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         public Token get(int i) {
  45             return this.tokens.get(i);
  46         }
  47     }
  48
  49
  50     public class TokenReader : Object
  51     {
  52
  53
  54
  55         /*
  56          *
  57          * I wonder if this will accept the prop: value, prop2 :value construxtor if we do not define one...
  58          */
  59
  60         /** @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token **/
  61         public bool collapseWhite = false, // only reduces white space...
  62         /** @cfg {Boolean} keepDocs keep JSDOC comments **/
  63         public bool keepDocs = true,
  64         /** @cfg {Boolean} keepWhite keep White space **/
  65         public bool keepWhite = false,
  66         /** @cfg {Boolean} keepComments  keep all comments **/
  67         public bool keepComments = false,
  68         /** @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] ) **/
  69         public bool sepIdents = false,
  70         /** @cfg {String} filename name of file being parsed. **/
  71         public string filename = "";
  72         /** @config {Boolean} ignoreBadGrammer do not throw errors if we find stuff that might break compression **/
  73         public bool ignoreBadGrammer = false,
  74
  75
  76         int line = 0;
  77
  78         /**
  79          * tokenize a stream
  80          * @return {Array} of tokens
  81          *
  82          * ts = new TextStream(File.read(str));
  83          * tr = TokenReader({ keepComments : true, keepWhite : true });
  84          * tr.tokenize(ts)
  85          *
  86          */
  87         public TokenArray tokenize(TextStream stream)
  88         {
  89             this.line =1;
  90             var tokens = new TokenArray();
  91
  92             bool eof;
  93             while (!stream.lookEOF()) {
  94
  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         {
 129             n = n || tokens.length -1;
 130             var stack = 0;
 131             while (n > -1) {
 132
 133                 if (!stack && tokens.get(n).data == data) {
 134                     return n;
 135                 }
 136
 137                 if (tokens.get(n).data  == ')' || tokens.get(n).data  == '}') {
 138                     stack++;
 139                     n--;
 140                     continue;
 141                 }
 142                 if (stack && (tokens.get(n).data  == '{' || tokens.get(n).data  == '(')) {
 143                     stack--;
 144                     n--;
 145                     continue;
 146                 }
 147
 148
 149                 n--;
 150             }
 151             return -1;
 152         }
 153         /**
 154          * lastSym - find the last token symbol
 155          * need to back check syntax..
 156          *
 157          * @arg {Array} tokens the array of tokens.
 158          * @arg {Number} offset where to start..
 159          * @return {Token} the token
 160          */
 161         public Token lastSym(TokenArray tokens, int n)
 162         {
 163             for (var i = n-1; i >= 0; i--) {
 164                 if (!(tokens.get(i).is("WHIT") || tokens.get(i).is("COMM"))) {
 165                     return tokens.get(i);
 166                 }
 167             }
 168             return null;
 169         }
 170
 171
 172
 173         /**
 174             @returns {Boolean} Was the token found?
 175          */
 176         public bool read_word (TokenStream stream, TokenArray tokens)
 177         {
 178             string found = "";
 179             while (!stream.lookEOF() && Lang.isWordChar(stream.look())) {
 180                 found += stream.next();
 181             }
 182
 183             if (found == "") {
 184                 return false;
 185             }
 186
 187             var name = Lang.keyword(found);
 188             if (name != null) {
 189
 190                 // look for "()return" ?? why ???
 191
 192                 if (found == "return" && tokens.lastSym().data == ")") {
 193                     //Seed.print('@' + tokens.length);
 194                     var n = this.findPuncToken(tokens, ")");
 195                     //Seed.print(')@' + n);
 196                     n = this.findPuncToken(tokens, "(", n-1);
 197                     //Seed.print('(@' + n);
 198
 199                     var lt = this.lastSym(tokens, n);
 200
 201                     //print(JSON.stringify(lt));
 202                     if (lt.type != "KEYW" || ["IF", 'WHILE'].indexOf(lt.name) < -1) {
 203                         if (!this.ignoreBadGrammer) {
 204                             throw {
 205                                 name : "ArgumentError",
 206                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 207                             }
 208                         }
 209                     }
 210
 211
 212
 213                 }
 214
 215                 tokens.push(new Token(found, "KEYW", name, this.line));
 216                 return true;
 217             }
 218             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 219                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 220                 return true;
 221             }
 222             var n = found.split('.');
 223             var p = false;
 224             var _this = this;
 225             n.forEach(function(nm) {
 226                 if (p) {
 227                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 228                 }
 229                 p=true;
 230                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 231             });
 232             return true;
 233
 234
 235         }
 236
 237         /**
 238             @returns {Boolean} Was the token found?
 239          */
 240         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 241             var found = "";
 242             var name;
 243             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 244                 found += stream.next();
 245             }
 246
 247
 248             if (found === "") {
 249                 return false;
 250             }
 251
 252             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 253                 //print("Error - comma found before " + found);
 254                 //print(JSON.stringify(tokens.lastSym(), null,4));
 255                 if (this.ignoreBadGrammer) {
 256                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 257                 } else {
 258
 259                     throw {
 260                         name : "ArgumentError",
 261                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 262                     }
 263                 }
 264             }
 265
 266             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 267             return true;
 268
 269         },
 270
 271         /**
 272             @returns {Boolean} Was the token found?
 273          */
 274         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 275             var found = "";
 276
 277             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 278                 found += stream.next();
 279             }
 280
 281             if (found === "") {
 282                 return false;
 283             }
 284             //print("WHITE = " + JSON.stringify(found));
 285             if (this.collapseWhite) found = " ";
 286             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 287             return true;
 288
 289         },
 290
 291         /**
 292             @returns {Boolean} Was the token found?
 293          */
 294         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 295             var found = "";
 296             var line = this.line;
 297             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 298                 this.line++;
 299                 found += stream.next();
 300             }
 301
 302             if (found === "") {
 303                 return false;
 304             }
 305             //this.line++;
 306             if (this.collapseWhite) {
 307                 found = "\n";
 308             }
 309              if (this.keepWhite) {
 310                 var last = tokens ? tokens.pop() : false;
 311                 if (last && last.name != "WHIT") {
 312                     tokens.push(last);
 313                 }
 314
 315                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 316             }
 317             return true;
 318         },
 319
 320         /**
 321             @returns {Boolean} Was the token found?
 322          */
 323         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 324             if (stream.look() == "/" && stream.look(1) == "*") {
 325                 var found = stream.next(2);
 326                 var c = '';
 327                 var line = this.line;
 328                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 329                     c = stream.next();
 330                     if (c == "\n") this.line++;
 331                     found += c;
 332                 }
 333
 334                 // to start doclet we allow /** or /*** but not /**/ or /****
 335                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 336                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 337                 return true;
 338             }
 339             return false;
 340         },
 341
 342         /**
 343             @returns {Boolean} Was the token found?
 344          */
 345         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 346             var found;
 347             if (
 348                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 349                 ||
 350                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 351             ) {
 352                 var line = this.line;
 353                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 354                     found += stream.next();
 355                 }
 356                 if (!stream.look().eof) {
 357                     found += stream.next();
 358                 }
 359                 if (this.keepComments) {
 360                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 361                 }
 362                 this.line++;
 363                 return true;
 364             }
 365             return false;
 366         },
 367
 368         /**
 369             @returns {Boolean} Was the token found?
 370          */
 371         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 372             if (stream.look() == "\"") {
 373                 // find terminator
 374                 var string = stream.next();
 375
 376                 while (!stream.look().eof) {
 377                     if (stream.look() == "\\") {
 378                         if (Lang.isNewline(stream.look(1))) {
 379                             do {
 380                                 stream.next();
 381                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 382                             string += "\\\n";
 383                         }
 384                         else {
 385                             string += stream.next(2);
 386                         }
 387                     }
 388                     else if (stream.look() == "\"") {
 389                         string += stream.next();
 390                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 391                         return true;
 392                     }
 393                     else {
 394                         string += stream.next();
 395                     }
 396                 }
 397             }
 398             return false; // error! unterminated string
 399         },
 400
 401         /**
 402             @returns {Boolean} Was the token found?
 403          */
 404         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 405             if (stream.look() == "'") {
 406                 // find terminator
 407                 var string = stream.next();
 408
 409                 while (!stream.look().eof) {
 410                     if (stream.look() == "\\") { // escape sequence
 411                         string += stream.next(2);
 412                     }
 413                     else if (stream.look() == "'") {
 414                         string += stream.next();
 415                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 416                         return true;
 417                     }
 418                     else {
 419                         string += stream.next();
 420                     }
 421                 }
 422             }
 423             return false; // error! unterminated string
 424         },
 425
 426         /**
 427             @returns {Boolean} Was the token found?
 428          */
 429         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 430             if (stream.look() === "0" && stream.look(1) == "x") {
 431                 return this.read_hex(stream, tokens);
 432             }
 433
 434             var found = "";
 435
 436             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 437                 found += stream.next();
 438             }
 439
 440             if (found === "") {
 441                 return false;
 442             }
 443             else {
 444                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 445                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 446                 return true;
 447             }
 448         },
 449         /*t:
 450             requires("../lib/JSDOC/TextStream.js");
 451             requires("../lib/JSDOC/Token.js");
 452             requires("../lib/JSDOC/Lang.js");
 453
 454             plan(3, "testing read_numb");
 455
 456             //// setup
 457             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 458             var tr = new TokenReader();
 459             var tokens = tr.tokenize(new TextStream(src));
 460
 461             var hexToken, octToken, decToken;
 462             for (var i = 0; i < tokens.length; i++) {
 463                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 464                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 465                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 466             }
 467             ////
 468
 469             is(decToken.data, "8.0", "decimal number is found in source.");
 470             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 471             is(octToken.data, "0777", "octal number is found in source.");
 472         */
 473
 474         /**
 475             @returns {Boolean} Was the token found?
 476          */
 477         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 478             var found = stream.next(2);
 479
 480             while (!stream.look().eof) {
 481                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 482                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 483                     return true;
 484                 }
 485                 else {
 486                     found += stream.next();
 487                 }
 488             }
 489             return false;
 490         },
 491
 492         /**
 493             @returns {Boolean} Was the token found?
 494          */
 495         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 496             var last;
 497             if (
 498                 stream.look() == "/"
 499                 &&
 500                 (
 501
 502                     (
 503                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 504                         ||
 505                         (
 506                                !last.is("NUMB")
 507                             && !last.is("NAME")
 508                             && !last.is("RIGHT_PAREN")
 509                             && !last.is("RIGHT_BRACKET")
 510                         )
 511                     )
 512                 )
 513             ) {
 514                 var regex = stream.next();
 515
 516                 while (!stream.look().eof) {
 517                     if (stream.look() == "\\") { // escape sequence
 518                         regex += stream.next(2);
 519                     }
 520                     else if (stream.look() == "/") {
 521                         regex += stream.next();
 522
 523                         while (/[gmi]/.test(stream.look())) {
 524                             regex += stream.next();
 525                         }
 526
 527                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 528                         return true;
 529                     }
 530                     else {
 531                         regex += stream.next();
 532                     }
 533                 }
 534                 // error: unterminated regex
 535             }
 536             return false;
 537         }
 538 });