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             //this.line++;
 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                 tokens.push(new Token(found, "WHIT", "NEWLINE", this.line));
 172             }
 173             return true;
 174         },
 175
 176         /**
 177             @returns {Boolean} Was the token found?
 178          */
 179         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 180             if (stream.look() == "/" && stream.look(1) == "*") {
 181                 var found = stream.next(2);
 182                 var c = '';
 183                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 184                     c = stream.next();
 185                     if (c == "\n") this.line++;
 186                     found += c;
 187                 }
 188
 189                 // to start doclet we allow /** or /*** but not /**/ or /****
 190                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 191                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", this.line));
 192                 return true;
 193             }
 194             return false;
 195         },
 196
 197         /**
 198             @returns {Boolean} Was the token found?
 199          */
 200         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 201             var found;
 202             if (
 203                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 204                 ||
 205                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 206             ) {
 207
 208                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 209                     found += stream.next();
 210                 }
 211
 212                 if (this.keepComments) {
 213                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", this.line));
 214                 }
 215                 this.line++;
 216                 return true;
 217             }
 218             return false;
 219         },
 220
 221         /**
 222             @returns {Boolean} Was the token found?
 223          */
 224         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 225             if (stream.look() == "\"") {
 226                 // find terminator
 227                 var string = stream.next();
 228
 229                 while (!stream.look().eof) {
 230                     if (stream.look() == "\\") {
 231                         if (Lang.isNewline(stream.look(1))) {
 232                             do {
 233                                 stream.next();
 234                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 235                             string += "\\\n";
 236                         }
 237                         else {
 238                             string += stream.next(2);
 239                         }
 240                     }
 241                     else if (stream.look() == "\"") {
 242                         string += stream.next();
 243                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 244                         return true;
 245                     }
 246                     else {
 247                         string += stream.next();
 248                     }
 249                 }
 250             }
 251             return false; // error! unterminated string
 252         },
 253
 254         /**
 255             @returns {Boolean} Was the token found?
 256          */
 257         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 258             if (stream.look() == "'") {
 259                 // find terminator
 260                 var string = stream.next();
 261
 262                 while (!stream.look().eof) {
 263                     if (stream.look() == "\\") { // escape sequence
 264                         string += stream.next(2);
 265                     }
 266                     else if (stream.look() == "'") {
 267                         string += stream.next();
 268                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 269                         return true;
 270                     }
 271                     else {
 272                         string += stream.next();
 273                     }
 274                 }
 275             }
 276             return false; // error! unterminated string
 277         },
 278
 279         /**
 280             @returns {Boolean} Was the token found?
 281          */
 282         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 283             if (stream.look() === "0" && stream.look(1) == "x") {
 284                 return this.read_hex(stream, tokens);
 285             }
 286
 287             var found = "";
 288
 289             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 290                 found += stream.next();
 291             }
 292
 293             if (found === "") {
 294                 return false;
 295             }
 296             else {
 297                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 298                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 299                 return true;
 300             }
 301         },
 302         /*t:
 303             requires("../lib/JSDOC/TextStream.js");
 304             requires("../lib/JSDOC/Token.js");
 305             requires("../lib/JSDOC/Lang.js");
 306
 307             plan(3, "testing read_numb");
 308
 309             //// setup
 310             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 311             var tr = new TokenReader();
 312             var tokens = tr.tokenize(new TextStream(src));
 313
 314             var hexToken, octToken, decToken;
 315             for (var i = 0; i < tokens.length; i++) {
 316                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 317                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 318                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 319             }
 320             ////
 321
 322             is(decToken.data, "8.0", "decimal number is found in source.");
 323             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 324             is(octToken.data, "0777", "octal number is found in source.");
 325         */
 326
 327         /**
 328             @returns {Boolean} Was the token found?
 329          */
 330         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 331             var found = stream.next(2);
 332
 333             while (!stream.look().eof) {
 334                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 335                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 336                     return true;
 337                 }
 338                 else {
 339                     found += stream.next();
 340                 }
 341             }
 342             return false;
 343         },
 344
 345         /**
 346             @returns {Boolean} Was the token found?
 347          */
 348         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 349             var last;
 350             if (
 351                 stream.look() == "/"
 352                 &&
 353                 (
 354
 355                     (
 356                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 357                         ||
 358                         (
 359                                !last.is("NUMB")
 360                             && !last.is("NAME")
 361                             && !last.is("RIGHT_PAREN")
 362                             && !last.is("RIGHT_BRACKET")
 363                         )
 364                     )
 365                 )
 366             ) {
 367                 var regex = stream.next();
 368
 369                 while (!stream.look().eof) {
 370                     if (stream.look() == "\\") { // escape sequence
 371                         regex += stream.next(2);
 372                     }
 373                     else if (stream.look() == "/") {
 374                         regex += stream.next();
 375
 376                         while (/[gmi]/.test(stream.look())) {
 377                             regex += stream.next();
 378                         }
 379
 380                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 381                         return true;
 382                     }
 383                     else {
 384                         regex += stream.next();
 385                     }
 386                 }
 387                 // error: unterminated regex
 388             }
 389             return false;
 390         }
 391 });