© 2005 Goetz Heller
Table of Contents Description Examples Copyright Note back///
///
// This file implements base64 encoding and decoding.
// Encoding is done by the function base64Encode(), decoding
// by base64Decode(). The naming mimics closely the corresponding
// library functions found in PHP. However, this implementation allows
// for a more flexible use.
//
// This implementation follows RFC 3548 (http://www.faqs.org/rfcs/rfc3548.html),
// so the copyright formulated therein applies.
//
// Dr.Heller Information Management, 2005 (http://www.hellerim.de).
//
///
try {
if (system.utility.base64 == null) {
throw new Error('');
};
} catch(e) { // don't install twice
// dependencies
try {
eval('core');
} catch(e) {
throw new Error('base64: class \'core\' not installed');
}
var base64 = function(){};
// provide for class information
base64.classID = function() {
return 'system.utility.base64';
};
//disallow subclassing
base64.isFinal = function() {
return true;
};
// original base64 encoding
base64.encString = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/';
// URL and file name safe encoding
base64.encStringS = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_';
/// BEGIN_DOC(base64).METHOD(encode)
///
// method String base64.encode(INPUTTYPE inp [, bool uc [, bool safe]])
//
// Encode input data into a base64 character string.
//
// Function arguments:
// INPUTTYPE inp: data to be encoded. INPUTTYPE may be String or Array.
// Any other INPUTTYPE results in an output value of null.
// If INPUTTYPE is String each character is converted into
// two bytes each of which is encoded separately.
// bool uc Optional. If this parameter has a value of 'true' which is
// the default, code of each character is treated as a 16-bit
// entity (UniCode), i.e. as two bytes. Otherwise, the codes
// are truncated to one byte (8-bit character set) which
// may result in information loss. If INPUTTYPE is Array,
// the value of this parameter has no effect.
// bool safe: Optioanal. If this parameter is set to true, the standard base64
// character set is replaced with a modified version where
// the characters '+' and '/' are replace with '-' and '_',
// repectively, in order to avoid problems with file system
// namings which otherwise could occur on some systems.
// By default, the value of this argument is assumed to be
// false.
// Return value: The function returns a character string consisting of
// the base64 representaion of the input. Its length is a
// multiple of 4. If the encoding yields less than this
// the string is stuffed with the '=' character. In each case,
// the string maybe empty but not null if no error occurred.
// Errors: Whenever an error occurs, null is returned. Parameter values
// not defined above are considered errors.
// Remarks: If the input array contains something different from
// a byte at some position the first 8 bits only of this entity are
// processed silently without returning an error, which probably
// results in garbage converted to base64.
//
/// END_DOC
base64.encode = function(inp, uc, safe) {
// do some argument checking
if (arguments.length < 1) return null;
var readBuf = new Array(); // read buffer
if (arguments.length >= 3 && safe != true && safe != false) return null;
var enc = (arguments.length >= 3 && safe) ? this.encStringS : this.encString; // character set used
var b = (typeof inp == "string"); // how input is to be processed
if (!b && (typeof inp != "object") && !(inp instanceof Array)) return null; // bad input
if (arguments.length < 2) {
uc = true; // set default
} // otherwise its value is passed from the caller
if (uc != true && uc != false) return null;
var n = (!b || !uc) ? 1 : 2; // length of read buffer
var out = ''; // output string
var c = 0; // holds character code (maybe 16 bit or 8 bit)
var j = 1; // sextett counter
var l = 0; // work buffer
var s = 0; // holds sextett
// convert
for (var i = 0; i < inp.length; i++) { // read input
c = (b) ? inp.charCodeAt(i) : inp[i]; // fill read buffer
for (var k = n - 1; k >= 0; k--) {
readBuf[k] = c & 0xff;
c >>= 8;
}
for (var m = 0; m < n; m++) { // run through read buffer
// process bytes from read buffer
l = ((l<<8)&0xff00) | readBuf[m]; // shift remaining bits one byte to the left and append next byte
s = (0x3f<<(2*j)) & l; // extract sextett from buffer
l -=s; // remove those bits from buffer;
out += enc.charAt(s>>(2*j)); // convert leftmost sextett and append it to output
j++;
if (j==4) { // another sextett is complete
out += enc.charAt(l&0x3f); // convert and append it
j = 1;
}
}
}
switch (j) { // handle left-over sextetts
case 2:
s = 0x3f & (16 * l); // extract sextett from buffer
out += enc.charAt(s); // convert leftmost sextett and append it to output
out += '=='; // stuff
break;
case 3:
s = 0x3f & (4 * l); // extract sextett from buffer
out += enc.charAt(s); // convert leftmost sextett and append it to output
out += '='; // stuff
break;
default:
break;
}
return out;
}
/// BEGIN_DOC(base64).METHOD(decode)
///
// method RETURNTYPE base64.decode(String inp [, enum outType [, bool safe [, bool lax]]])
//
// Encode input data into a base64 character string.
//
// Function arguments:
// String inp: base64 encoded data string to be decoded.
// enum outType Optional. This parameter specifies the type of the output and determines
// how the input data is to be interpreted.:
// 0 - binary data; create a byte array (default)
// 1 - 8-bit character string, assuming 1-byte characters encoded in inp
// 2 - 16-bit (UniCode) character string, assuming 2-byte
// characters encoded in inp
// If 2 is passed to the function, but the number of base64 characters
// is odd, a value of null is returned.
// bool safe Optional. If this parameter is set to true, the standard base64
// character set is replaced with a modified version where
// the characters '+' and '/' are replaced with '-' and '_',
// repectively, in order to avoid problems with file system
// namings which otherwise could occur on some systems.
// By default, the value of this argument is assumed to be
// false.
// bool lax Optional. If set to true, the function skips all input characters which
// cannot be processed. This includes the character '=', too, if
// it is followed by at least one different character before the string
// ends. However, if skipping infeasible characters amounts to a number
// of allowed base64 characters which is not amultiple of 4,
// this is considered an error and null is returned.
// If lax is set to false (the default), null is returned
// whenever an infeasible character is found.
// The purpose of this parameter is to give support in cases
// where data has been base64 encoded and later on was folded by
// some other software, e.g. '\r\n\'s have been inserted in email.
// exchange.
// Return value: The function's processing result value is stored in a string or in
// a byte array before it is returned, depending on the value
// assigned to the type parameter. In each case, the value
// maybe empty but not null if no error occurred.
// Errors: Whenever an error occurs, null is returned. Parameter values
// not defined above are considered errors.
//
/// END_DOC
base64.decode = function(inp, outType, safe, lax) {
// do some argument checking
if (arguments.length < 1) return null;
if (arguments.length < 2) outType = 0 ;// produce character array by default
if (outType != 0 && outType != 1 && outType != 2) return null;
if (arguments.length >= 3 && safe != true && safe != false) return null;
var sEnc = (arguments.length >= 3 && safe) ? this.encStringS : this.encString; // select encoding character set
if (arguments.length >= 4 && lax != true && lax != false) return null;
var aDec = new Object(); // create an associative array for decoding
for (var p = 0; p < sEnc.length; p++) { // populate array
aDec[sEnc.charAt(p)] = p;
}
var out = (outType == 0) ? new Array() : '';
lax = (arguments.length == 4 && lax); // ignore non-base64 characters
var l = 0; // work area
var i = 0; // index into input
var j = 0; // sextett counter
var c = 0; // input buffer
var k = 0; // index into work area
var end = inp.length; // one position past the last character to be processed
var C = '';
// check input
if (lax) {
var inpS = ''; // shadow input
var ignore = false; // determines wether '=' must be counted
var cnt = 0;
for (var p = 1; p <= inp.length; p++) { // check and cleanup string before trying to decode
c = inp.charAt(end - p);
if (c == '=') {
if (!ignore) {
if (++cnt > 1) ignore = true;
} else {
continue;
}
} else if (undefined != aDec[c]) { // the character is base64, hence feasible
if (!ignore) ignore = true; // no more '=' allowed
inpS = c + inpS; // prepend c to shadow input
}
}
for (var p = 0; p <= cnt; p++) { // at most cnt '=''s were garbage, a number in
if (p == 2) return null; // [inpS.length, inpS.length + cnt] must be a
if ((inpS.length + cnt)%4 == 0) break; // multiple of 4
}
if (inpS.length%4==1) return null; // must be 0, 2, or 3 for inpS to contain correctly base64 encoded data
inp = inpS; // inp now contains feasible characters only
end = inp.length;
} else {
if (inp.length%4 > 0) return null; // invalid length
for (var p = 0; p < 2; p++) { // search for trailing '=''s
if (inp.charAt(end - 1) == '=') {
end--;
} else {
break;
}
}
}
// convert
for (i = 0; i < end; i++) {
l <<= 6; // clear space for next sextett
if (undefined == (c = aDec[inp.charAt(i)])) return null; // lax must be false at this place!
l |= (c&0x3f); // append it
if (j == 0) {
j++;
continue; // work area contains incomplete byte only
}
if (outType == 2) {
if (k == 1) { // work area contains complete double byte
out += String.fromCharCode(l>>(2*(3-j))); // convert leftmost 16 bits and append them to string
l &= ~(0xffff<<(2*(3-j))); // clear the 16 processed bits
}
k = ++k%2;
} else { // work area contains complete byte
if (outType == 0) {
out.push(l>>(2*(3-j))); // append byte to array
} else {
out += String.fromCharCode(l>>(2*(3-j))); // convert leftmost 8 bits and append them to String
}
l &= ~(0xff<<(2*(3-j))); // clear the 8 processed bits
}
j = ++j%4; // increment sextett counter cyclically
}
if (outType == 2 && k == 1) return null; // incomplete double byte in work area
return out;
}
core.installClass(base64);
} // end catch