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