CocoaDev

I use the following function in my application:

(header file can be found at: http://www.opendarwin.org/cgi-bin/cvsweb.cgi/src/CoreFoundation/StringEncodings.subproj/CFStringEncodingConverter.h?rev=1.1.1.3 )

extern UInt32 CFStringEncodingUnicodeToBytes( UInt32 encoding, UInt32 flags, const UniChar *characters, UInt32 numChars, UInt32 *usedCharLen, UInt8 *bytes, UInt32 maxByteLen, UInt32 *usedByteLen);

if (CFStringEncodingUnicodeToBytes(encoding, (1 « 6), &character, 1, &ucl, NULL, 0, &ubl) == 0) { // Some code } else { // Some other code }

It works just fine, but:

A) Is it a private header file(to me it doesn’t seem to be)? How to include it properly so it doesn’t give me a warning?

B) It’s undocumented?

C) Is there a better way to check if a character is available in the specified encoding?

– JP

Stuff the character into an NSString and the call -canBeConvertedToEncoding: on it. – Bo

Yeah, but thats not very efficent when you have to do that upto 4000 times.

– JP

I doubt that will be a problem; I benchmarked checking all 65536 possible values of the unichar type and it took under half a second on my computer. Checking a ‘mere’ 4000 values took 1/20 of a second. I’ve included the whole main function below so you can try it out on yours (just create a Foundation Tool project and paste it into main.m, replacing the default main function) – Bo

int main (int argc, const char * argv[]) { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; unichar c; NSString* charString; NSMutableString* validString = [[NSMutableString alloc] init]; NSMutableString* invalidString = [[NSMutableString alloc] init]; NSDate* startDate = [NSDate date]; for (c = 0; c < 65535; c++) { charString = [[NSString alloc] initWithCharacters:&c length:1]; BOOL valid = [charString canBeConvertedToEncoding:NSMacOSRomanStringEncoding]; if (valid) { [validString appendString:charString]; } else { [invalidString appendString:charString]; } [charString release]; } NSDate* stopDate = [NSDate date]; NSTimeInterval interval = [stopDate timeIntervalSinceDate:startDate]; NSLog(@”stop Date: %@\ntime to execute: %f secs”, stopDate, interval); NSLog(@”number of valid chars: %d\nnumber of invalid chars: %d”, [validString length], [invalidString length]); [pool release]; return 0; }

I tested both versions and:

A) COCOA_WAY

*time to execute: 0.166083 secs *number of valid chars: 257 *number of invalid chars: 65278

B) !COCOA_WAY

*time to execute: 0.021796 secs *number of valid chars: 257 *number of invalid chars: 65278

Conclusion: Cocoa version is about 8 times slower, BUT 0.166 sec isn’t very bad!!! – JP

#import <Foundation/Foundation.h> #include <CoreFoundation/CoreFoundation.h>

#define COCOA_WAY

int main (int argc, const char * argv[]) { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; unichar c; UInt32 valid = 0, invalid = 0; NSDate* startDate = [NSDate date];

    for (c = 0; c < 65535; c++)
    {
        #ifdef COCOA_WAY
            NSString *charString = [[NSString alloc] initWithCharacters:&c length:1];
            BOOL isValidChar = [charString canBeConvertedToEncoding:NSMacOSRomanStringEncoding];
            
            if(isValidChar) valid++;
            else invalid++;
            
            [charString release];
        #else
            UInt32 ucl, ubl;
            BOOL isNotValidChar = CFStringEncodingUnicodeToBytes(kCFStringEncodingMacRoman,  (1 << 6), &c, 1, &ucl, NULL, 0, &ubl);
            
            if(isNotValidChar) invalid++;
            else valid++;
        #endif
    }
    
    NSDate* stopDate = [NSDate date];
    NSTimeInterval interval = [stopDate timeIntervalSinceDate:startDate];
    NSLog(@"stop Date: %@\ntime to execute: %f secs", stopDate, interval);
    NSLog(@"number of valid chars: %d\nnumber of invalid chars: %d", valid, invalid);
[pool release];
return 0; }

Man. I do love my Tibook but it’s sure not the fastest boat in the pond. The other way to do this would be to create a buffer with all the characters in the encoding, make an NSString from it, create an NSCharacterSet using the +characterSetWithCharactersInString: method and then just testing for membership in the character set. Obviously, this would only be easy for fixed-length encodings like ASCII, ISO Latin-1 and Mac Roman, but it would probably be significantly faster. – Bo

Well, maybe, but the following is working fine. THX!: – JP

static inline bool IsUniCharAvailable(NSStringEncoding encoding, UniChar ch) { NSString *charString = [[NSString allocWithZone:NULL] initWithCharacters:&ch length:1]; BOOL isValid = [charString canBeConvertedToEncoding:encoding]; [charString release]; return isValid; }