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             var found = "";
 178             while (!stream.lookEOF() && Lang.isWordChar(stream.look())) {
 179                 found += stream.next();
 180             }
 181
 182             if (found === "") {
 183                 return false;
 184             }
 185
 186             var name;
 187             if ((name = Lang.keyword(found))) {
 188                 if (found == 'return' && tokens.lastSym().data == ')') {
 189                     //Seed.print('@' + tokens.length);
 190                     var n = this.findPuncToken(tokens, ')');
 191                     //Seed.print(')@' + n);
 192                     n = this.findPuncToken(tokens, '(', n-1);
 193                     //Seed.print('(@' + n);
 194
 195                     var lt = this.lastSym(tokens, n);
 196                     print(JSON.stringify(lt));
 197                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 198                         if (!this.ignoreBadGrammer) {
 199                             throw {
 200                                 name : "ArgumentError",
 201                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 202                             }
 203                         }
 204                     }
 205
 206
 207
 208                 }
 209
 210                 tokens.push(new Token(found, "KEYW", name, this.line));
 211                 return true;
 212             }
 213             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 214                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 215                 return true;
 216             }
 217             var n = found.split('.');
 218             var p = false;
 219             var _this = this;
 220             n.forEach(function(nm) {
 221                 if (p) {
 222                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 223                 }
 224                 p=true;
 225                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 226             });
 227             return true;
 228
 229
 230         }
 231
 232         /**
 233             @returns {Boolean} Was the token found?
 234          */
 235         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 236             var found = "";
 237             var name;
 238             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 239                 found += stream.next();
 240             }
 241
 242
 243             if (found === "") {
 244                 return false;
 245             }
 246
 247             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 248                 //print("Error - comma found before " + found);
 249                 //print(JSON.stringify(tokens.lastSym(), null,4));
 250                 if (this.ignoreBadGrammer) {
 251                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 252                 } else {
 253
 254                     throw {
 255                         name : "ArgumentError",
 256                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 257                     }
 258                 }
 259             }
 260
 261             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 262             return true;
 263
 264         },
 265
 266         /**
 267             @returns {Boolean} Was the token found?
 268          */
 269         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 270             var found = "";
 271
 272             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 273                 found += stream.next();
 274             }
 275
 276             if (found === "") {
 277                 return false;
 278             }
 279             //print("WHITE = " + JSON.stringify(found));
 280             if (this.collapseWhite) found = " ";
 281             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 282             return true;
 283
 284         },
 285
 286         /**
 287             @returns {Boolean} Was the token found?
 288          */
 289         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 290             var found = "";
 291             var line = this.line;
 292             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 293                 this.line++;
 294                 found += stream.next();
 295             }
 296
 297             if (found === "") {
 298                 return false;
 299             }
 300             //this.line++;
 301             if (this.collapseWhite) {
 302                 found = "\n";
 303             }
 304              if (this.keepWhite) {
 305                 var last = tokens ? tokens.pop() : false;
 306                 if (last && last.name != "WHIT") {
 307                     tokens.push(last);
 308                 }
 309
 310                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 311             }
 312             return true;
 313         },
 314
 315         /**
 316             @returns {Boolean} Was the token found?
 317          */
 318         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 319             if (stream.look() == "/" && stream.look(1) == "*") {
 320                 var found = stream.next(2);
 321                 var c = '';
 322                 var line = this.line;
 323                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 324                     c = stream.next();
 325                     if (c == "\n") this.line++;
 326                     found += c;
 327                 }
 328
 329                 // to start doclet we allow /** or /*** but not /**/ or /****
 330                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 331                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 332                 return true;
 333             }
 334             return false;
 335         },
 336
 337         /**
 338             @returns {Boolean} Was the token found?
 339          */
 340         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 341             var found;
 342             if (
 343                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 344                 ||
 345                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 346             ) {
 347                 var line = this.line;
 348                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 349                     found += stream.next();
 350                 }
 351                 if (!stream.look().eof) {
 352                     found += stream.next();
 353                 }
 354                 if (this.keepComments) {
 355                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 356                 }
 357                 this.line++;
 358                 return true;
 359             }
 360             return false;
 361         },
 362
 363         /**
 364             @returns {Boolean} Was the token found?
 365          */
 366         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 367             if (stream.look() == "\"") {
 368                 // find terminator
 369                 var string = stream.next();
 370
 371                 while (!stream.look().eof) {
 372                     if (stream.look() == "\\") {
 373                         if (Lang.isNewline(stream.look(1))) {
 374                             do {
 375                                 stream.next();
 376                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 377                             string += "\\\n";
 378                         }
 379                         else {
 380                             string += stream.next(2);
 381                         }
 382                     }
 383                     else if (stream.look() == "\"") {
 384                         string += stream.next();
 385                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 386                         return true;
 387                     }
 388                     else {
 389                         string += stream.next();
 390                     }
 391                 }
 392             }
 393             return false; // error! unterminated string
 394         },
 395
 396         /**
 397             @returns {Boolean} Was the token found?
 398          */
 399         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 400             if (stream.look() == "'") {
 401                 // find terminator
 402                 var string = stream.next();
 403
 404                 while (!stream.look().eof) {
 405                     if (stream.look() == "\\") { // escape sequence
 406                         string += stream.next(2);
 407                     }
 408                     else if (stream.look() == "'") {
 409                         string += stream.next();
 410                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 411                         return true;
 412                     }
 413                     else {
 414                         string += stream.next();
 415                     }
 416                 }
 417             }
 418             return false; // error! unterminated string
 419         },
 420
 421         /**
 422             @returns {Boolean} Was the token found?
 423          */
 424         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 425             if (stream.look() === "0" && stream.look(1) == "x") {
 426                 return this.read_hex(stream, tokens);
 427             }
 428
 429             var found = "";
 430
 431             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 432                 found += stream.next();
 433             }
 434
 435             if (found === "") {
 436                 return false;
 437             }
 438             else {
 439                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 440                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 441                 return true;
 442             }
 443         },
 444         /*t:
 445             requires("../lib/JSDOC/TextStream.js");
 446             requires("../lib/JSDOC/Token.js");
 447             requires("../lib/JSDOC/Lang.js");
 448
 449             plan(3, "testing read_numb");
 450
 451             //// setup
 452             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 453             var tr = new TokenReader();
 454             var tokens = tr.tokenize(new TextStream(src));
 455
 456             var hexToken, octToken, decToken;
 457             for (var i = 0; i < tokens.length; i++) {
 458                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 459                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 460                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 461             }
 462             ////
 463
 464             is(decToken.data, "8.0", "decimal number is found in source.");
 465             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 466             is(octToken.data, "0777", "octal number is found in source.");
 467         */
 468
 469         /**
 470             @returns {Boolean} Was the token found?
 471          */
 472         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 473             var found = stream.next(2);
 474
 475             while (!stream.look().eof) {
 476                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 477                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 478                     return true;
 479                 }
 480                 else {
 481                     found += stream.next();
 482                 }
 483             }
 484             return false;
 485         },
 486
 487         /**
 488             @returns {Boolean} Was the token found?
 489          */
 490         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 491             var last;
 492             if (
 493                 stream.look() == "/"
 494                 &&
 495                 (
 496
 497                     (
 498                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 499                         ||
 500                         (
 501                                !last.is("NUMB")
 502                             && !last.is("NAME")
 503                             && !last.is("RIGHT_PAREN")
 504                             && !last.is("RIGHT_BRACKET")
 505                         )
 506                     )
 507                 )
 508             ) {
 509                 var regex = stream.next();
 510
 511                 while (!stream.look().eof) {
 512                     if (stream.look() == "\\") { // escape sequence
 513                         regex += stream.next(2);
 514                     }
 515                     else if (stream.look() == "/") {
 516                         regex += stream.next();
 517
 518                         while (/[gmi]/.test(stream.look())) {
 519                             regex += stream.next();
 520                         }
 521
 522                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 523                         return true;
 524                     }
 525                     else {
 526                         regex += stream.next();
 527                     }
 528                 }
 529                 // error: unterminated regex
 530             }
 531             return false;
 532         }
 533 });