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

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