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