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