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         XObject.extend(this, o || {});
  18
  19     },
  20     Object,
  21     {
  22         /** @cfg {Boolean} collapseWhite merge multiple whitespace/comments into a single token **/
  23         collapseWhite : false, // only reduces white space...
  24         /** @cfg {Boolean} keepDocs keep JSDOC comments **/
  25         keepDocs : true,
  26         /** @cfg {Boolean} keepWhite keep White space **/
  27         keepWhite : false,
  28         /** @cfg {Boolean} keepComments  keep all comments **/
  29         keepComments : false,
  30         /** @cfg {Boolean} sepIdents seperate identifiers (eg. a.b.c into ['a', '.', 'b', '.', 'c'] ) **/
  31         sepIdents : false,
  32         /** @cfg {String} filename name of file being parsed. **/
  33         filename : '',
  34
  35         /**
  36          * tokenize a stream
  37          * @return {Array} of tokens
  38          *
  39          * ts = new TextStream(File.read(str));
  40          * tr = TokenReader({ keepComments : true, keepWhite : true });
  41          * tr.tokenize(ts)
  42          *
  43          */
  44
  45
  46
  47         tokenize : function(/**JSDOC.TextStream*/stream) {
  48             this.line =1;
  49             var tokens = [];
  50             /**@ignore*/
  51             tokens.last    = function() { return tokens[tokens.length-1]; }
  52             /**@ignore*/
  53             tokens.lastSym = function() {
  54                 for (var i = tokens.length-1; i >= 0; i--) {
  55                     if (!(tokens[i].is("WHIT") || tokens[i].is("COMM"))) return tokens[i];
  56                 }
  57             }
  58
  59             while (!stream.look().eof) {
  60                 if (this.read_mlcomment(stream, tokens)) continue;
  61                 if (this.read_slcomment(stream, tokens)) continue;
  62                 if (this.read_dbquote(stream, tokens))   continue;
  63                 if (this.read_snquote(stream, tokens))   continue;
  64                 if (this.read_regx(stream, tokens))      continue;
  65                 if (this.read_numb(stream, tokens))      continue;
  66                 if (this.read_punc(stream, tokens))      continue;
  67                 if (this.read_newline(stream, tokens))   continue;
  68                 if (this.read_space(stream, tokens))     continue;
  69                 if (this.read_word(stream, tokens))      continue;
  70
  71                 // if execution reaches here then an error has happened
  72                 tokens.push(new Token(stream.next(), "TOKN", "UNKNOWN_TOKEN", this.line));
  73             }
  74
  75
  76
  77             return tokens;
  78         },
  79
  80         /**
  81             @returns {Boolean} Was the token found?
  82          */
  83         read_word : function(/**JSDOC.TokenStream*/stream, tokens) {
  84             var found = "";
  85             while (!stream.look().eof && Lang.isWordChar(stream.look())) {
  86                 found += stream.next();
  87             }
  88
  89             if (found === "") {
  90                 return false;
  91             }
  92
  93             var name;
  94             if ((name = Lang.keyword(found))) {
  95                 if (found == 'return' && tokens.lastSym().data == ')') {
  96                     throw {
  97                         name : "ArgumentError",
  98                         message: "\n" + this.filename + ':' + this.line + " Error - return found after )"
  99                     }
 100                 }
 101
 102                 tokens.push(new Token(found, "KEYW", name, this.line));
 103                 return true;
 104             }
 105             if (!this.sepIdents || found.indexOf('.') < 0 ) {
 106                 tokens.push(new Token(found, "NAME", "NAME", this.line));
 107                 return true;
 108             }
 109             var n = found.split('.');
 110             var p = false;
 111             var _this = this;
 112             n.forEach(function(nm) {
 113                 if (p) {
 114                     tokens.push(new Token('.', "PUNC", "DOT", _this.line));
 115                 }
 116                 p=true;
 117                 tokens.push(new Token(nm, "NAME", "NAME", _this.line));
 118             });
 119             return true;
 120
 121
 122         },
 123
 124         /**
 125             @returns {Boolean} Was the token found?
 126          */
 127         read_punc : function(/**JSDOC.TokenStream*/stream, tokens) {
 128             var found = "";
 129             var name;
 130             while (!stream.look().eof && Lang.punc(found+stream.look())) {
 131                 found += stream.next();
 132             }
 133
 134
 135             if (found === "") {
 136                 return false;
 137             }
 138
 139             if ((found == '}' || found == ']') && tokens.lastSym().data == ',') {
 140                 //print("Error - comma found before " + found);
 141                 //print(JSON.stringify(tokens.lastSym(), null,4));
 142                 throw {
 143                     name : "ArgumentError",
 144                     message: "\n" + this.filename + ':' + this.line + " Error - comma found before " + found
 145                 }
 146             }
 147
 148             tokens.push(new Token(found, "PUNC", Lang.punc(found), this.line));
 149             return true;
 150
 151         },
 152
 153         /**
 154             @returns {Boolean} Was the token found?
 155          */
 156         read_space : function(/**JSDOC.TokenStream*/stream, tokens) {
 157             var found = "";
 158
 159             while (!stream.look().eof && Lang.isSpace(stream.look()) && !Lang.isNewline(stream.look())) {
 160                 found += stream.next();
 161             }
 162
 163             if (found === "") {
 164                 return false;
 165             }
 166             //print("WHITE = " + JSON.stringify(found));
 167             if (this.collapseWhite) found = " ";
 168             if (this.keepWhite) tokens.push(new Token(found, "WHIT", "SPACE", this.line));
 169             return true;
 170
 171         },
 172
 173         /**
 174             @returns {Boolean} Was the token found?
 175          */
 176         read_newline : function(/**JSDOC.TokenStream*/stream, tokens) {
 177             var found = "";
 178             var line = this.line;
 179             while (!stream.look().eof && Lang.isNewline(stream.look())) {
 180                 this.line++;
 181                 found += stream.next();
 182             }
 183
 184             if (found === "") {
 185                 return false;
 186             }
 187             //this.line++;
 188             if (this.collapseWhite) {
 189                 found = "\n";
 190             }
 191             if (this.keepWhite) {
 192                 var last = tokens.pop();
 193                 if (last && last.name != "WHIT") {
 194                     tokens.push(last);
 195                 }
 196
 197                 tokens.push(new Token(found, "WHIT", "NEWLINE", line));
 198             }
 199             return true;
 200         },
 201
 202         /**
 203             @returns {Boolean} Was the token found?
 204          */
 205         read_mlcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 206             if (stream.look() == "/" && stream.look(1) == "*") {
 207                 var found = stream.next(2);
 208                 var c = '';
 209                 var line = this.line;
 210                 while (!stream.look().eof && !(stream.look(-1) == "/" && stream.look(-2) == "*")) {
 211                     c = stream.next();
 212                     if (c == "\n") this.line++;
 213                     found += c;
 214                 }
 215
 216                 // to start doclet we allow /** or /*** but not /**/ or /****
 217                 if (/^\/\*\*([^\/]|\*[^*])/.test(found) && this.keepDocs) tokens.push(new Token(found, "COMM", "JSDOC", this.line));
 218                 else if (this.keepComments) tokens.push(new Token(found, "COMM", "MULTI_LINE_COMM", line));
 219                 return true;
 220             }
 221             return false;
 222         },
 223
 224         /**
 225             @returns {Boolean} Was the token found?
 226          */
 227         read_slcomment : function(/**JSDOC.TokenStream*/stream, tokens) {
 228             var found;
 229             if (
 230                 (stream.look() == "/" && stream.look(1) == "/" && (found=stream.next(2)))
 231                 ||
 232                 (stream.look() == "<" && stream.look(1) == "!" && stream.look(2) == "-" && stream.look(3) == "-" && (found=stream.next(4)))
 233             ) {
 234                 var line = this.line;
 235                 while (!stream.look().eof && !Lang.isNewline(stream.look())) {
 236                     found += stream.next();
 237                 }
 238                 if (!stream.look().eof) {
 239                     found += stream.next();
 240                 }
 241                 if (this.keepComments) {
 242                     tokens.push(new Token(found, "COMM", "SINGLE_LINE_COMM", line));
 243                 }
 244                 this.line++;
 245                 return true;
 246             }
 247             return false;
 248         },
 249
 250         /**
 251             @returns {Boolean} Was the token found?
 252          */
 253         read_dbquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 254             if (stream.look() == "\"") {
 255                 // find terminator
 256                 var string = stream.next();
 257
 258                 while (!stream.look().eof) {
 259                     if (stream.look() == "\\") {
 260                         if (Lang.isNewline(stream.look(1))) {
 261                             do {
 262                                 stream.next();
 263                             } while (!stream.look().eof && Lang.isNewline(stream.look()));
 264                             string += "\\\n";
 265                         }
 266                         else {
 267                             string += stream.next(2);
 268                         }
 269                     }
 270                     else if (stream.look() == "\"") {
 271                         string += stream.next();
 272                         tokens.push(new Token(string, "STRN", "DOUBLE_QUOTE", this.line));
 273                         return true;
 274                     }
 275                     else {
 276                         string += stream.next();
 277                     }
 278                 }
 279             }
 280             return false; // error! unterminated string
 281         },
 282
 283         /**
 284             @returns {Boolean} Was the token found?
 285          */
 286         read_snquote : function(/**JSDOC.TokenStream*/stream, tokens) {
 287             if (stream.look() == "'") {
 288                 // find terminator
 289                 var string = stream.next();
 290
 291                 while (!stream.look().eof) {
 292                     if (stream.look() == "\\") { // escape sequence
 293                         string += stream.next(2);
 294                     }
 295                     else if (stream.look() == "'") {
 296                         string += stream.next();
 297                         tokens.push(new Token(string, "STRN", "SINGLE_QUOTE", this.line));
 298                         return true;
 299                     }
 300                     else {
 301                         string += stream.next();
 302                     }
 303                 }
 304             }
 305             return false; // error! unterminated string
 306         },
 307
 308         /**
 309             @returns {Boolean} Was the token found?
 310          */
 311         read_numb : function(/**JSDOC.TokenStream*/stream, tokens) {
 312             if (stream.look() === "0" && stream.look(1) == "x") {
 313                 return this.read_hex(stream, tokens);
 314             }
 315
 316             var found = "";
 317
 318             while (!stream.look().eof && Lang.isNumber(found+stream.look())){
 319                 found += stream.next();
 320             }
 321
 322             if (found === "") {
 323                 return false;
 324             }
 325             else {
 326                 if (/^0[0-7]/.test(found)) tokens.push(new Token(found, "NUMB", "OCTAL", this.line));
 327                 else tokens.push(new Token(found, "NUMB", "DECIMAL", this.line));
 328                 return true;
 329             }
 330         },
 331         /*t:
 332             requires("../lib/JSDOC/TextStream.js");
 333             requires("../lib/JSDOC/Token.js");
 334             requires("../lib/JSDOC/Lang.js");
 335
 336             plan(3, "testing read_numb");
 337
 338             //// setup
 339             var src = "function foo(num){while (num+8.0 >= 0x20 && num < 0777){}}";
 340             var tr = new TokenReader();
 341             var tokens = tr.tokenize(new TextStream(src));
 342
 343             var hexToken, octToken, decToken;
 344             for (var i = 0; i < tokens.length; i++) {
 345                 if (tokens[i].name == "HEX_DEC") hexToken = tokens[i];
 346                 if (tokens[i].name == "OCTAL") octToken = tokens[i];
 347                 if (tokens[i].name == "DECIMAL") decToken = tokens[i];
 348             }
 349             ////
 350
 351             is(decToken.data, "8.0", "decimal number is found in source.");
 352             is(hexToken.data, "0x20", "hexdec number is found in source (issue #99).");
 353             is(octToken.data, "0777", "octal number is found in source.");
 354         */
 355
 356         /**
 357             @returns {Boolean} Was the token found?
 358          */
 359         read_hex : function(/**JSDOC.TokenStream*/stream, tokens) {
 360             var found = stream.next(2);
 361
 362             while (!stream.look().eof) {
 363                 if (Lang.isHexDec(found) && !Lang.isHexDec(found+stream.look())) { // done
 364                     tokens.push(new Token(found, "NUMB", "HEX_DEC", this.line));
 365                     return true;
 366                 }
 367                 else {
 368                     found += stream.next();
 369                 }
 370             }
 371             return false;
 372         },
 373
 374         /**
 375             @returns {Boolean} Was the token found?
 376          */
 377         read_regx : function(/**JSDOC.TokenStream*/stream, tokens) {
 378             var last;
 379             if (
 380                 stream.look() == "/"
 381                 &&
 382                 (
 383
 384                     (
 385                         !(last = tokens.lastSym()) // there is no last, the regex is the first symbol
 386                         ||
 387                         (
 388                                !last.is("NUMB")
 389                             && !last.is("NAME")
 390                             && !last.is("RIGHT_PAREN")
 391                             && !last.is("RIGHT_BRACKET")
 392                         )
 393                     )
 394                 )
 395             ) {
 396                 var regex = stream.next();
 397
 398                 while (!stream.look().eof) {
 399                     if (stream.look() == "\\") { // escape sequence
 400                         regex += stream.next(2);
 401                     }
 402                     else if (stream.look() == "/") {
 403                         regex += stream.next();
 404
 405                         while (/[gmi]/.test(stream.look())) {
 406                             regex += stream.next();
 407                         }
 408
 409                         tokens.push(new Token(regex, "REGX", "REGX", this.line));
 410                         return true;
 411                     }
 412                     else {
 413                         regex += stream.next();
 414                     }
 415                 }
 416                 // error: unterminated regex
 417             }
 418             return false;
 419         }
 420 });