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 (true) {
  94
  95                 stream.look(0, out eof)
  96                 if (eof) {
  97                     break;
  98                 }
  99                 if (this.read_mlcomment(stream, tokens)) continue;
 100                 if (this.read_slcomment(stream, tokens)) continue;
 101                 if (this.read_dbquote(stream, tokens))   continue;
 102                 if (this.read_snquote(stream, tokens))   continue;
 103                 if (this.read_regx(stream, tokens))      continue;
 104                 if (this.read_numb(stream, tokens))      continue;
 105                 if (this.read_punc(stream, tokens))      continue;
 106                 if (this.read_newline(stream, tokens))   continue;
 107                 if (this.read_space(stream, tokens))     continue;
 108                 if (this.read_word(stream, tokens))      continue;
 109
 110                 // if execution reaches here then an error has happened
 111                 tokens.push(
 112                         new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line)
 113                 );
 114             }
 115
 116
 117
 118             return tokens;
 119         },
 120
 121         /**
 122          * findPuncToken - find the id of a token (previous to current)
 123          * need to back check syntax..
 124          *
 125          * @arg {Array} tokens the array of tokens.
 126          * @arg {String} token data (eg. '(')
 127          * @arg {Number} offset where to start reading from
 128          * @return {Number} position of token
 129          */
 130         public int findPuncToken(TokenArray tokens, string data, int n) {
 131             n = n || tokens.length -1;
 132             var stack = 0;
 133             while (n > -1) {
 134
 135                 if (!stack && tokens[n].data == data) {
 136                     return n;
 137                 }
 138
 139                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
 140                     stack++;
 141                     n--;
 142                     continue;
 143                 }
 144                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 145                     stack--;
 146                     n--;
 147                     continue;
 148                 }
 149
 150
 151                 n--;
 152             }
 153             return -1;
 154         },
 155         /**
 156          * lastSym - find the last token symbol
 157          * need to back check syntax..
 158          *
 159          * @arg {Array} tokens the array of tokens.
 160          * @arg {Number} offset where to start..
 161          * @return {Token} the token
 162          */
 163         lastSym : function(tokens, n) {
 164             for (var i = n-1; i >= 0; i--) {
 165                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 166             }
 167             return null;
 168         },
 169
 170
 171
 172         /**
 173             @returns {Boolean} Was the token found?
 174          */
 175         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 176             var found = "";
 177             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 178                 found += stream.next();
 179             }
 180
 181             if (found === "") {
 182                 return false;
 183             }
 184
 185             var name;
 186             if ((name = Lang.keyword(found))) {
 187                 if (found == 'return' && tokens.lastSym().data == ')') {
 188                     //Seed.print('@' + tokens.length);
 189                     var n = this.findPuncToken(tokens, ')');
 190                     //Seed.print(')@' + n);
 191                     n = this.findPuncToken(tokens, '(', n-1);
 192                     //Seed.print('(@' + n);
 193
 194                     var lt = this.lastSym(tokens, n);
 195                     print(JSON.stringify(lt));
 196                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 197                         if (!this.ignoreBadGrammer) {
 198                             throw {
 199                                 name : "ArgumentError",
 200                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 201                             }
 202                         }
 203                     }
 204
 205
 206
 207                 }
 208
 209                 tokens.push(new Token(found, "KEYW", name, this.line));
 210                 return true;
 211             }
 212             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 213                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 214                 return true;
 215             }
 216             var n = found.split('.');
 217             var p = false;
 218             var _this = this;
 219             n.forEach(function(nm) {
 220                 if (p) {
 221                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 222                 }
 223                 p=true;
 224                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 225             });
 226             return true;
 227
 228
 229         },
 230
 231         /**
 232             @returns {Boolean} Was the token found?
 233          */
 234         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 235             var found = "";
 236             var name;
 237             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 238                 found += stream.next();
 239             }
 240
 241
 242             if (found === "") {
 243                 return false;
 244             }
 245
 246             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 247                 //print("Error - comma found before " + found);
 248                 //print(JSON.stringify(tokens.lastSym(), null,4));
 249                 if (this.ignoreBadGrammer) {
 250                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 251                 } else {
 252
 253                     throw {
 254                         name : "ArgumentError",
 255                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 256                     }
 257                 }
 258             }
 259
 260             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 261             return true;
 262
 263         },
 264
 265         /**
 266             @returns {Boolean} Was the token found?
 267          */
 268         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 269             var found = "";
 270
 271             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 272                 found += stream.next();
 273             }
 274
 275             if (found === "") {
 276                 return false;
 277             }
 278             //print("WHITE = " + JSON.stringify(found));
 279             if (this.collapseWhite) found = " ";
 280             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 281             return true;
 282
 283         },
 284
 285         /**
 286             @returns {Boolean} Was the token found?
 287          */
 288         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 289             var found = "";
 290             var line = this.line;
 291             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 292                 this.line++;
 293                 found += stream.next();
 294             }
 295
 296             if (found === "") {
 297                 return false;
 298             }
 299             //this.line++;
 300             if (this.collapseWhite) {
 301                 found = "\n";
 302             }
 303              if (this.keepWhite) {
 304                 var last = tokens ? tokens.pop() : false;
 305                 if (last && last.name != "WHIT") {
 306                     tokens.push(last);
 307                 }
 308
 309                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 310             }
 311             return true;
 312         },
 313
 314         /**
 315             @returns {Boolean} Was the token found?
 316          */
 317         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 318             if (stream.look() == "/" && stream.look(1) == "*") {
 319                 var found = stream.next(2);
 320                 var c = '';
 321                 var line = this.line;
 322                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 323                     c = stream.next();
 324                     if (c == "\n") this.line++;
 325                     found += c;
 326                 }
 327
 328                 // to start doclet we allow /** or /*** but not /**/ or /****
 329                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 330                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 331                 return true;
 332             }
 333             return false;
 334         },
 335
 336         /**
 337             @returns {Boolean} Was the token found?
 338          */
 339         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 340             var found;
 341             if (
 342                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 343                 ||
 344                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 345             ) {
 346                 var line = this.line;
 347                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 348                     found += stream.next();
 349                 }
 350                 if (!stream.look().eof) {
 351                     found += stream.next();
 352                 }
 353                 if (this.keepComments) {
 354                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 355                 }
 356                 this.line++;
 357                 return true;
 358             }
 359             return false;
 360         },
 361
 362         /**
 363             @returns {Boolean} Was the token found?
 364          */
 365         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 366             if (stream.look() == "\"") {
 367                 // find terminator
 368                 var string = stream.next();
 369
 370                 while (!stream.look().eof) {
 371                     if (stream.look() == "\\") {
 372                         if (Lang.isNewline(stream.look(1))) {
 373                             do {
 374                                 stream.next();
 375                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 376                             string += "\\\n";
 377                         }
 378                         else {
 379                             string += stream.next(2);
 380                         }
 381                     }
 382                     else if (stream.look() == "\"") {
 383                         string += stream.next();
 384                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 385                         return true;
 386                     }
 387                     else {
 388                         string += stream.next();
 389                     }
 390                 }
 391             }
 392             return false; // error! unterminated string
 393         },
 394
 395         /**
 396             @returns {Boolean} Was the token found?
 397          */
 398         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 399             if (stream.look() == "'") {
 400                 // find terminator
 401                 var string = stream.next();
 402
 403                 while (!stream.look().eof) {
 404                     if (stream.look() == "\\") { // escape sequence
 405                         string += stream.next(2);
 406                     }
 407                     else if (stream.look() == "'") {
 408                         string += stream.next();
 409                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 410                         return true;
 411                     }
 412                     else {
 413                         string += stream.next();
 414                     }
 415                 }
 416             }
 417             return false; // error! unterminated string
 418         },
 419
 420         /**
 421             @returns {Boolean} Was the token found?
 422          */
 423         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 424             if (stream.look() === "0" && stream.look(1) == "x") {
 425                 return this.read_hex(stream, tokens);
 426             }
 427
 428             var found = "";
 429
 430             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 431                 found += stream.next();
 432             }
 433
 434             if (found === "") {
 435                 return false;
 436             }
 437             else {
 438                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 439                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 440                 return true;
 441             }
 442         },
 443         /*t:
 444             requires("../lib/JSDOC/TextStream.js");
 445             requires("../lib/JSDOC/Token.js");
 446             requires("../lib/JSDOC/Lang.js");
 447
 448             plan(3, "testing read_numb");
 449
 450             //// setup
 451             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 452             var tr = new TokenReader();
 453             var tokens = tr.tokenize(new TextStream(src));
 454
 455             var hexToken, octToken, decToken;
 456             for (var i = 0; i < tokens.length; i++) {
 457                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 458                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 459                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 460             }
 461             ////
 462
 463             is(decToken.data, "8.0", "decimal number is found in source.");
 464             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 465             is(octToken.data, "0777", "octal number is found in source.");
 466         */
 467
 468         /**
 469             @returns {Boolean} Was the token found?
 470          */
 471         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 472             var found = stream.next(2);
 473
 474             while (!stream.look().eof) {
 475                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 476                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 477                     return true;
 478                 }
 479                 else {
 480                     found += stream.next();
 481                 }
 482             }
 483             return false;
 484         },
 485
 486         /**
 487             @returns {Boolean} Was the token found?
 488          */
 489         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 490             var last;
 491             if (
 492                 stream.look() == "/"
 493                 &&
 494                 (
 495
 496                     (
 497                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 498                         ||
 499                         (
 500                                !last.is("NUMB")
 501                             && !last.is("NAME")
 502                             && !last.is("RIGHT_PAREN")
 503                             && !last.is("RIGHT_BRACKET")
 504                         )
 505                     )
 506                 )
 507             ) {
 508                 var regex = stream.next();
 509
 510                 while (!stream.look().eof) {
 511                     if (stream.look() == "\\") { // escape sequence
 512                         regex += stream.next(2);
 513                     }
 514                     else if (stream.look() == "/") {
 515                         regex += stream.next();
 516
 517                         while (/[gmi]/.test(stream.look())) {
 518                             regex += stream.next();
 519                         }
 520
 521                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 522                         return true;
 523                     }
 524                     else {
 525                         regex += stream.next();
 526                     }
 527                 }
 528                 // error: unterminated regex
 529             }
 530             return false;
 531         }
 532 });