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