How to unescape backslashes on macOS?

Question

Localized.strings files may contain escaped chars like \\n and \\" .

How do I efficiently unescape them?

I know I could write my own function that looks for a \\ and removes it and then keeps searching at the second-next character (so that I don't turn \\\\ into nothing), but that won't handle special escape methods such as an octal character numbers (eg \\012 for a LF) and possibly other forms.

I'd think that NSString would offer a function for that but I can't find any.

Actually, this appears to be a duplicate of NSString strip regular \\ escape characters how? , but that question has a lot of bad answers and not a single correct answer. And the user is not active any more, meaning no one will ever accept a correct answer there. How shall that be handled on SO?

Answer 1

Here's a self-written ObjC version that works for most cases. It defines an unescaped method for NSString.

#import <Cocoa/Cocoa.h>

@interface NSString (BackslashUnescaping)
- (NSString*) unescaped;
@end

@implementation NSString (BackslashUnescaping)

- (NSString*) unescaped
{
    NSString *text = self;
    NSInteger charIndex = 0;
    while (true) {
        NSInteger textLen = text.length;
        if (charIndex >= textLen) {
            break;
        }
        NSRange remainingRange = NSMakeRange (charIndex, textLen-charIndex);
        NSRange charRange = [text rangeOfString:@"\\"options:0 range:remainingRange];
        if (charRange.length == 0) {
            // no more backslashes -> done
            break;
        }
        charIndex = charRange.location + 1;
        if (charIndex >= textLen) {
            // reached end of string -> exit loop
            break;
        }
        // check char following the backslash
        unichar nextChar = [text characterAtIndex:charIndex];
        unichar replacementChar;
        NSInteger skipLen = 1;
        if (nextChar >= 'a' && nextChar <= 'z') {
            if (nextChar == 'n') {
                replacementChar = '\n'; // LF
            } else if (nextChar == 'r') {
                replacementChar = '\r'; // CR
            } else if (nextChar == 't') {
                replacementChar = '\t'; // TAB
            } else if (nextChar == 'x') {
                // A hex char code
                const NSInteger xtraLen = 2;
                if (charIndex+xtraLen >= textLen) break;
                // Note: Does not make sure that both chars are valid hex chars
                NSString *code = [text substringWithRange:NSMakeRange(charIndex+1, 2)];
                char ch = strtol(code.UTF8String, NULL, 16);
                replacementChar = ch;
                skipLen += xtraLen;
            } else if (nextChar == 'u') {
                // A unicode char code
                const NSInteger xtraLen = 4;
                if (charIndex+xtraLen >= textLen) break;
                // Note: Does not make sure that all four chars are valid hex chars
                NSString *code = [text substringWithRange:NSMakeRange(charIndex+1, 4)];
                unichar ch = strtol(code.UTF8String, NULL, 16);
                replacementChar = ch;
                skipLen += xtraLen;
            } else {
                // an unknown escape code - this should be fixed
                NSAssert(false, @"There's a case missing for escaping \\%c", nextChar);
            }
        } else if (nextChar >= '0' && nextChar <= '9') {
            unichar nextChar2 = 0;
            if (charIndex > textLen) {  // get the second octal char
                nextChar2 = [text characterAtIndex:charIndex+1];
            }
            if (nextChar == '0' && (nextChar2 < '0' || nextChar2 > '9')) {
                // A short NUL (\0) char
                replacementChar = 0;
            } else {
                // An octal char code
                const NSInteger xtraLen = 2;
                if (charIndex+xtraLen >= textLen) break;
                // Note: Does not make sure that the last char is a valid octal char
                NSString *code = [text substringWithRange:NSMakeRange(charIndex, 3)];
                char ch = strtol(code.UTF8String, NULL, 8); // https://stackoverflow.com/a/12820646/43615
                replacementChar = ch;
                skipLen += xtraLen;
            }
        } else {
            // Handle all generic escapes, like for \\ and \"
            replacementChar = nextChar;
        }
        #if 0 // Use string concatenation
            charIndex += skipLen-1;
            NSString *head = [text substringToIndex:charRange.location];
            NSString *tail = [text substringFromIndex:charIndex+1];
            text = [NSString stringWithFormat:@"%@%C%@", head, replacementChar, tail];
        #else // Use a mutable string
            if (text == self) {
                text = text.mutableCopy;
            }
            NSRange replacedRange = NSMakeRange(charRange.location, skipLen+1);
            NSString *replacement = [NSString stringWithCharacters:&replacementChar length:1];
            [(NSMutableString*)text replaceCharactersInRange:replacedRange withString:replacement];
            charIndex += 1;
        #endif 
    }
    return text;
}

@end 

int main(int argc, const char * argv[]) {
    @autoreleasepool {
        NSArray *testValues = @[
            @"CR:\\rLF:\\n",
            @"\\\"quoted\\\"",
            @"Backslash: \\\\",
            @"Octal x (\170):\\170",
            @"Hex x (\x78):\\x78",
            @"Unicode Ф (\u0424):\\u0424",
            @"NUL char:\\0.",
            @"Bad Hex:\\x7x",   // this is not detected being invalid
            @"Bad Hex:\\x7",
            @"Incomplete :\\13"
        ];
        for (NSString *s in testValues) {
            NSString *s2 = [NSString stringWithFormat:@"Escaped:   %@\nUnescaped: %@", s, s.unescaped];
            printf("\n%s\n", s2.UTF8String);
        }
    }
    return 0;
}