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