深入理解苹果系统（Unicode）字符串的排序方法

Localized string comparisons are based on the Unicode Collation Algorithm, as tailored for different languages by CLDR (Common Locale Data Repository). Both are projects of the Unicode Consortium. Unicode is a registered trademark of Unicode, Inc.

Collation is the general term for the process and function of determining the sorting order of strings of characters.

NSArray *rawArray = @[@"爱你", @"一生一世",@"㊀", @"上",@"㊤",@"μ",@"язык",@"..",@"123",@"@",@"AA",@"abc",@"abb"];
//1. 默认排序方式
NSArray *defaultedSortedArray = [rawArray sortedArrayUsingComparator:^NSComparisonResult(id  _Nonnull obj1, id  _Nonnull obj2) {
    return [obj1 compare:obj2 options:NSCaseInsensitiveSearch];
}];
__block NSMutableArray *codeUnits = [NSMutableArray array];
[defaultedSortedArray enumerateObjectsUsingBlock:^(NSString*  _Nonnull obj, NSUInteger idx, BOOL * _Nonnull stop) {
    [codeUnits addObject:@([obj characterAtIndex:0])];
}];
NSLog(@"默认Unicode码点值排序 %@ 对应的各个字符串的首字符码点值是 %@", [defaultedSortedArray descriptionWithLocale:cnLocal], codeUnits);

排序结果 
 ..  123  @  AA  abb  abc  μ  язык  ㊀  ㊤  一生一世  上  爱你 
 对应的各个字符串的首字符码点值是 
 46  49  64  65  97  97  956  1103  12928  12964  19968  19978  29233

0000  ; [.0000.0000.0000] # <NULL>
0001  ; [.0000.0000.0000] # <START OF HEADING>
0002  ; [.0000.0000.0000] # <START OF TEXT>
0003  ; [.0000.0000.0000] # <END OF TEXT>

NSLocale *enLocale = [[NSLocale alloc] initWithLocaleIdentifier:@"en"];
defaultedSortedArray = [rawArray sortedArrayUsingComparator:^NSComparisonResult(NSString*  _Nonnull obj1, NSString*  _Nonnull obj2) {
    return [obj1 compare:obj2 options:0 range:NSMakeRange(0, obj1.length) locale:enLocale];
}];
NSLog(@"默认排序规则或者指定地区为locale后的排序结果是 %@", [defaultedSortedArray descriptionWithLocale:cnLocal]);

默认排序规则或者指定地区为en后的排序结果是 
 ..  （ch  (en  @  123  AA  abb  abc  μ  язык  ㊀  一生一世  上  ㊤  爱你

默认排序规则或者指定地区为zh_CN后的排序结果是 
 ..  （ch  (en  @  0124  123  艾你  爱你  産  上  ㊤  ㊀  一生一世  重逢  重要  aa  AA  abb  μ  язык 
 
 默认排序规则或者指定地区为ru_CN后的排序结果是 
 ..  （ch  (en  @  0124  123  язык  aa  AA  abb  μ  ㊀  一生一世  上  ㊤  爱你  産  艾你  重要  重逢

public func localizedStandardCompare(_ string: String) -> ComparisonResult {
    return compare(string, options: [.caseInsensitive, .numeric, .widthInsensitive, .forcedOrdering], range: NSRange(location: 0, length: length), locale: Locale.current._bridgeToObjectiveC())
}

NSCaseInsensitiveSearch  //大小写不敏感
NSNumericSearch //对字符串中出现的数字字符进行数字化的大小比较，比如Foo2.txt < Foo7.txt < Foo25.txt
NSWidthInsensitiveSearch //忽略宽度，按照实际表示的意思来对比，如'a' = UFF41
NSForcedOrderingSearch //强制返回Ascending或者Descending，和NSCaseInsensitiveSearch结合起来就是例如"aaa" > "AAA"

if (numerically && ((0 == strBuf1Len) && (str1Char <= '9') && (str1Char >= '0')) && ((0 == strBuf2Len) && (str2Char <= '9') && (str2Char >= '0'))) { // If both are not ASCII digits, then don't do numerical comparison here
    uint64_t intValue1 = 0, intValue2 = 0;	// !!! Doesn't work if numbers are > max uint64_t
    CFIndex str1NumRangeIndex = str1Index;
    CFIndex str2NumRangeIndex = str2Index;

    do {
        intValue1 = (intValue1 * 10) + (str1Char - '0');
        str1Char = CFStringGetCharacterFromInlineBuffer(&inlineBuf1, ++str1Index);
    } while ((str1Char <= '9') && (str1Char >= '0'));

    do {
        intValue2 = intValue2 * 10 + (str2Char - '0');
        str2Char = CFStringGetCharacterFromInlineBuffer(&inlineBuf2, ++str2Index);
    } while ((str2Char <= '9') && (str2Char >= '0'));

    if (intValue1 == intValue2) {
        if (forceOrdering && (kCFCompareEqualTo == compareResult) && ((str1Index - str1NumRangeIndex) != (str2Index - str2NumRangeIndex))) {
            compareResult = (((str1Index - str1NumRangeIndex) < (str2Index - str2NumRangeIndex)) ? kCFCompareLessThan : kCFCompareGreaterThan);
            numericEquivalence = true;
            forcedIndex1 = str1NumRangeIndex;
            forcedIndex2 = str2NumRangeIndex;
        }

        continue;
    } else if (intValue1 < intValue2) {
        if (freeLocale && locale) {
            CFRelease(locale);
        }
            return kCFCompareLessThan;
        } else {
            if (freeLocale && locale) {
                CFRelease(locale);
            }
        return kCFCompareGreaterThan;
    }
}