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         collapseWhite : false, // only reduces white space...
  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                 var _this = this;
  96                 n.forEach(function(nm) {
  97                     if (p) {
  98                         tokens.push(new Token('.', "PUNC", "DOT", _this.line));
  99                     }
 100                     p=true;
 101                     tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 102                 });
 103                 return true;
 104
 105             }
 106         },
 107
 108         /**
 109             @returns {Boolean} Was the token found?
 110          */
 111         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 112             var found = "";
 113             var name;
 114             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 115                 found += stream.next();
 116             }
 117
 118             if (found === "") {
 119                 return false;
 120             }
 121             else {
 122                 tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 123                 return true;
 124             }
 125         },
 126
 127         /**
 128             @returns {Boolean} Was the token found?
 129          */
 130         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 131             var found = "";
 132
 133             while (!stream.look().eof && Lang.isSpace(stream.look())) {
 134                 found += stream.next();
 135             }
 136
 137             if (found === "") {
 138                 return false;
 139             }
 140             else {
 141                 if (this.collapseWhite) found = " ";
 142                 if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 143                 return true;
 144             }
 145         },
 146
 147         /**
 148             @returns {Boolean} Was the token found?
 149          */
 150         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 151             var found = "";
 152
 153             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 154                 this.line++;
 155                 found += stream.next();
 156             }
 157
 158             if (found === "") {
 159                 return false;
 160             }
 161             else {
 162                 if (this.collapseWhite) {
 163                     found = "\n";
 164                 }
 165                 if (this.keepWhite) {
 166                     var last = tokens.pop();
 167                     if (last.name != "WHIT") {
 168                         tokens.push(last);
 169                     }
 170
 171
 172                     tokens.push(new Token(found, "WHIT", "NEWLINE", this.line));
 173                 }
 174                 return true;
 175             }
 176         },
 177
 178         /**
 179             @returns {Boolean} Was the token found?
 180          */
 181         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 182             if (stream.look() == "/" && stream.look(1) == "*") {
 183                 var found = stream.next(2);
 184                 var c = '';
 185                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 186                     c = stream.next();
 187                     if (c == "\n") this.line++;
 188                     found += c;
 189                 }
 190
 191                 // to start doclet we allow /** or /*** but not /**/ or /****
 192                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 193                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", this.line));
 194                 return true;
 195             }
 196             return false;
 197         },
 198
 199         /**
 200             @returns {Boolean} Was the token found?
 201          */
 202         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 203             var found;
 204             if (
 205                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 206                 ||
 207                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 208             ) {
 209
 210                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 211                     found += stream.next();
 212                 }
 213
 214                 if (this.keepComments) {
 215                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", this.line));
 216                 }
 217                 this.line++;
 218                 return true;
 219             }
 220             return false;
 221         },
 222
 223         /**
 224             @returns {Boolean} Was the token found?
 225          */
 226         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 227             if (stream.look() == "\"") {
 228                 // find terminator
 229                 var string = stream.next();
 230
 231                 while (!stream.look().eof) {
 232                     if (stream.look() == "\\") {
 233                         if (Lang.isNewline(stream.look(1))) {
 234                             do {
 235                                 stream.next();
 236                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 237                             string += "\\\n";
 238                         }
 239                         else {
 240                             string += stream.next(2);
 241                         }
 242                     }
 243                     else if (stream.look() == "\"") {
 244                         string += stream.next();
 245                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 246                         return true;
 247                     }
 248                     else {
 249                         string += stream.next();
 250                     }
 251                 }
 252             }
 253             return false; // error! unterminated string
 254         },
 255
 256         /**
 257             @returns {Boolean} Was the token found?
 258          */
 259         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 260             if (stream.look() == "'") {
 261                 // find terminator
 262                 var string = stream.next();
 263
 264                 while (!stream.look().eof) {
 265                     if (stream.look() == "\\") { // escape sequence
 266                         string += stream.next(2);
 267                     }
 268                     else if (stream.look() == "'") {
 269                         string += stream.next();
 270                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 271                         return true;
 272                     }
 273                     else {
 274                         string += stream.next();
 275                     }
 276                 }
 277             }
 278             return false; // error! unterminated string
 279         },
 280
 281         /**
 282             @returns {Boolean} Was the token found?
 283          */
 284         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 285             if (stream.look() === "0" && stream.look(1) == "x") {
 286                 return this.read_hex(stream, tokens);
 287             }
 288
 289             var found = "";
 290
 291             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 292                 found += stream.next();
 293             }
 294
 295             if (found === "") {
 296                 return false;
 297             }
 298             else {
 299                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 300                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 301                 return true;
 302             }
 303         },
 304         /*t:
 305             requires("../lib/JSDOC/TextStream.js");
 306             requires("../lib/JSDOC/Token.js");
 307             requires("../lib/JSDOC/Lang.js");
 308
 309             plan(3, "testing read_numb");
 310
 311             //// setup
 312             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 313             var tr = new TokenReader();
 314             var tokens = tr.tokenize(new TextStream(src));
 315
 316             var hexToken, octToken, decToken;
 317             for (var i = 0; i < tokens.length; i++) {
 318                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 319                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 320                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 321             }
 322             ////
 323
 324             is(decToken.data, "8.0", "decimal number is found in source.");
 325             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 326             is(octToken.data, "0777", "octal number is found in source.");
 327         */
 328
 329         /**
 330             @returns {Boolean} Was the token found?
 331          */
 332         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 333             var found = stream.next(2);
 334
 335             while (!stream.look().eof) {
 336                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 337                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 338                     return true;
 339                 }
 340                 else {
 341                     found += stream.next();
 342                 }
 343             }
 344             return false;
 345         },
 346
 347         /**
 348             @returns {Boolean} Was the token found?
 349          */
 350         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 351             var last;
 352             if (
 353                 stream.look() == "/"
 354                 &&
 355                 (
 356
 357                     (
 358                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 359                         ||
 360                         (
 361                                !last.is("NUMB")
 362                             && !last.is("NAME")
 363                             && !last.is("RIGHT_PAREN")
 364                             && !last.is("RIGHT_BRACKET")
 365                         )
 366                     )
 367                 )
 368             ) {
 369                 var regex = stream.next();
 370
 371                 while (!stream.look().eof) {
 372                     if (stream.look() == "\\") { // escape sequence
 373                         regex += stream.next(2);
 374                     }
 375                     else if (stream.look() == "/") {
 376                         regex += stream.next();
 377
 378                         while (/[gmi]/.test(stream.look())) {
 379                             regex += stream.next();
 380                         }
 381
 382                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 383                         return true;
 384                     }
 385                     else {
 386                         regex += stream.next();
 387                     }
 388                 }
 389                 // error: unterminated regex
 390             }
 391             return false;
 392         }
 393 });