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