git.roojs.org Git - gnome.introspection-doc-generator/blob - JSDOC/TokenReader.js

   1 //<script type="text/javascript">
   2
   3
   4 const XObject = imports.XObject.XObject;
   5 const console = imports.console.console;
   6
   7
   8 const Token   = imports.Token.Token;
   9 const Lang    = imports.Lang.Lang;
  10
  11 /**
  12         @class Search a {@link JSDOC.TextStream} for language tokens.
  13 */
  14 const TokenReader = XObject.define(
  15     function(o) {
  16
  17         XObject.extend(this, o || {});
  18
  19     },
  20     Object,
  21     {
  22         /** @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token **/
  23         collapseWhite : false, // only reduces white space...
  24         /** @cfg {Boolean} keepDocs keep JSDOC comments **/
  25         keepDocs : true,
  26         /** @cfg {Boolean} keepWhite keep White space **/
  27         keepWhite : false,
  28         /** @cfg {Boolean} keepComments  keep all comments **/
  29         keepComments : false,
  30         /** @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] ) **/
  31         sepIdents : false,
  32         /** @cfg {String} filename name of file being parsed. **/
  33         filename : '',
  34         /** @config {Boolean} ignoreBadGrammer do not throw errors if we find stuff that might break compression **/
  35         ignoreBadGrammer : false,
  36         /**
  37          * tokenize a stream
  38          * @return {Array} of tokens
  39          *
  40          * ts = new TextStream(File.read(str));
  41          * tr = TokenReader({ keepComments : true, keepWhite : true });
  42          * tr.tokenize(ts)
  43          *
  44          */
  45         tokenize : function(/**JSDOC.TextStream*/stream) {
  46             this.line =1;
  47             var tokens = [];
  48             /**@ignore*/
  49             tokens.last    = function() { return tokens[tokens.length-1]; }
  50             /**@ignore*/
  51             tokens.lastSym = function() {
  52                 for (var i = tokens.length-1; i >= 0; i--) {
  53                     if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
  54                 }
  55                 return true;
  56             }
  57
  58             while (!stream.look().eof) {
  59                 print_r(tokens);
  60                 if (this.read_mlcomment(stream, tokens)) continue;
  61                 if (this.read_slcomment(stream, tokens)) continue;
  62                 if (this.read_dbquote(stream, tokens))   continue;
  63                 if (this.read_snquote(stream, tokens))   continue;
  64                 if (this.read_regx(stream, tokens))      continue;
  65                 if (this.read_numb(stream, tokens))      continue;
  66                 if (this.read_punc(stream, tokens))      continue;
  67                 if (this.read_newline(stream, tokens))   continue;
  68                 if (this.read_space(stream, tokens))     continue;
  69                 if (this.read_word(stream, tokens))      continue;
  70
  71                 // if execution reaches here then an error has happened
  72                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
  73             }
  74
  75
  76
  77             return tokens;
  78         },
  79
  80         /**
  81          * findPuncToken - find the id of a token (previous to current)
  82          * need to back check syntax..
  83          *
  84          * @arg {Array} tokens the array of tokens.
  85          * @arg {String} token data (eg. '(')
  86          * @arg {Number} offset where to start reading from
  87          * @return {Number} position of token
  88          */
  89         findPuncToken : function(tokens, data, n) {
  90             n = n || tokens.length -1;
  91             var stack = 0;
  92             while (n > -1) {
  93
  94                 if (!stack && tokens[n].data == data) {
  95                     return n;
  96                 }
  97
  98                 if (tokens[n].data  == ')' || tokens[n].data  == '}') {
  99                     stack++;
 100                     n--;
 101                     continue;
 102                 }
 103                 if (stack && (tokens[n].data  == '{' || tokens[n].data  == '(')) {
 104                     stack--;
 105                     n--;
 106                     continue;
 107                 }
 108
 109
 110                 n--;
 111             }
 112             return -1;
 113         },
 114         /**
 115          * lastSym - find the last token symbol
 116          * need to back check syntax..
 117          *
 118          * @arg {Array} tokens the array of tokens.
 119          * @arg {Number} offset where to start..
 120          * @return {Token} the token
 121          */
 122         lastSym : function(tokens, n) {
 123             for (var i = n-1; i >= 0; i--) {
 124                 if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
 125             }
 126             return null;
 127         },
 128
 129
 130
 131         /**
 132             @returns {Boolean} Was the token found?
 133          */
 134         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
 135             var found = "";
 136             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
 137                 found += stream.next();
 138             }
 139
 140             if (found === "") {
 141                 return false;
 142             }
 143
 144             var name;
 145             if ((name = Lang.keyword(found))) {
 146                 if (found == 'return' && tokens.lastSym().data == ')') {
 147                     //Seed.print('@' + tokens.length);
 148                     var n = this.findPuncToken(tokens, ')');
 149                     //Seed.print(')@' + n);
 150                     n = this.findPuncToken(tokens, '(', n-1);
 151                     //Seed.print('(@' + n);
 152
 153                     var lt = this.lastSym(tokens, n);
 154                     print(JSON.stringify(lt));
 155                     if (lt.type != 'KEYW' || ['IF', 'WHILE'].indexOf(lt.name) < -1) {
 156                         if (!this.ignoreBadGrammer) {
 157                             throw {
 158                                 name : "ArgumentError",
 159                                 message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
 160                             }
 161                         }
 162                     }
 163
 164
 165
 166                 }
 167
 168                 tokens.push(new Token(found, "KEYW", name, this.line));
 169                 return true;
 170             }
 171             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 172                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 173                 return true;
 174             }
 175             var n = found.split('.');
 176             var p = false;
 177             var _this = this;
 178             n.forEach(function(nm) {
 179                 if (p) {
 180                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 181                 }
 182                 p=true;
 183                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 184             });
 185             return true;
 186
 187
 188         },
 189
 190         /**
 191             @returns {Boolean} Was the token found?
 192          */
 193         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 194             var found = "";
 195             var name;
 196             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 197                 found += stream.next();
 198             }
 199
 200
 201             if (found === "") {
 202                 return false;
 203             }
 204
 205             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 206                 //print("Error - comma found before " + found);
 207                 //print(JSON.stringify(tokens.lastSym(), null,4));
 208                 if (this.ignoreBadGrammer) {
 209                     print("\n" + this.filename + ':' + this.line + " Error - comma found before " + found);
 210                 } else {
 211
 212                     throw {
 213                         name : "ArgumentError",
 214                         message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 215                     }
 216                 }
 217             }
 218
 219             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 220             return true;
 221
 222         },
 223
 224         /**
 225             @returns {Boolean} Was the token found?
 226          */
 227         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 228             var found = "";
 229
 230             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 231                 found += stream.next();
 232             }
 233
 234             if (found === "") {
 235                 return false;
 236             }
 237             //print("WHITE = " + JSON.stringify(found));
 238             if (this.collapseWhite) found = " ";
 239             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 240             return true;
 241
 242         },
 243
 244         /**
 245             @returns {Boolean} Was the token found?
 246          */
 247         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 248             var found = "";
 249             var line = this.line;
 250             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 251                 this.line++;
 252                 found += stream.next();
 253             }
 254
 255             if (found === "") {
 256                 return false;
 257             }
 258             //this.line++;
 259             if (this.collapseWhite) {
 260                 found = "\n";
 261             }
 262             if (this.keepWhite) {
 263                 var last = tokens.pop();
 264                 if (last && last.name != "WHIT") {
 265                     tokens.push(last);
 266                 }
 267
 268                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 269             }
 270             return true;
 271         },
 272
 273         /**
 274             @returns {Boolean} Was the token found?
 275          */
 276         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 277             if (stream.look() == "/" && stream.look(1) == "*") {
 278                 var found = stream.next(2);
 279                 var c = '';
 280                 var line = this.line;
 281                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 282                     c = stream.next();
 283                     if (c == "\n") this.line++;
 284                     found += c;
 285                 }
 286
 287                 // to start doclet we allow /** or /*** but not /**/ or /****
 288                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 289                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 290                 return true;
 291             }
 292             return false;
 293         },
 294
 295         /**
 296             @returns {Boolean} Was the token found?
 297          */
 298         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 299             var found;
 300             if (
 301                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 302                 ||
 303                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 304             ) {
 305                 var line = this.line;
 306                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 307                     found += stream.next();
 308                 }
 309                 if (!stream.look().eof) {
 310                     found += stream.next();
 311                 }
 312                 if (this.keepComments) {
 313                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 314                 }
 315                 this.line++;
 316                 return true;
 317             }
 318             return false;
 319         },
 320
 321         /**
 322             @returns {Boolean} Was the token found?
 323          */
 324         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 325             if (stream.look() == "\"") {
 326                 // find terminator
 327                 var string = stream.next();
 328
 329                 while (!stream.look().eof) {
 330                     if (stream.look() == "\\") {
 331                         if (Lang.isNewline(stream.look(1))) {
 332                             do {
 333                                 stream.next();
 334                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 335                             string += "\\\n";
 336                         }
 337                         else {
 338                             string += stream.next(2);
 339                         }
 340                     }
 341                     else if (stream.look() == "\"") {
 342                         string += stream.next();
 343                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 344                         return true;
 345                     }
 346                     else {
 347                         string += stream.next();
 348                     }
 349                 }
 350             }
 351             return false; // error! unterminated string
 352         },
 353
 354         /**
 355             @returns {Boolean} Was the token found?
 356          */
 357         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 358             if (stream.look() == "'") {
 359                 // find terminator
 360                 var string = stream.next();
 361
 362                 while (!stream.look().eof) {
 363                     if (stream.look() == "\\") { // escape sequence
 364                         string += stream.next(2);
 365                     }
 366                     else if (stream.look() == "'") {
 367                         string += stream.next();
 368                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 369                         return true;
 370                     }
 371                     else {
 372                         string += stream.next();
 373                     }
 374                 }
 375             }
 376             return false; // error! unterminated string
 377         },
 378
 379         /**
 380             @returns {Boolean} Was the token found?
 381          */
 382         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 383             if (stream.look() === "0" && stream.look(1) == "x") {
 384                 return this.read_hex(stream, tokens);
 385             }
 386
 387             var found = "";
 388
 389             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 390                 found += stream.next();
 391             }
 392
 393             if (found === "") {
 394                 return false;
 395             }
 396             else {
 397                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 398                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 399                 return true;
 400             }
 401         },
 402         /*t:
 403             requires("../lib/JSDOC/TextStream.js");
 404             requires("../lib/JSDOC/Token.js");
 405             requires("../lib/JSDOC/Lang.js");
 406
 407             plan(3, "testing read_numb");
 408
 409             //// setup
 410             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 411             var tr = new TokenReader();
 412             var tokens = tr.tokenize(new TextStream(src));
 413
 414             var hexToken, octToken, decToken;
 415             for (var i = 0; i < tokens.length; i++) {
 416                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 417                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 418                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 419             }
 420             ////
 421
 422             is(decToken.data, "8.0", "decimal number is found in source.");
 423             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 424             is(octToken.data, "0777", "octal number is found in source.");
 425         */
 426
 427         /**
 428             @returns {Boolean} Was the token found?
 429          */
 430         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 431             var found = stream.next(2);
 432
 433             while (!stream.look().eof) {
 434                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 435                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 436                     return true;
 437                 }
 438                 else {
 439                     found += stream.next();
 440                 }
 441             }
 442             return false;
 443         },
 444
 445         /**
 446             @returns {Boolean} Was the token found?
 447          */
 448         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 449             var last;
 450             if (
 451                 stream.look() == "/"
 452                 &&
 453                 (
 454
 455                     (
 456                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 457                         ||
 458                         (
 459                                !last.is("NUMB")
 460                             && !last.is("NAME")
 461                             && !last.is("RIGHT_PAREN")
 462                             && !last.is("RIGHT_BRACKET")
 463                         )
 464                     )
 465                 )
 466             ) {
 467                 var regex = stream.next();
 468
 469                 while (!stream.look().eof) {
 470                     if (stream.look() == "\\") { // escape sequence
 471                         regex += stream.next(2);
 472                     }
 473                     else if (stream.look() == "/") {
 474                         regex += stream.next();
 475
 476                         while (/[gmi]/.test(stream.look())) {
 477                             regex += stream.next();
 478                         }
 479
 480                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 481                         return true;
 482                     }
 483                     else {
 484                         regex += stream.next();
 485                     }
 486                 }
 487                 // error: unterminated regex
 488             }
 489             return false;
 490         }
 491 });