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         public Gee.ArrayList<Token> tokens;
  18         public int length {
  19             get { return this.tokens.size }
  20         }
  21
  22         public TokenArray()
  23         {
  24             this.items = new Gee.ArrayList<Token>();
  25         }
  26
  27         public Token? last() {
  28             if (this.tokens > 0) {
  29                 return this.tokens[this.tokens.length-1];
  30             }
  31             return null;
  32         }
  33         public Token? lastSym () {
  34             for (var i = this.tokens.length-1; i >= 0; i--) {
  35                 if (!(this.tokens.get(i).is("WHIT") || this.tokens.get(i).is("COMM")))  {
  36                     return this.tokens.get(i);
  37                 }
  38             }
  39             return null;
  40         }
  41         public void push (Token t) {
  42             this.tokens.add(t);
  43         }
  44     }
  45
  46
  47     public class TokenReader : Object
  48     {
  49
  50
  51
  52         /*
  53          *
  54          * I wonder if this will accept the prop: value, prop2 :value construxtor if we do not define one...
  55          */
  56
  57         /** @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token **/
  58         public bool collapseWhite = false, // only reduces white space...
  59         /** @cfg {Boolean} keepDocs keep JSDOC comments **/
  60         public bool keepDocs = true,
  61         /** @cfg {Boolean} keepWhite keep White space **/
  62         public bool keepWhite = false,
  63         /** @cfg {Boolean} keepComments  keep all comments **/
  64         public bool keepComments = false,
  65         /** @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] ) **/
  66         public bool sepIdents = false,
  67         /** @cfg {String} filename name of file being parsed. **/
  68         public string filename = "";
  69         /** @config {Boolean} ignoreBadGrammer do not throw errors if we find stuff that might break compression **/
  70         public bool ignoreBadGrammer = false,
  71
  72
  73         int line = 0;
  74
  75         /**
  76          * tokenize a stream
  77          * @return {Array} of tokens
  78          *
  79          * ts = new TextStream(File.read(str));
  80          * tr = TokenReader({ keepComments : true, keepWhite : true });
  81          * tr.tokenize(ts)
  82          *
  83          */
  84         public TokenArray tokenize(TextStream stream)
  85         {
  86             this.line =1;
  87             var tokens = new TokenArray();
  88
  89             bool eof;
  90             while (true) {
  91
  92                 stream.look(0, out eof)
  93                 if (eof) {
  94                     break;
  95                 }
  96                 if (this.read_mlcomment(stream, tokens)) continue;
  97                 if (this.read_slcomment(stream, tokens)) continue;
  98                 if (this.read_dbquote(stream, tokens))   continue;
  99                 if (this.read_snquote(stream, tokens))   continue;
 100                 if (this.read_regx(stream, tokens))      continue;
 101                 if (this.read_numb(stream, tokens))      continue;
 102                 if (this.read_punc(stream, tokens))      continue;
 103                 if (this.read_newline(stream, tokens))   continue;
 104                 if (this.read_space(stream, tokens))     continue;
 105                 if (this.read_word(stream, tokens))      continue;
 106
 107                 // if execution reaches here then an error has happened
 108                 tokens.push(
 109                         new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line)
 110                 );
 111             }
 112
 113
 114
 115             return tokens;
 116         },
 117
 118         /**
 119          * findPuncToken - find the id of a token (previous to current)
 120          * need to back check syntax..
 121          *
 122          * @arg {Array} tokens the array of tokens.
 123          * @arg {String} token data (eg. '(')
 124          * @arg {Number} offset where to start reading from
 125          * @return {Number} position of token
 126          */
 127         public int findPuncToken(TokenArray tokens, string data, int n) {
 128             n = n || tokens.length -1;
 129             var stack = 0;
 130             while (n > -1) {
 131
 132                 if (!stack && tokens[n].data == data) {
 133                     return n;
 134                 }
 135
 136                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
 137                     stack++;
 138                     n--;
 139                     continue;
 140                 }
 141                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 142                     stack--;
 143                     n--;
 144                     continue;
 145                 }
 146
 147
 148                 n--;
 149             }
 150             return -1;
 151         },
 152         /**
 153          * lastSym - find the last token symbol
 154          * need to back check syntax..
 155          *
 156          * @arg {Array} tokens the array of tokens.
 157          * @arg {Number} offset where to start..
 158          * @return {Token} the token
 159          */
 160         lastSym : function(tokens, n) {
 161             for (var i = n-1; i >= 0; i--) {
 162                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 163             }
 164             return null;
 165         },
 166
 167
 168
 169         /**
 170             @returns {Boolean} Was the token found?
 171          */
 172         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 173             var found = "";
 174             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 175                 found += stream.next();
 176             }
 177
 178             if (found === "") {
 179                 return false;
 180             }
 181
 182             var name;
 183             if ((name = Lang.keyword(found))) {
 184                 if (found == 'return' && tokens.lastSym().data == ')') {
 185                     //Seed.print('@' + tokens.length);
 186                     var n = this.findPuncToken(tokens, ')');
 187                     //Seed.print(')@' + n);
 188                     n = this.findPuncToken(tokens, '(', n-1);
 189                     //Seed.print('(@' + n);
 190
 191                     var lt = this.lastSym(tokens, n);
 192                     print(JSON.stringify(lt));
 193                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 194                         if (!this.ignoreBadGrammer) {
 195                             throw {
 196                                 name : "ArgumentError",
 197                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 198                             }
 199                         }
 200                     }
 201
 202
 203
 204                 }
 205
 206                 tokens.push(new Token(found, "KEYW", name, this.line));
 207                 return true;
 208             }
 209             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 210                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 211                 return true;
 212             }
 213             var n = found.split('.');
 214             var p = false;
 215             var _this = this;
 216             n.forEach(function(nm) {
 217                 if (p) {
 218                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 219                 }
 220                 p=true;
 221                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 222             });
 223             return true;
 224
 225
 226         },
 227
 228         /**
 229             @returns {Boolean} Was the token found?
 230          */
 231         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 232             var found = "";
 233             var name;
 234             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 235                 found += stream.next();
 236             }
 237
 238
 239             if (found === "") {
 240                 return false;
 241             }
 242
 243             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 244                 //print("Error - comma found before " + found);
 245                 //print(JSON.stringify(tokens.lastSym(), null,4));
 246                 if (this.ignoreBadGrammer) {
 247                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 248                 } else {
 249
 250                     throw {
 251                         name : "ArgumentError",
 252                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 253                     }
 254                 }
 255             }
 256
 257             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 258             return true;
 259
 260         },
 261
 262         /**
 263             @returns {Boolean} Was the token found?
 264          */
 265         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 266             var found = "";
 267
 268             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 269                 found += stream.next();
 270             }
 271
 272             if (found === "") {
 273                 return false;
 274             }
 275             //print("WHITE = " + JSON.stringify(found));
 276             if (this.collapseWhite) found = " ";
 277             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 278             return true;
 279
 280         },
 281
 282         /**
 283             @returns {Boolean} Was the token found?
 284          */
 285         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 286             var found = "";
 287             var line = this.line;
 288             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 289                 this.line++;
 290                 found += stream.next();
 291             }
 292
 293             if (found === "") {
 294                 return false;
 295             }
 296             //this.line++;
 297             if (this.collapseWhite) {
 298                 found = "\n";
 299             }
 300              if (this.keepWhite) {
 301                 var last = tokens ? tokens.pop() : false;
 302                 if (last && last.name != "WHIT") {
 303                     tokens.push(last);
 304                 }
 305
 306                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 307             }
 308             return true;
 309         },
 310
 311         /**
 312             @returns {Boolean} Was the token found?
 313          */
 314         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 315             if (stream.look() == "/" && stream.look(1) == "*") {
 316                 var found = stream.next(2);
 317                 var c = '';
 318                 var line = this.line;
 319                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 320                     c = stream.next();
 321                     if (c == "\n") this.line++;
 322                     found += c;
 323                 }
 324
 325                 // to start doclet we allow /** or /*** but not /**/ or /****
 326                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 327                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 328                 return true;
 329             }
 330             return false;
 331         },
 332
 333         /**
 334             @returns {Boolean} Was the token found?
 335          */
 336         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 337             var found;
 338             if (
 339                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 340                 ||
 341                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 342             ) {
 343                 var line = this.line;
 344                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 345                     found += stream.next();
 346                 }
 347                 if (!stream.look().eof) {
 348                     found += stream.next();
 349                 }
 350                 if (this.keepComments) {
 351                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 352                 }
 353                 this.line++;
 354                 return true;
 355             }
 356             return false;
 357         },
 358
 359         /**
 360             @returns {Boolean} Was the token found?
 361          */
 362         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 363             if (stream.look() == "\"") {
 364                 // find terminator
 365                 var string = stream.next();
 366
 367                 while (!stream.look().eof) {
 368                     if (stream.look() == "\\") {
 369                         if (Lang.isNewline(stream.look(1))) {
 370                             do {
 371                                 stream.next();
 372                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 373                             string += "\\\n";
 374                         }
 375                         else {
 376                             string += stream.next(2);
 377                         }
 378                     }
 379                     else if (stream.look() == "\"") {
 380                         string += stream.next();
 381                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 382                         return true;
 383                     }
 384                     else {
 385                         string += stream.next();
 386                     }
 387                 }
 388             }
 389             return false; // error! unterminated string
 390         },
 391
 392         /**
 393             @returns {Boolean} Was the token found?
 394          */
 395         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 396             if (stream.look() == "'") {
 397                 // find terminator
 398                 var string = stream.next();
 399
 400                 while (!stream.look().eof) {
 401                     if (stream.look() == "\\") { // escape sequence
 402                         string += stream.next(2);
 403                     }
 404                     else if (stream.look() == "'") {
 405                         string += stream.next();
 406                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 407                         return true;
 408                     }
 409                     else {
 410                         string += stream.next();
 411                     }
 412                 }
 413             }
 414             return false; // error! unterminated string
 415         },
 416
 417         /**
 418             @returns {Boolean} Was the token found?
 419          */
 420         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 421             if (stream.look() === "0" && stream.look(1) == "x") {
 422                 return this.read_hex(stream, tokens);
 423             }
 424
 425             var found = "";
 426
 427             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 428                 found += stream.next();
 429             }
 430
 431             if (found === "") {
 432                 return false;
 433             }
 434             else {
 435                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 436                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 437                 return true;
 438             }
 439         },
 440         /*t:
 441             requires("../lib/JSDOC/TextStream.js");
 442             requires("../lib/JSDOC/Token.js");
 443             requires("../lib/JSDOC/Lang.js");
 444
 445             plan(3, "testing read_numb");
 446
 447             //// setup
 448             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 449             var tr = new TokenReader();
 450             var tokens = tr.tokenize(new TextStream(src));
 451
 452             var hexToken, octToken, decToken;
 453             for (var i = 0; i < tokens.length; i++) {
 454                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 455                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 456                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 457             }
 458             ////
 459
 460             is(decToken.data, "8.0", "decimal number is found in source.");
 461             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 462             is(octToken.data, "0777", "octal number is found in source.");
 463         */
 464
 465         /**
 466             @returns {Boolean} Was the token found?
 467          */
 468         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 469             var found = stream.next(2);
 470
 471             while (!stream.look().eof) {
 472                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 473                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 474                     return true;
 475                 }
 476                 else {
 477                     found += stream.next();
 478                 }
 479             }
 480             return false;
 481         },
 482
 483         /**
 484             @returns {Boolean} Was the token found?
 485          */
 486         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 487             var last;
 488             if (
 489                 stream.look() == "/"
 490                 &&
 491                 (
 492
 493                     (
 494                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 495                         ||
 496                         (
 497                                !last.is("NUMB")
 498                             && !last.is("NAME")
 499                             && !last.is("RIGHT_PAREN")
 500                             && !last.is("RIGHT_BRACKET")
 501                         )
 502                     )
 503                 )
 504             ) {
 505                 var regex = stream.next();
 506
 507                 while (!stream.look().eof) {
 508                     if (stream.look() == "\\") { // escape sequence
 509                         regex += stream.next(2);
 510                     }
 511                     else if (stream.look() == "/") {
 512                         regex += stream.next();
 513
 514                         while (/[gmi]/.test(stream.look())) {
 515                             regex += stream.next();
 516                         }
 517
 518                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 519                         return true;
 520                     }
 521                     else {
 522                         regex += stream.next();
 523                     }
 524                 }
 525                 // error: unterminated regex
 526             }
 527             return false;
 528         }
 529 });