How can I escape unicode characters in a NSString?

Question

When I store a NSString inside some NSDictionary and log that dictionary to the console like this:

NSString *someString = @\"Münster\";  
NSDictionary *som         


        

Answer 1

The problem could be solved using a loop on a UniChar-string representation of the given string. Implemented as extension on NSString it would look something like this:

- (NSString *) escapedUnicode  
{  
    NSMutableString *uniString = [ [ NSMutableString alloc ] init ];  
    UniChar *uniBuffer = (UniChar *) malloc ( sizeof(UniChar) * [ self length ] );  
    CFRange stringRange = CFRangeMake ( 0, [ self length ] );  

    CFStringGetCharacters ( (CFStringRef)self, stringRange, uniBuffer );  

    for ( int i = 0; i < [ self length ]; i++ ) {  
        if ( uniBuffer[i] > 0x7e )  
            [ uniString appendFormat: @"\\u%04x", uniBuffer[i] ];  
        else  
            [ uniString appendFormat: @"%c", uniBuffer[i] ];  
    }  

    free ( uniBuffer );  

    NSString *retString = [ NSString stringWithString: uniString ];  
    [ uniString release ];  

    return retString;  
}

Answer 2

NSDictionary *someDict = [ NSDictionary dictionaryWithObjectsAndKeys: 
    someString, @"thestring" ];

Don't forget the nil sentinel. ;)

The console output looks like this:

unicode_test[3621:903] someDict:
{
    thestring = "M\U00fcnster";
}

with the string's unicode character escaped.

They're all Unicode characters.

Is there any method to convert a NSString to this escaped representation?

That's the dictionary (or some private method of NSPropertyListSerialization or private function of CFPropertyList) doing that, not the string. The \U sequence in that output is part of the OpenStep plist format. If you output the plist as XML using NSPropertyListSerialization, you'll find the ü (currently) encoded as naked UTF-8.

As far as I know, there is no built-in method, public or private, that will do the same escaping for you on a string alone. The closest thing is the strvis function, but that works byte-by-byte; it doesn't understand Unicode or UTFs.