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

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