Home | History | Annotate | Download | only in src 
      1 // Copyright 2006-2008 the V8 project authors. All rights reserved.
      2 // Redistribution and use in source and binary forms, with or without
      3 // modification, are permitted provided that the following conditions are
      4 // met:
      5 //
      6 //     * Redistributions of source code must retain the above copyright
      7 //       notice, this list of conditions and the following disclaimer.
      8 //     * Redistributions in binary form must reproduce the above
      9 //       copyright notice, this list of conditions and the following
     10 //       disclaimer in the documentation and/or other materials provided
     11 //       with the distribution.
     12 //     * Neither the name of Google Inc. nor the names of its
     13 //       contributors may be used to endorse or promote products derived
     14 //       from this software without specific prior written permission.
     15 //
     16 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
     17 // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
     18 // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
     19 // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
     20 // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
     21 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
     22 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
     23 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
     24 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
     25 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
     26 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
     27 
     28 // This file contains support for URI manipulations written in
     29 // JavaScript.
     30 
     31 // Expect $String = global.String;
     32 
     33 // Lazily initialized.
     34 var hexCharArray = 0;
     35 var hexCharCodeArray = 0;
     36 
     37 
     38 function URIAddEncodedOctetToBuffer(octet, result, index) {
     39   result[index++] = 37; // Char code of '%'.
     40   result[index++] = hexCharCodeArray[octet >> 4];
     41   result[index++] = hexCharCodeArray[octet & 0x0F];
     42   return index;
     43 }
     44 
     45 
     46 function URIEncodeOctets(octets, result, index) {
     47   if (hexCharCodeArray === 0) {
     48     hexCharCodeArray = [48, 49, 50, 51, 52, 53, 54, 55, 56, 57,
     49                         65, 66, 67, 68, 69, 70];
     50   }
     51   index = URIAddEncodedOctetToBuffer(octets[0], result, index);
     52   if (octets[1]) index = URIAddEncodedOctetToBuffer(octets[1], result, index);
     53   if (octets[2]) index = URIAddEncodedOctetToBuffer(octets[2], result, index);
     54   if (octets[3]) index = URIAddEncodedOctetToBuffer(octets[3], result, index);
     55   return index;
     56 }
     57 
     58 
     59 function URIEncodeSingle(cc, result, index) {
     60   var x = (cc >> 12) & 0xF;
     61   var y = (cc >> 6) & 63;
     62   var z = cc & 63;
     63   var octets = new $Array(3);
     64   if (cc <= 0x007F) {
     65     octets[0] = cc;
     66   } else if (cc <= 0x07FF) {
     67     octets[0] = y + 192;
     68     octets[1] = z + 128;
     69   } else {
     70     octets[0] = x + 224;
     71     octets[1] = y + 128;
     72     octets[2] = z + 128;
     73   }
     74   return URIEncodeOctets(octets, result, index);
     75 }
     76 
     77 
     78 function URIEncodePair(cc1 , cc2, result, index) {
     79   var u = ((cc1 >> 6) & 0xF) + 1;
     80   var w = (cc1 >> 2) & 0xF;
     81   var x = cc1 & 3;
     82   var y = (cc2 >> 6) & 0xF;
     83   var z = cc2 & 63;
     84   var octets = new $Array(4);
     85   octets[0] = (u >> 2) + 240;
     86   octets[1] = (((u & 3) << 4) | w) + 128;
     87   octets[2] = ((x << 4) | y) + 128;
     88   octets[3] = z + 128;
     89   return URIEncodeOctets(octets, result, index);
     90 }
     91 
     92 
     93 function URIHexCharsToCharCode(highChar, lowChar) {
     94   var highCode = HexValueOf(highChar);
     95   var lowCode = HexValueOf(lowChar);
     96   if (highCode == -1 || lowCode == -1) {
     97     throw new $URIError("URI malformed");
     98   }
     99   return (highCode << 4) | lowCode;
    100 }
    101 
    102 
    103 function URIDecodeOctets(octets, result, index) {
    104   var value;
    105   var o0 = octets[0];
    106   if (o0 < 0x80) {
    107     value = o0;
    108   } else if (o0 < 0xc2) {
    109     throw new $URIError("URI malformed");
    110   } else {
    111     var o1 = octets[1];
    112     if (o0 < 0xe0) {
    113       var a = o0 & 0x1f;
    114       if ((o1 < 0x80) || (o1 > 0xbf))
    115         throw new $URIError("URI malformed");
    116       var b = o1 & 0x3f;
    117       value = (a << 6) + b;
    118       if (value < 0x80 || value > 0x7ff)
    119         throw new $URIError("URI malformed");
    120     } else {
    121       var o2 = octets[2];
    122       if (o0 < 0xf0) {
    123         var a = o0 & 0x0f;
    124         if ((o1 < 0x80) || (o1 > 0xbf))
    125           throw new $URIError("URI malformed");
    126         var b = o1 & 0x3f;
    127         if ((o2 < 0x80) || (o2 > 0xbf))
    128           throw new $URIError("URI malformed");
    129         var c = o2 & 0x3f;
    130         value = (a << 12) + (b << 6) + c;
    131         if ((value < 0x800) || (value > 0xffff))
    132           throw new $URIError("URI malformed");
    133       } else {
    134         var o3 = octets[3];
    135         if (o0 < 0xf8) {
    136           var a = (o0 & 0x07);
    137           if ((o1 < 0x80) || (o1 > 0xbf))
    138             throw new $URIError("URI malformed");
    139           var b = (o1 & 0x3f);
    140           if ((o2 < 0x80) || (o2 > 0xbf))
    141             throw new $URIError("URI malformed");
    142           var c = (o2 & 0x3f);
    143           if ((o3 < 0x80) || (o3 > 0xbf))
    144             throw new $URIError("URI malformed");
    145           var d = (o3 & 0x3f);
    146           value = (a << 18) + (b << 12) + (c << 6) + d;
    147           if ((value < 0x10000) || (value > 0x10ffff))
    148             throw new $URIError("URI malformed");
    149         } else {
    150           throw new $URIError("URI malformed");
    151         }
    152       }
    153     }
    154   }
    155   if (value < 0x10000) {
    156     result[index++] = value;
    157     return index;
    158   } else {
    159     result[index++] = (value >> 10) + 0xd7c0;
    160     result[index++] = (value & 0x3ff) + 0xdc00;
    161     return index;
    162   }
    163 }
    164 
    165 
    166 // ECMA-262, section 15.1.3
    167 function Encode(uri, unescape) {
    168   var uriLength = uri.length;
    169   var result = new $Array(uriLength);
    170   var index = 0;
    171   for (var k = 0; k < uriLength; k++) {
    172     var cc1 = uri.charCodeAt(k);
    173     if (unescape(cc1)) {
    174       result[index++] = cc1;
    175     } else {
    176       if (cc1 >= 0xDC00 && cc1 <= 0xDFFF) throw new $URIError("URI malformed");
    177       if (cc1 < 0xD800 || cc1 > 0xDBFF) {
    178         index = URIEncodeSingle(cc1, result, index);
    179       } else {
    180         k++;
    181         if (k == uriLength) throw new $URIError("URI malformed");
    182         var cc2 = uri.charCodeAt(k);
    183         if (cc2 < 0xDC00 || cc2 > 0xDFFF) throw new $URIError("URI malformed");
    184         index = URIEncodePair(cc1, cc2, result, index);
    185       }
    186     }
    187   }
    188   return %StringFromCharCodeArray(result);
    189 }
    190 
    191 
    192 // ECMA-262, section 15.1.3
    193 function Decode(uri, reserved) {
    194   var uriLength = uri.length;
    195   var result = new $Array(uriLength);
    196   var index = 0;
    197   for (var k = 0; k < uriLength; k++) {
    198     var ch = uri.charAt(k);
    199     if (ch == '%') {
    200       if (k + 2 >= uriLength) throw new $URIError("URI malformed");
    201       var cc = URIHexCharsToCharCode(uri.charCodeAt(++k), uri.charCodeAt(++k));
    202       if (cc >> 7) {
    203         var n = 0;
    204         while (((cc << ++n) & 0x80) != 0) ;
    205         if (n == 1 || n > 4) throw new $URIError("URI malformed");
    206         var octets = new $Array(n);
    207         octets[0] = cc;
    208         if (k + 3 * (n - 1) >= uriLength) throw new $URIError("URI malformed");
    209         for (var i = 1; i < n; i++) {
    210           if (uri.charAt(++k) != '%') throw new $URIError("URI malformed");
    211           octets[i] = URIHexCharsToCharCode(uri.charCodeAt(++k), uri.charCodeAt(++k));
    212         }
    213         index = URIDecodeOctets(octets, result, index);
    214       } else {
    215         if (reserved(cc)) {
    216           result[index++] = 37; // Char code of '%'.
    217           result[index++] = uri.charCodeAt(k - 1);
    218           result[index++] = uri.charCodeAt(k);
    219         } else {
    220           result[index++] = cc;
    221         }
    222       }
    223     } else {
    224       result[index++] = ch.charCodeAt(0);
    225     }
    226   }
    227   result.length = index;
    228   return %StringFromCharCodeArray(result);
    229 }
    230 
    231 
    232 // ECMA-262 - 15.1.3.1.
    233 function URIDecode(uri) {
    234   function reservedPredicate(cc) {
    235     // #$
    236     if (35 <= cc && cc <= 36) return true;
    237     // &
    238     if (cc == 38) return true;
    239     // +,
    240     if (43 <= cc && cc <= 44) return true;
    241     // /
    242     if (cc == 47) return true;
    243     // :;
    244     if (58 <= cc && cc <= 59) return true;
    245     // =
    246     if (cc == 61) return true;
    247     // ?@
    248     if (63 <= cc && cc <= 64) return true;
    249 
    250     return false;
    251   };
    252   var string = ToString(uri);
    253   return Decode(string, reservedPredicate);
    254 }
    255 
    256 
    257 // ECMA-262 - 15.1.3.2.
    258 function URIDecodeComponent(component) {
    259   function reservedPredicate(cc) { return false; };
    260   var string = ToString(component);
    261   return Decode(string, reservedPredicate);
    262 }
    263 
    264 
    265 // Does the char code correspond to an alpha-numeric char.
    266 function isAlphaNumeric(cc) {
    267   // a - z
    268   if (97 <= cc && cc <= 122) return true;
    269   // A - Z
    270   if (65 <= cc && cc <= 90) return true;
    271   // 0 - 9
    272   if (48 <= cc && cc <= 57) return true;
    273 
    274   return false;
    275 }
    276 
    277 
    278 // ECMA-262 - 15.1.3.3.
    279 function URIEncode(uri) {
    280   function unescapePredicate(cc) {
    281     if (isAlphaNumeric(cc)) return true;
    282     // !
    283     if (cc == 33) return true;
    284     // #$
    285     if (35 <= cc && cc <= 36) return true;
    286     // &'()*+,-./
    287     if (38 <= cc && cc <= 47) return true;
    288     // :;
    289     if (58 <= cc && cc <= 59) return true;
    290     // =
    291     if (cc == 61) return true;
    292     // ?@
    293     if (63 <= cc && cc <= 64) return true;
    294     // _
    295     if (cc == 95) return true;
    296     // ~
    297     if (cc == 126) return true;
    298 
    299     return false;
    300   };
    301 
    302   var string = ToString(uri);
    303   return Encode(string, unescapePredicate);
    304 }
    305 
    306 
    307 // ECMA-262 - 15.1.3.4
    308 function URIEncodeComponent(component) {
    309   function unescapePredicate(cc) {
    310     if (isAlphaNumeric(cc)) return true;
    311     // !
    312     if (cc == 33) return true;
    313     // '()*
    314     if (39 <= cc && cc <= 42) return true;
    315     // -.
    316     if (45 <= cc && cc <= 46) return true;
    317     // _
    318     if (cc == 95) return true;
    319     // ~
    320     if (cc == 126) return true;
    321 
    322     return false;
    323   };
    324 
    325   var string = ToString(component);
    326   return Encode(string, unescapePredicate);
    327 }
    328 
    329 
    330 function HexValueOf(code) {
    331   // 0-9
    332   if (code >= 48 && code <= 57) return code - 48;
    333   // A-F
    334   if (code >= 65 && code <= 70) return code - 55;
    335   // a-f
    336   if (code >= 97 && code <= 102) return code - 87;
    337 
    338   return -1;
    339 }
    340 
    341 
    342 // Convert a character code to 4-digit hex string representation
    343 // 64 -> 0040, 62234 -> F31A.
    344 function CharCodeToHex4Str(cc) {
    345   var r = "";
    346   if (hexCharArray === 0) {
    347     hexCharArray = ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9",
    348                     "A", "B", "C", "D", "E", "F"];
    349   }
    350   for (var i = 0; i < 4; ++i) {
    351     var c = hexCharArray[cc & 0x0F];
    352     r = c + r;
    353     cc = cc >>> 4;
    354   }
    355   return r;
    356 }
    357 
    358 
    359 // Returns true if all digits in string s are valid hex numbers
    360 function IsValidHex(s) {
    361   for (var i = 0; i < s.length; ++i) {
    362     var cc = s.charCodeAt(i);
    363     if ((48 <= cc && cc <= 57) || (65 <= cc && cc <= 70) || (97 <= cc && cc <= 102)) {
    364       // '0'..'9', 'A'..'F' and 'a' .. 'f'.
    365     } else {
    366       return false;
    367     }
    368   }
    369   return true;
    370 }
    371 
    372 
    373 // ECMA-262 - B.2.1.
    374 function URIEscape(str) {
    375   var s = ToString(str);
    376   return %URIEscape(s);
    377 }
    378 
    379 
    380 // ECMA-262 - B.2.2.
    381 function URIUnescape(str) {
    382   var s = ToString(str);
    383   return %URIUnescape(s);
    384 }
    385 
    386 
    387 // -------------------------------------------------------------------
    388 
    389 function SetupURI() {
    390   // Setup non-enumerable URI functions on the global object and set
    391   // their names.
    392   InstallFunctions(global, DONT_ENUM, $Array(
    393     "escape", URIEscape,
    394     "unescape", URIUnescape,
    395     "decodeURI", URIDecode,
    396     "decodeURIComponent", URIDecodeComponent,
    397     "encodeURI", URIEncode,
    398     "encodeURIComponent", URIEncodeComponent
    399   ));
    400 }
    401 
    402 SetupURI();
    403