runtime/JSGlobalObjectFunctions.cpp

   1 /*
   2  *  Copyright (C) 1999-2002 Harri Porten (porten@kde.org)
   3  *  Copyright (C) 2001 Peter Kelly (pmk@post.com)
   4  *  Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008, 2009 Apple Inc. All rights reserved.
   5  *  Copyright (C) 2007 Cameron Zwarich (cwzwarich@uwaterloo.ca)
   6  *  Copyright (C) 2007 Maks Orlovich
   7  *
   8  *  This library is free software; you can redistribute it and/or
   9  *  modify it under the terms of the GNU Library General Public
  10  *  License as published by the Free Software Foundation; either
  11  *  version 2 of the License, or (at your option) any later version.
  12  *
  13  *  This library is distributed in the hope that it will be useful,
  14  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  16  *  Library General Public License for more details.
  17  *
  18  *  You should have received a copy of the GNU Library General Public License
  19  *  along with this library; see the file COPYING.LIB.  If not, write to
  20  *  the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  21  *  Boston, MA 02110-1301, USA.
  22  *
  23  */
  24
  25 #include "config.h"
  26 #include "JSGlobalObjectFunctions.h"
  27
  28 #include "CallFrame.h"
  29 #include "GlobalEvalFunction.h"
  30 #include "Interpreter.h"
  31 #include "JSGlobalObject.h"
  32 #include "JSString.h"
  33 #include "JSStringBuilder.h"
  34 #include "Lexer.h"
  35 #include "LiteralParser.h"
  36 #include "Nodes.h"
  37 #include "Parser.h"
  38 #include "StringBuilder.h"
  39 #include "StringExtras.h"
  40 #include "dtoa.h"
  41 #include <stdio.h>
  42 #include <stdlib.h>
  43 #include <string.h>
  44 #include <wtf/ASCIICType.h>
  45 #include <wtf/Assertions.h>
  46 #include <wtf/MathExtras.h>
  47 #include <wtf/unicode/UTF8.h>
  48
  49 using namespace WTF;
  50 using namespace Unicode;
  51
  52 namespace JSC {
  53
  54 static JSValue encode(ExecState* exec, const ArgList& args, const char* doNotEscape)
  55 {
  56     UString str = args.at(0).toString(exec);
  57     CString cstr = str.UTF8String(true);
  58     if (!cstr.data())
  59         return throwError(exec, URIError, "String contained an illegal UTF-16 sequence.");
  60
  61     JSStringBuilder builder;
  62     const char* p = cstr.data();
  63     for (size_t k = 0; k < cstr.length(); k++, p++) {
  64         char c = *p;
  65         if (c && strchr(doNotEscape, c))
  66             builder.append(c);
  67         else {
  68             char tmp[4];
  69             snprintf(tmp, sizeof(tmp), "%%%02X", static_cast<unsigned char>(c));
  70             builder.append(tmp);
  71         }
  72     }
  73     return builder.build(exec);
  74 }
  75
  76 static JSValue decode(ExecState* exec, const ArgList& args, const char* doNotUnescape, bool strict)
  77 {
  78     JSStringBuilder builder;
  79     UString str = args.at(0).toString(exec);
  80     int k = 0;
  81     int len = str.size();
  82     const UChar* d = str.data();
  83     UChar u = 0;
  84     while (k < len) {
  85         const UChar* p = d + k;
  86         UChar c = *p;
  87         if (c == '%') {
  88             int charLen = 0;
  89             if (k <= len - 3 && isASCIIHexDigit(p[1]) && isASCIIHexDigit(p[2])) {
  90                 const char b0 = Lexer::convertHex(p[1], p[2]);
  91                 const int sequenceLen = UTF8SequenceLength(b0);
  92                 if (sequenceLen != 0 && k <= len - sequenceLen * 3) {
  93                     charLen = sequenceLen * 3;
  94                     char sequence[5];
  95                     sequence[0] = b0;
  96                     for (int i = 1; i < sequenceLen; ++i) {
  97                         const UChar* q = p + i * 3;
  98                         if (q[0] == '%' && isASCIIHexDigit(q[1]) && isASCIIHexDigit(q[2]))
  99                             sequence[i] = Lexer::convertHex(q[1], q[2]);
 100                         else {
 101                             charLen = 0;
 102                             break;
 103                         }
 104                     }
 105                     if (charLen != 0) {
 106                         sequence[sequenceLen] = 0;
 107                         const int character = decodeUTF8Sequence(sequence);
 108                         if (character < 0 || character >= 0x110000)
 109                             charLen = 0;
 110                         else if (character >= 0x10000) {
 111                             // Convert to surrogate pair.
 112                             builder.append(static_cast<UChar>(0xD800 | ((character - 0x10000) >> 10)));
 113                             u = static_cast<UChar>(0xDC00 | ((character - 0x10000) & 0x3FF));
 114                         } else
 115                             u = static_cast<UChar>(character);
 116                     }
 117                 }
 118             }
 119             if (charLen == 0) {
 120                 if (strict)
 121                     return throwError(exec, URIError);
 122                 // The only case where we don't use "strict" mode is the "unescape" function.
 123                 // For that, it's good to support the wonky "%u" syntax for compatibility with WinIE.
 124                 if (k <= len - 6 && p[1] == 'u'
 125                         && isASCIIHexDigit(p[2]) && isASCIIHexDigit(p[3])
 126                         && isASCIIHexDigit(p[4]) && isASCIIHexDigit(p[5])) {
 127                     charLen = 6;
 128                     u = Lexer::convertUnicode(p[2], p[3], p[4], p[5]);
 129                 }
 130             }
 131             if (charLen && (u == 0 || u >= 128 || !strchr(doNotUnescape, u))) {
 132                 c = u;
 133                 k += charLen - 1;
 134             }
 135         }
 136         k++;
 137         builder.append(c);
 138     }
 139     return builder.build(exec);
 140 }
 141
 142 bool isStrWhiteSpace(UChar c)
 143 {
 144     switch (c) {
 145         case 0x0009:
 146         case 0x000A:
 147         case 0x000B:
 148         case 0x000C:
 149         case 0x000D:
 150         case 0x0020:
 151         case 0x00A0:
 152         case 0x2028:
 153         case 0x2029:
 154             return true;
 155         default:
 156             return c > 0xff && isSeparatorSpace(c);
 157     }
 158 }
 159
 160 static int parseDigit(unsigned short c, int radix)
 161 {
 162     int digit = -1;
 163
 164     if (c >= '0' && c <= '9')
 165         digit = c - '0';
 166     else if (c >= 'A' && c <= 'Z')
 167         digit = c - 'A' + 10;
 168     else if (c >= 'a' && c <= 'z')
 169         digit = c - 'a' + 10;
 170
 171     if (digit >= radix)
 172         return -1;
 173     return digit;
 174 }
 175
 176 double parseIntOverflow(const char* s, int length, int radix)
 177 {
 178     double number = 0.0;
 179     double radixMultiplier = 1.0;
 180
 181     for (const char* p = s + length - 1; p >= s; p--) {
 182         if (radixMultiplier == Inf) {
 183             if (*p != '0') {
 184                 number = Inf;
 185                 break;
 186             }
 187         } else {
 188             int digit = parseDigit(*p, radix);
 189             number += digit * radixMultiplier;
 190         }
 191
 192         radixMultiplier *= radix;
 193     }
 194
 195     return number;
 196 }
 197
 198 static double parseInt(const UString& s, int radix)
 199 {
 200     int length = s.size();
 201     const UChar* data = s.data();
 202     int p = 0;
 203
 204     while (p < length && isStrWhiteSpace(data[p]))
 205         ++p;
 206
 207     double sign = 1;
 208     if (p < length) {
 209         if (data[p] == '+')
 210             ++p;
 211         else if (data[p] == '-') {
 212             sign = -1;
 213             ++p;
 214         }
 215     }
 216
 217     if ((radix == 0 || radix == 16) && length - p >= 2 && data[p] == '0' && (data[p + 1] == 'x' || data[p + 1] == 'X')) {
 218         radix = 16;
 219         p += 2;
 220     } else if (radix == 0) {
 221         if (p < length && data[p] == '0')
 222             radix = 8;
 223         else
 224             radix = 10;
 225     }
 226
 227     if (radix < 2 || radix > 36)
 228         return NaN;
 229
 230     int firstDigitPosition = p;
 231     bool sawDigit = false;
 232     double number = 0;
 233     while (p < length) {
 234         int digit = parseDigit(data[p], radix);
 235         if (digit == -1)
 236             break;
 237         sawDigit = true;
 238         number *= radix;
 239         number += digit;
 240         ++p;
 241     }
 242
 243     if (number >= mantissaOverflowLowerBound) {
 244         // FIXME: It is incorrect to use UString::ascii() here because it's not thread-safe.
 245         if (radix == 10)
 246             number = WTF::strtod(s.substr(firstDigitPosition, p - firstDigitPosition).ascii(), 0);
 247         else if (radix == 2 || radix == 4 || radix == 8 || radix == 16 || radix == 32)
 248             number = parseIntOverflow(s.substr(firstDigitPosition, p - firstDigitPosition).ascii(), p - firstDigitPosition, radix);
 249     }
 250
 251     if (!sawDigit)
 252         return NaN;
 253
 254     return sign * number;
 255 }
 256
 257 static double parseFloat(const UString& s)
 258 {
 259     // Check for 0x prefix here, because toDouble allows it, but we must treat it as 0.
 260     // Need to skip any whitespace and then one + or - sign.
 261     int length = s.size();
 262     const UChar* data = s.data();
 263     int p = 0;
 264     while (p < length && isStrWhiteSpace(data[p]))
 265         ++p;
 266
 267     if (p < length && (data[p] == '+' || data[p] == '-'))
 268         ++p;
 269
 270     if (length - p >= 2 && data[p] == '0' && (data[p + 1] == 'x' || data[p + 1] == 'X'))
 271         return 0;
 272
 273     // FIXME: UString::toDouble will ignore leading ASCII spaces, but we need to ignore
 274     // other StrWhiteSpaceChar values as well.
 275     return s.toDouble(true /*tolerant*/, false /* NaN for empty string */);
 276 }
 277
 278 JSValue JSC_HOST_CALL globalFuncEval(ExecState* exec, JSObject* function, JSValue thisValue, const ArgList& args)
 279 {
 280     JSObject* thisObject = thisValue.toThisObject(exec);
 281     JSObject* unwrappedObject = thisObject->unwrappedObject();
 282     if (!unwrappedObject->isGlobalObject() || static_cast<JSGlobalObject*>(unwrappedObject)->evalFunction() != function)
 283         return throwError(exec, EvalError, "The \"this\" value passed to eval must be the global object from which eval originated");
 284
 285     JSValue x = args.at(0);
 286     if (!x.isString())
 287         return x;
 288
 289     UString s = x.toString(exec);
 290
 291     LiteralParser preparser(exec, s, LiteralParser::NonStrictJSON);
 292     if (JSValue parsedObject = preparser.tryLiteralParse())
 293         return parsedObject;
 294
 295     RefPtr<EvalExecutable> eval = EvalExecutable::create(exec, makeSource(s));
 296     JSObject* error = eval->compile(exec, static_cast<JSGlobalObject*>(unwrappedObject)->globalScopeChain().node());
 297     if (error)
 298         return throwError(exec, error);
 299
 300     return exec->interpreter()->execute(eval.get(), exec, thisObject, static_cast<JSGlobalObject*>(unwrappedObject)->globalScopeChain().node(), exec->exceptionSlot());
 301 }
 302
 303 JSValue JSC_HOST_CALL globalFuncParseInt(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 304 {
 305     JSValue value = args.at(0);
 306     int32_t radix = args.at(1).toInt32(exec);
 307
 308     if (radix != 0 && radix != 10)
 309         return jsNumber(exec, parseInt(value.toString(exec), radix));
 310
 311     if (value.isInt32())
 312         return value;
 313
 314     if (value.isDouble()) {
 315         double d = value.asDouble();
 316         if (isfinite(d))
 317             return jsNumber(exec, (d > 0) ? floor(d) : ceil(d));
 318         if (isnan(d) || isinf(d))
 319             return jsNaN(exec);
 320         return jsNumber(exec, 0);
 321     }
 322
 323     return jsNumber(exec, parseInt(value.toString(exec), radix));
 324 }
 325
 326 JSValue JSC_HOST_CALL globalFuncParseFloat(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 327 {
 328     return jsNumber(exec, parseFloat(args.at(0).toString(exec)));
 329 }
 330
 331 JSValue JSC_HOST_CALL globalFuncIsNaN(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 332 {
 333     return jsBoolean(isnan(args.at(0).toNumber(exec)));
 334 }
 335
 336 JSValue JSC_HOST_CALL globalFuncIsFinite(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 337 {
 338     double n = args.at(0).toNumber(exec);
 339     return jsBoolean(!isnan(n) && !isinf(n));
 340 }
 341
 342 JSValue JSC_HOST_CALL globalFuncDecodeURI(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 343 {
 344     static const char do_not_unescape_when_decoding_URI[] =
 345         "#$&+,/:;=?@";
 346
 347     return decode(exec, args, do_not_unescape_when_decoding_URI, true);
 348 }
 349
 350 JSValue JSC_HOST_CALL globalFuncDecodeURIComponent(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 351 {
 352     return decode(exec, args, "", true);
 353 }
 354
 355 JSValue JSC_HOST_CALL globalFuncEncodeURI(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 356 {
 357     static const char do_not_escape_when_encoding_URI[] =
 358         "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 359         "abcdefghijklmnopqrstuvwxyz"
 360         "0123456789"
 361         "!#$&'()*+,-./:;=?@_~";
 362
 363     return encode(exec, args, do_not_escape_when_encoding_URI);
 364 }
 365
 366 JSValue JSC_HOST_CALL globalFuncEncodeURIComponent(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 367 {
 368     static const char do_not_escape_when_encoding_URI_component[] =
 369         "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 370         "abcdefghijklmnopqrstuvwxyz"
 371         "0123456789"
 372         "!'()*-._~";
 373
 374     return encode(exec, args, do_not_escape_when_encoding_URI_component);
 375 }
 376
 377 JSValue JSC_HOST_CALL globalFuncEscape(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 378 {
 379     static const char do_not_escape[] =
 380         "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 381         "abcdefghijklmnopqrstuvwxyz"
 382         "0123456789"
 383         "*+-./@_";
 384
 385     JSStringBuilder builder;
 386     UString str = args.at(0).toString(exec);
 387     const UChar* c = str.data();
 388     for (unsigned k = 0; k < str.size(); k++, c++) {
 389         int u = c[0];
 390         if (u > 255) {
 391             char tmp[7];
 392             snprintf(tmp, sizeof(tmp), "%%u%04X", u);
 393             builder.append(tmp);
 394         } else if (u != 0 && strchr(do_not_escape, static_cast<char>(u)))
 395             builder.append(c, 1);
 396         else {
 397             char tmp[4];
 398             snprintf(tmp, sizeof(tmp), "%%%02X", u);
 399             builder.append(tmp);
 400         }
 401     }
 402
 403     return builder.build(exec);
 404 }
 405
 406 JSValue JSC_HOST_CALL globalFuncUnescape(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 407 {
 408     StringBuilder builder;
 409     UString str = args.at(0).toString(exec);
 410     int k = 0;
 411     int len = str.size();
 412     while (k < len) {
 413         const UChar* c = str.data() + k;
 414         UChar u;
 415         if (c[0] == '%' && k <= len - 6 && c[1] == 'u') {
 416             if (isASCIIHexDigit(c[2]) && isASCIIHexDigit(c[3]) && isASCIIHexDigit(c[4]) && isASCIIHexDigit(c[5])) {
 417                 u = Lexer::convertUnicode(c[2], c[3], c[4], c[5]);
 418                 c = &u;
 419                 k += 5;
 420             }
 421         } else if (c[0] == '%' && k <= len - 3 && isASCIIHexDigit(c[1]) && isASCIIHexDigit(c[2])) {
 422             u = UChar(Lexer::convertHex(c[1], c[2]));
 423             c = &u;
 424             k += 2;
 425         }
 426         k++;
 427         builder.append(*c);
 428     }
 429
 430     return jsString(exec, builder.build());
 431 }
 432
 433 #ifndef NDEBUG
 434 JSValue JSC_HOST_CALL globalFuncJSCPrint(ExecState* exec, JSObject*, JSValue, const ArgList& args)
 435 {
 436     CString string = args.at(0).toString(exec).UTF8String();
 437     puts(string.data());
 438     return jsUndefined();
 439 }
 440 #endif
 441
 442 } // namespace JSC