Remove Emoji in NSString

gistfile1.m

// XCode 4.2.1

@implementation NSString(EmojiExtension)

- (NSString*)removeEmoji {
    __block NSMutableString* temp = [NSMutableString string];
    
    [self enumerateSubstringsInRange: NSMakeRange(0, [self length]) options:NSStringEnumerationByComposedCharacterSequences usingBlock:
     ^(NSString *substring, NSRange substringRange, NSRange enclosingRange, BOOL *stop){
         
         const unichar hs = [substring characterAtIndex: 0];
         
         // surrogate pair
         if (0xd800 <= hs && hs <= 0xdbff) {
             const unichar ls = [substring characterAtIndex: 1];
             const int uc = ((hs - 0xd800) * 0x400) + (ls - 0xdc00) + 0x10000;
             
             [temp appendString: (0x1d000 <= uc && uc <= 0x1f77f)? @"": substring]; // U+1D000-1F77F
             
         // non surrogate
         } else {
             [temp appendString: (0x2100 <= hs && hs <= 0x26ff)? @"": substring]; // U+2100-26FF
         }
     }];
    
    return temp;
}

@end

'Measuring length of a string' at the Apple docs https://developer.apple.com/documentation/swift/string brought me to another solution, without the need for knowledge about the unicode pages. I just want letters to to remain in the string and skip all that is an icon:

#include <string.h>

inline static NSString* _Nonnull nsstring_remove_emoji_v2(NSString* const _Nonnull origString) {
    NSMutableString* const result = [NSMutableString stringWithCapacity:0];
    NSUInteger const len = origString.length;
    NSString* subStr;
    for (NSUInteger index = 0; index < len; index++) {
        subStr = [origString substringWithRange:NSMakeRange(index, 1)];
        const char* utf8Rep = subStr.UTF8String;  // will return NULL for icons that consist of 2 chars
        if (utf8Rep != NULL) {
            unsigned long const length = strlen(utf8Rep);
            if (length <= 2) {
                [result appendString:subStr];
            }
        }
    }
    return result.copy;
}

I have no clue, what this does with chinese or japanese text. But it works for all german letters.

I have no clue, what this does with chinese or japanese text. But it works for all german letters.

Came across this gist, and I happen to have strings with Chinese + emoji, this code will remove all Chinese character due to there strlen are 3 :)

Much simpler way is to use a string transform: -- this will move all emoji code points, and preserve non-latin characters, accents etc

Eg
[@"🤯!!! ক❤️testé᏷🧡💚💛せぬ❤️‍🔥👩🏿‍🦰" stringByApplyingTransform: @"[:emoji:] remove" reverse:NO]

returns
!!! ক️testé᏷せぬ️‍‍

just to followup. -- apparently the [:emoji:] property used in the ICU transform includes digits, some punctuation, other things not generally though to be emoji.

I am finding this method on an NSString category working better

- (NSString *)stringByRemovingEmoji {
    static NSRegularExpression * regex = nil;
    static dispatch_once_t onceToken;
    dispatch_once(&onceToken, ^{
        // remove all emoji less those that are digits, punctuation, letters, latin 1 supplement or letter like symbols
        // or BIDI Non-Spacing Mark
        NSError * error = nil;
        regex = [NSRegularExpression regularExpressionWithPattern:@"([[:emoji:]--[:digit:]--[:punctuation:]--[:letter:]--[:block=Latin-1_sup:]--[:block=letter-like-symbols:]]|\\uFE0F)" options: 0 error:&error];
        if (error){
            NSLog(@"Error forming regex");
        }
    });
    
    return [regex stringByReplacingMatchesInString:self options:0 range:NSMakeRange(0, self.length) withTemplate:@""];
}