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();
  33          * tr = TokenReader({ keepComments : true, keepWhite : true });
  34          * tr.tokenize(ts)
  35          *
  36          */
  37             @type {JSDOC.Token[]}
  38
  39
  40          */
  41
  42
  43         tokenize : function(/**JSDOC.TextStream*/stream) {
  44             this.line =1;
  45             var tokens = [];
  46             /**@ignore*/ tokens.last    = function() { return tokens[tokens.length-1]; }
  47             /**@ignore*/ tokens.lastSym = function() {
  48                 for (var i = tokens.length-1; i >= 0; i--) {
  49                     if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
  50                 }
  51             }
  52
  53             while (!stream.look().eof) {
  54                 if (this.read_mlcomment(stream, tokens)) continue;
  55                 if (this.read_slcomment(stream, tokens)) continue;
  56                 if (this.read_dbquote(stream, tokens))   continue;
  57                 if (this.read_snquote(stream, tokens))   continue;
  58                 if (this.read_regx(stream, tokens))      continue;
  59                 if (this.read_numb(stream, tokens))      continue;
  60                 if (this.read_punc(stream, tokens))      continue;
  61                 if (this.read_newline(stream, tokens))   continue;
  62                 if (this.read_space(stream, tokens))     continue;
  63                 if (this.read_word(stream, tokens))      continue;
  64
  65                 // if execution reaches here then an error has happened
  66                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
  67             }
  68
  69
  70
  71             return tokens;
  72         },
  73
  74         /**
  75             @returns {Boolean} Was the token found?
  76          */
  77         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
  78             var found = "";
  79             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
  80                 found += stream.next();
  81             }
  82
  83             if (found === "") {
  84                 return false;
  85             }
  86             else {
  87                 var name;
  88                 if ((name = Lang.keyword(found))) {
  89                     tokens.push(new Token(found, "KEYW", name, this.line));
  90                     return true;
  91                 }
  92                 if (!this.sepIdents || found.indexOf('.') < 0 ) {
  93                     tokens.push(new Token(found, "NAME", "NAME", this.line));
  94                     return true;
  95                 }
  96                 var n = found.split('.');
  97                 var p = false;
  98                 n.forEach(function(nm) {
  99                     if (p) {
 100                         tokens.push(new Token('.', "PUNC", "DOT", this.line));
 101                     }
 102                     p=true;
 103                     tokens.push(new Token(nm, "NAME", "NAME", this.line));
 104                 });
 105                 return true;
 106
 107             }
 108         },
 109
 110         /**
 111             @returns {Boolean} Was the token found?
 112          */
 113         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 114             var found = "";
 115             var name;
 116             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 117                 found += stream.next();
 118             }
 119
 120             if (found === "") {
 121                 return false;
 122             }
 123             else {
 124                 tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 125                 return true;
 126             }
 127         },
 128
 129         /**
 130             @returns {Boolean} Was the token found?
 131          */
 132         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 133             var found = "";
 134
 135             while (!stream.look().eof && Lang.isSpace(stream.look())) {
 136                 found += stream.next();
 137             }
 138
 139             if (found === "") {
 140                 return false;
 141             }
 142             else {
 143                 if (this.collapseWhite) found = " ";
 144                 if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 145                 return true;
 146             }
 147         },
 148
 149         /**
 150             @returns {Boolean} Was the token found?
 151          */
 152         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 153             var found = "";
 154
 155             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 156                 this.line++;
 157                 found += stream.next();
 158             }
 159
 160             if (found === "") {
 161                 return false;
 162             }
 163             else {
 164                 if (this.collapseWhite) found = "\n";
 165                 if (this.keepWhite) tokens.push(new Token(found, "WHIT", "NEWLINE", this.line));
 166                 return true;
 167             }
 168         },
 169
 170         /**
 171             @returns {Boolean} Was the token found?
 172          */
 173         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 174             if (stream.look() == "/" && stream.look(1) == "*") {
 175                 var found = stream.next(2);
 176                 var c = '';
 177                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 178                     c = stream.next();
 179                     if (c == "\n") this.line++;
 180                     found += c;
 181                 }
 182
 183                 // to start doclet we allow /** or /*** but not /**/ or /****
 184                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 185                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", this.line));
 186                 return true;
 187             }
 188             return false;
 189         },
 190
 191         /**
 192             @returns {Boolean} Was the token found?
 193          */
 194         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 195             var found;
 196             if (
 197                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 198                 ||
 199                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 200             ) {
 201
 202                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 203                     found += stream.next();
 204                 }
 205
 206                 if (this.keepComments) {
 207                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", this.line));
 208                 }
 209                 this.line++;
 210                 return true;
 211             }
 212             return false;
 213         },
 214
 215         /**
 216             @returns {Boolean} Was the token found?
 217          */
 218         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 219             if (stream.look() == "\"") {
 220                 // find terminator
 221                 var string = stream.next();
 222
 223                 while (!stream.look().eof) {
 224                     if (stream.look() == "\\") {
 225                         if (Lang.isNewline(stream.look(1))) {
 226                             do {
 227                                 stream.next();
 228                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 229                             string += "\\\n";
 230                         }
 231                         else {
 232                             string += stream.next(2);
 233                         }
 234                     }
 235                     else if (stream.look() == "\"") {
 236                         string += stream.next();
 237                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 238                         return true;
 239                     }
 240                     else {
 241                         string += stream.next();
 242                     }
 243                 }
 244             }
 245             return false; // error! unterminated string
 246         },
 247
 248         /**
 249             @returns {Boolean} Was the token found?
 250          */
 251         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 252             if (stream.look() == "'") {
 253                 // find terminator
 254                 var string = stream.next();
 255
 256                 while (!stream.look().eof) {
 257                     if (stream.look() == "\\") { // escape sequence
 258                         string += stream.next(2);
 259                     }
 260                     else if (stream.look() == "'") {
 261                         string += stream.next();
 262                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 263                         return true;
 264                     }
 265                     else {
 266                         string += stream.next();
 267                     }
 268                 }
 269             }
 270             return false; // error! unterminated string
 271         },
 272
 273         /**
 274             @returns {Boolean} Was the token found?
 275          */
 276         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 277             if (stream.look() === "0" && stream.look(1) == "x") {
 278                 return this.read_hex(stream, tokens);
 279             }
 280
 281             var found = "";
 282
 283             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 284                 found += stream.next();
 285             }
 286
 287             if (found === "") {
 288                 return false;
 289             }
 290             else {
 291                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 292                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 293                 return true;
 294             }
 295         },
 296         /*t:
 297             requires("../lib/JSDOC/TextStream.js");
 298             requires("../lib/JSDOC/Token.js");
 299             requires("../lib/JSDOC/Lang.js");
 300
 301             plan(3, "testing read_numb");
 302
 303             //// setup
 304             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 305             var tr = new TokenReader();
 306             var tokens = tr.tokenize(new TextStream(src));
 307
 308             var hexToken, octToken, decToken;
 309             for (var i = 0; i < tokens.length; i++) {
 310                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 311                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 312                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 313             }
 314             ////
 315
 316             is(decToken.data, "8.0", "decimal number is found in source.");
 317             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 318             is(octToken.data, "0777", "octal number is found in source.");
 319         */
 320
 321         /**
 322             @returns {Boolean} Was the token found?
 323          */
 324         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 325             var found = stream.next(2);
 326
 327             while (!stream.look().eof) {
 328                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 329                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 330                     return true;
 331                 }
 332                 else {
 333                     found += stream.next();
 334                 }
 335             }
 336             return false;
 337         },
 338
 339         /**
 340             @returns {Boolean} Was the token found?
 341          */
 342         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 343             var last;
 344             if (
 345                 stream.look() == "/"
 346                 &&
 347                 (
 348
 349                     (
 350                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 351                         ||
 352                         (
 353                                !last.is("NUMB")
 354                             && !last.is("NAME")
 355                             && !last.is("RIGHT_PAREN")
 356                             && !last.is("RIGHT_BRACKET")
 357                         )
 358                     )
 359                 )
 360             ) {
 361                 var regex = stream.next();
 362
 363                 while (!stream.look().eof) {
 364                     if (stream.look() == "\\") { // escape sequence
 365                         regex += stream.next(2);
 366                     }
 367                     else if (stream.look() == "/") {
 368                         regex += stream.next();
 369
 370                         while (/[gmi]/.test(stream.look())) {
 371                             regex += stream.next();
 372                         }
 373
 374                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 375                         return true;
 376                     }
 377                     else {
 378                         regex += stream.next();
 379                     }
 380                 }
 381                 // error: unterminated regex
 382             }
 383             return false;
 384         }
 385 });