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