/// <summary>
/// Validate that a character is valid for a uri at a specific encoding level.
/// </summary>
/// <param name="ch">Character to check.</param>
/// <param name="level">Encoding level.</param>
/// <returns><see langword="True"/> if the character is valid for the uri.</returns>
public static bool IsValidCharInUri(char ch, UriEncoding level)
{
if((((ch >= 'a') && (ch <= 'z')) || ((ch >= 'A') && (ch <= 'Z'))) || ((ch >= '0') && (ch <= '9'))) {
return true;
}
// the following characters are always safe
switch(ch) {
case '\'':
case '(':
case ')':
case '*':
case '-':
case '.':
case '_':
case '!':
return true;
}
// based on encoding level, additional character may be safe
switch(level) {
case UriEncoding.Fragment:
// the following characters are safe when used in the fragment of a uri
switch(ch) {
case '#':
return true;
}
// all characters safe for UriEncoding.Query are also safe for UriEncoding.Fragment
goto case UriEncoding.Query;
case UriEncoding.Query:
// the following characters are safe when used in the query of a uri
switch(ch) {
case '/':
case ':':
case '~':
case '$':
case ',':
case ';':
// NOTE (steveb): we don't encode | characters
case '|':
// NOTE (steveb): don't decode '?', because it's handling is different on various web-servers (e.g. Apache vs. IIS)
// case '?':
return true;
}
// all characters safe for UriEncoding.Segment are also safe for UriEncoding.Query
goto case UriEncoding.Segment;
case UriEncoding.Segment:
// the following characters are safe when used in a segment of a uri
switch(ch) {
case '@':
// NOTE (steveb): we don't encode ^ characters
case '^':
return true;
}
break;
case UriEncoding.UserInfo:
// the following characters are safe when used in the UserInfo part of a uri
switch(ch) {
case '&':
case '=':
return true;
}
break;
}
return false;
}
/// <summary>
/// Uri encode a string.
/// </summary>
/// <param name="text">Input text.</param>
/// <param name="level">Encoding level.</param>
/// <returns>Encoded string.</returns>
public static string Encode(string text, UriEncoding level)
{
if(string.IsNullOrEmpty(text)) {
return text;
}
byte[] original = Encoding.UTF8.GetBytes(text);
// count how many characters are affected by the encoding
int charsToReplace = 0;
int charsToEncode = 0;
int length = original.Length;
for(int i = 0; i < length; i++) {
var ch = (char)original[i];
if(ch == ' ') {
charsToReplace++;
} else if(!IsValidCharInUri(ch, level)) {
charsToEncode++;
}
}
// check if any characters are affected
if((charsToReplace == 0) && (charsToEncode == 0)) {
return text;
}
// copy, replace, and encode characters
var encoded = new byte[length + (charsToEncode * 2)];
int index = 0;
for(int j = 0; j < length; j++) {
byte asciiByte = original[j];
char asciiChar = (char)asciiByte;
if(IsValidCharInUri(asciiChar, level)) {
encoded[index++] = asciiByte;
} else if(asciiChar == ' ') {
// replace ' ' with '+'
encoded[index++] = 0x2b; // '+'
} else {
// replace char with '%' + code
encoded[index++] = 0x25; // '%'
encoded[index++] = (byte)StringUtil.IntToHexChar((asciiByte >> 4) & 15);
encoded[index++] = (byte)StringUtil.IntToHexChar(asciiByte & 15);
}
}
return Encoding.ASCII.GetString(encoded);
}
