// This function will encode all "special" characters in typical url
// encoding, that is %hh where h is a valid hex digit. It will also fold
// any duplicated slashes.
bool
nsUrlClassifierUtils::SpecialEncode(const nsACString & url,
bool foldSlashes,
nsACString & _retval)
{
bool changed = false;
const char* curChar = url.BeginReading();
const char* end = url.EndReading();
unsigned char lastChar = '\0';
while (curChar != end) {
unsigned char c = static_cast<unsigned char>(*curChar);
if (ShouldURLEscape(c)) {
_retval.Append('%');
_retval.Append(int_to_hex_digit(c / 16));
_retval.Append(int_to_hex_digit(c % 16));
changed = true;
} else if (foldSlashes && (c == '/' && lastChar == '/')) {
// skip
} else {
_retval.Append(*curChar);
}
lastChar = c;
curChar++;
}
return changed;
}
void TestEnc()
{
// Test empty string
TestEncodeHelper("", "");
// Test that all characters we shouldn't encode ([33-36],[38,126]) are not.
nsCString noenc;
for (PRInt32 i = 33; i < 127; i++) {
if (i != 37) { // skip %
noenc.Append(static_cast<char>(i));
}
}
nsUrlClassifierUtils utils;
nsCString out;
utils.SpecialEncode(noenc, false, out);
CheckEquals(noenc, out);
// Test that all the chars that we should encode [0,32],37,[127,255] are
nsCString yesAsString, yesExpectedString;
for (PRInt32 i = 1; i < 256; i++) {
if (i < 33 || i == 37 || i > 126) {
yesAsString.Append(static_cast<char>(i));
yesExpectedString.Append('%');
yesExpectedString.Append(int_to_hex_digit(i / 16));
yesExpectedString.Append(int_to_hex_digit(i % 16));
}
}
out.Truncate();
utils.SpecialEncode(yesAsString, false, out);
CheckEquals(yesExpectedString, out);
TestEncodeHelper("blah//blah", "blah/blah");
}
// Make sure Unescape from nsEncode.h's unescape does what the server does.
void TestUnescape()
{
// test empty string
TestUnescapeHelper("\0", "\0");
// Test docoding of all characters.
nsCString allCharsEncoded, allCharsEncodedLowercase, allCharsAsString;
for (PRInt32 i = 1; i < 256; ++i) {
allCharsEncoded.Append('%');
allCharsEncoded.Append(int_to_hex_digit(i / 16));
allCharsEncoded.Append((int_to_hex_digit(i % 16)));
allCharsEncodedLowercase.Append('%');
allCharsEncodedLowercase.Append(tolower(int_to_hex_digit(i / 16)));
allCharsEncodedLowercase.Append(tolower(int_to_hex_digit(i % 16)));
allCharsAsString.Append(static_cast<char>(i));
}
nsUrlClassifierUtils utils;
nsCString out;
NS_UnescapeURL(allCharsEncoded.get(), allCharsEncoded.Length(), esc_AlwaysCopy, out);
CheckEquals(allCharsAsString, out);
out.Truncate();
NS_UnescapeURL(allCharsEncodedLowercase.get(), allCharsEncodedLowercase.Length(), esc_AlwaysCopy, out);
CheckEquals(allCharsAsString, out);
// Test %-related edge cases
TestUnescapeHelper("%", "%");
TestUnescapeHelper("%xx", "%xx");
TestUnescapeHelper("%%", "%%");
TestUnescapeHelper("%%%", "%%%");
TestUnescapeHelper("%%%%", "%%%%");
TestUnescapeHelper("%1", "%1");
TestUnescapeHelper("%1z", "%1z");
TestUnescapeHelper("a%1z", "a%1z");
TestUnescapeHelper("abc%d%e%fg%hij%klmno%", "abc%d%e%fg%hij%klmno%");
// A few more tests
TestUnescapeHelper("%25", "%");
TestUnescapeHelper("%25%32%35", "%25");
}
size_t
url_encode_cstr(
char *dest, const char*src, size_t max_size
) {
size_t ret = 0;
if(!max_size--)
return 0;
while(true) {
const char src_char = *src++;
if(src_char==0)
break;
if(max_size==0) {
ret++;
break;
}
if(isurlchar(src_char)) {
*dest++ = src_char;
ret++;
max_size--;
#if URL_ENCODE_SPACES_AS_PLUSES
} else if(src_char == ' ') {
*dest++ = '+'; // Stupid legacy space encoding.
ret++;
max_size--;
#endif
} else {
if(max_size < 3) {
// Dest buffer too small for the next character.
ret++;
break;
}
*dest++ = '%';
*dest++ = int_to_hex_digit(src_char >> 4);
*dest++ = int_to_hex_digit(src_char & 0xF);
ret += 3;
max_size -= 3;
}
}
*dest = 0;
return ret;
}
