encoding string for URL

Question

I have to POST data to a php page from an Iphone application, so i need to encode the parameters properly, converting the special characters...

Specifically i need to send the UDID of the iphone.

I've found many helps on the web to encode the String to pass as parameter, but i got a strange error.

I'm using this function:

- (NSString *)URLEncodeString:(NSString *)string { 

NSString *result =
   (NSString*)CFURLCreateStringByAddingPercentEscapes(kCFAllocatorDefault, 
    (CFStringRef)string, NULL, CFSTR("'\"?=&+<>;:-"), kCFStringEncodingUTF8);

return [result autorelease];}

it seems correct but when i use it, the result isn't what i expect for.

This is the code:

UIDevice *device = [UIDevice currentDevice];
NSString *uniqueIdentifier = [device uniqueIdentifier];

NSLog(uniqueIdentifier);
NSLog([self URLEncodeString:uniqueIdentifier]);

and this is the Log generated when i hit this code:

2010-06-23 21:58:51.671 provaNavigation[2343:20b] 00000000-0000-1000-8000-0013775CE6D2
2010-06-23 21:58:51.672 provaNavigation[2343:20b] 0000000010003900009080176010001273086780001283520013775CE6D2

and again:

2010-06-23 21:59:25.614 provaNavigation[2343:20b] 00000000-0000-1000-8000-0013775CE6D2
2010-06-23 21:59:25.615 provaNavigation[2343:20b] 000000001000390000908192801000-18000912258750013775CE6D2

and again:

2010-06-23 21:59:40.848 provaNavigation[2343:20b] 00000000-0000-1000-8000-0013775CE6D2
2010-06-23 21:59:40.849 provaNavigation[2343:20b] 000000001000390000908866081000-18000912239630013775CE6D2

and again...

I get a different value every time and never correct with the right %number.

Answer 1

I've come across problems URL encoding before, so I ended up writing a category on NSString to properly encode URLs:

- (NSString *) URLEncodedString {
  NSMutableString * output = [NSMutableString string];
  const char * source = [self UTF8String];
  int sourceLen = strlen(source);
  for (int i = 0; i < sourceLen; ++i) {
    const unsigned char thisChar = (const unsigned char)source[i];
    if (thisChar == ' '){
      [output appendString:@"+"];
    } else if (thisChar == '.' || thisChar == '-' || thisChar == '_' || thisChar == '~' || 
           (thisChar >= 'a' && thisChar <= 'z') ||
           (thisChar >= 'A' && thisChar <= 'Z') ||
           (thisChar >= '0' && thisChar <= '9')) {
      [output appendFormat:@"%c", thisChar];
    } else {
      [output appendFormat:@"%%%02X", thisChar];
    }
  }
  return output;
}

- (NSString *) URLDecodedString {
  NSString * spacedDecoded = [self stringByReplacingOccurrencesOfString:@"+" withString:@" "];
  return [spacedDecoded stringByReplacingPercentEscapesUsingEncoding:NSUTF8StringEncoding];
}

There are a couple notes about this:

1. This will encode spaces as "+"
2. This encodes everything except a-z (upper and lowercase), numbers, period, hyphen, underscore, and tilde

3. This will properly encode unicode characters as multiple octets. For example, I can do:

   NSString * enc = [@"http://example.com/?data=÷¡¿" URLEncodedString];
   NSLog(@"%@", enc);
   NSLog(@"%@", [enc URLDecodedString]);

   Which will properly log:

   Testing.m:65 main() http%3A%2F%2Fexample.com%2F%3Fdata%3D%C3%B7%C2%A1%C2%BF
   Testing.m:66 main() http://example.com/?data=÷¡¿

EDIT I copied and pasted it into an empty project, like this:

NSString * string = @"00000000-0000-1000-8000-0013775CE6D2";
NSString *result = (NSString*)CFURLCreateStringByAddingPercentEscapes(kCFAllocatorDefault, (CFStringRef)string, NULL, CFSTR("'\"?=&+<>;:-"), kCFStringEncodingUTF8);
NSLog(@"encode? %@", result);
result = [result stringByReplacingPercentEscapesUsingEncoding:NSUTF8StringEncoding];
NSLog(@"decode? %@", result);
NSLog(@"equal? %d", [string isEqual:result]);

And it logs:

Testing[62541:a0f] encode? 00000000%2D0000%2D1000%2D8000%2D0013775CE6D2
Testing[62541:a0f] decode? 00000000-0000-1000-8000-0013775CE6D2
Testing[62541:a0f] equal? 1

The reason your code isn't working is because you're logging the string as the format string, which means it's seeing the %2 stuff and trying to replace it with other things. Observe:

NSString * string = @"00000000-0000-1000-8000-0013775CE6D2";
NSString *result = (NSString*)CFURLCreateStringByAddingPercentEscapes(kCFAllocatorDefault, (CFStringRef)string, NULL, CFSTR("'\"?=&+<>;:-"), kCFStringEncodingUTF8);
NSLog(@"%@", result);
NSLog(result);

Logs:

Testing[62585:a0f] 00000000%2D0000%2D1000%2D8000%2D0013775CE6D2
Testing[62585:a0f] 00000000879923200001897296640100041968968000 00013775CE6D2

So the string is getting properly encoded, but you're just logging it incorrectly.