
 338                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 339             ) {
 340                 var line = this.line;
 341                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 342                     found += stream.next();
 343                 }
 344                 if (!stream.look().eof) {
 345                     found += stream.next();
 346                 }
 347                 if (this.keepComments) {
 348                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 349                 }
 350                 this.line++;
 351                 return true;
 352             }
 353             return false;
 354         },
 355
 356         /**
 357             @returns {Boolean} Was the token found?
 358          */
 359         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 360             if (stream.look() == "\"") {
 361                 // find terminator
 362                 var string = stream.next();
 363
 364                 while (!stream.look().eof) {
 365                     if (stream.look() == "\\") {
 366                         if (Lang.isNewline(stream.look(1))) {
 367                             do {
 368                                 stream.next();
 369                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 370                             string += "\\\n";
 371                         }
 372                         else {
 373                             string += stream.next(2);
 374                         }
 375                     }
 376                     else if (stream.look() == "\"") {
 377                         string += stream.next();
 378                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 379                         return true;
 380                     }
 381                     else {
 382                         string += stream.next();
 383                     }
 384                 }
 385             }
 386             return false; // error! unterminated string
 387         },
 388
 389         /**
 390             @returns {Boolean} Was the token found?
 391          */
 392         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 393             if (stream.look() == "'") {
 394                 // find terminator
 395                 var string = stream.next();
 396
 397                 while (!stream.look().eof) {
 398                     if (stream.look() == "\\") { // escape sequence
 399                         string += stream.next(2);
 400                     }
 401                     else if (stream.look() == "'") {
 402                         string += stream.next();
 403                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 404                         return true;
 405                     }
 406                     else {
 407                         string += stream.next();
 408                     }
 409                 }
 410             }
 411             return false; // error! unterminated string
 412         },
 413
 414         /**
 415             @returns {Boolean} Was the token found?
 416          */
 417         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 418             if (stream.look() === "0" && stream.look(1) == "x") {
 419                 return this.read_hex(stream, tokens);
 420             }
 421
 422             var found = "";
 423
 424             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 425                 found += stream.next();
 426             }
 427
 428             if (found === "") {
 429                 return false;
 430             }
 431             else {
 432                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 433                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 434                 return true;
 435             }
 436         },
 437         /*t:
 438             requires("../lib/JSDOC/TextStream.js");
 439             requires("../lib/JSDOC/Token.js");
 440             requires("../lib/JSDOC/Lang.js");
 441
 442             plan(3, "testing read_numb");
 443
 444             //// setup
 445             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 446             var tr = new TokenReader();
 447             var tokens = tr.tokenize(new TextStream(src));
 448
 449             var hexToken, octToken, decToken;
 450             for (var i = 0; i < tokens.length; i++) {
 451                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 452                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 453                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 454             }
 455             ////
 456
 457             is(decToken.data, "8.0", "decimal number is found in source.");
 458             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 459             is(octToken.data, "0777", "octal number is found in source.");
 460         */
 461
 462         /**
 463             @returns {Boolean} Was the token found?
 464          */
 465         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 466             var found = stream.next(2);
 467
 468             while (!stream.look().eof) {
 469                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 470                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 471                     return true;
 472                 }
 473                 else {
 474                     found += stream.next();
 475                 }
 476             }
 477             return false;
 478         },
 479
 480         /**
 481             @returns {Boolean} Was the token found?
 482          */
 483         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 484             var last;
 485             if (
 486                 stream.look() == "/"
 487                 &&
 488                 (
 489
 490                     (
 491                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 492                         ||
 493                         (
 494                                !last.is("NUMB")
 495                             && !last.is("NAME")
 496                             && !last.is("RIGHT_PAREN")
 497                             && !last.is("RIGHT_BRACKET")
 498                         )
 499                     )
 500                 )
 501             ) {
 502                 var regex = stream.next();
 503
 504                 while (!stream.look().eof) {
 505                     if (stream.look() == "\\") { // escape sequence
 506                         regex += stream.next(2);
 507                     }
 508                     else if (stream.look() == "/") {
 509                         regex += stream.next();
 510
 511                         while (/[gmi]/.test(stream.look())) {
 512                             regex += stream.next();
 513                         }
 514
 515                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 516                         return true;
 517                     }
 518                     else {
 519                         regex += stream.next();
 520                     }
 521                 }
 522                 // error: unterminated regex
 523             }
 524             return false;
 525         }
 526 });