CFString​Transform

Written by Mattt Thompson on

There are two indicators that tell you everything you need to know about how nice a language is to use:

  1. API Consistency
  2. Quality of String Implementation

NSString is the crown jewel of Foundation. In an age where other languages still struggle to handle Unicode correctly, NSString is especially impressive. Not content to just work with whatever is thrown at it, NSString can parse strings into linguistic tags, determine the dominant language of the content, and convert between every string encoding imaginable. It's unfairly good.

But as powerful as NSString / NSMutableString are, one would be remiss not to mention their toll-free bridged cousin, CFMutableString—or more specifically, CFStringTransform.

As denoted by the CF prefix, CFStringTransform is part of Core Foundation. The function takes the following arguments, and returns a Boolean for whether or not the transform was successful:

  • string: The string to be transformed. Since this argument is a CFMutableStringRef, an NSMutableString can be passed using toll-free bridging cast.
  • range: The range of the string over which the transformation should be applied. This argument is a CFRange, rather than an NSRange value.
  • transform: The transformation to apply. This argument takes an ICU transform string, including any one of the string constants described below.
  • reverse: Whether to run the transformation in reverse, where applicable.

CFStringTransform covers a lot of ground with its transform argument. Here's a rundown of what it can do:

Strip Accents and Diacritics

Énġlišh långuãge lẳcks iñterêßţing diaçrïtičş. As such, it can be useful to normalize extended Latin characters into ASCII-friendly representations. Rid any string of its squiggly bits using the kCFStringTransformStripCombiningMarks transformation.

Name Unicode Characters

kCFStringTransformToUnicodeName allows you to determine the Unicode standard name for special characters, including Emoji. For instance, "🐑💨✨" is transformed into "{SHEEP} {DASH SYMBOL} {SPARKLES}", and "🐷" becomes "{PIG FACE}".

Transliterate Between Orthographies

With the notable exception of English (and its delightful spelling inconsistencies), writing systems generally encode speech sounds into a consistent written representation. European languages generally use the Latin alphabet (with a few added diacritics), Russian uses Cyrillic, Japanese uses Hiragana & Katakana, and Thai, Korean, & Arabic each have their own scripts.

Although each language has a particular inventory of sounds, some of which other languages may lack, the overlap across all of the major writing systems is remarkably high—enough so that one can rather effectively transliterate (not to be confused with translation) from one script to another.

CFStringTransform can transliterate back and forth between Latin and Arabic, Cyrillic, Greek, Korean (Hangul), Hebrew, Japanese (Hiragana & Katakana), Mandarin Chinese, and Thai.

Transformation Input Output
kCFStringTransformLatinArabic mrḥbạ مرحبا
kCFStringTransformLatinCyrillic privet привет
kCFStringTransformLatinGreek geiá sou γειά σου
kCFStringTransformLatinHangul annyeonghaseyo 안녕하세요
kCFStringTransformLatinHebrew şlwm שלום
kCFStringTransformLatinHiragana hiragana ひらがな
kCFStringTransformLatinKatakana katakana カタカナ
kCFStringTransformLatinThai s̄wạs̄dī สวัสดี
kCFStringTransformHiraganaKatakana にほんご ニホンゴ
kCFStringTransformMandarinLatin 中文 zhōng wén

And that's only using the constants defined in Core Foundation! By passing an ICU transform directly, CFStringTransform can transliterate between Latin and Arabic, Armenian, Bopomofo, Cyrillic, Georgian, Greek, Han, Hangul, Hebrew, Hiragana, Indic ( Devanagari, Gujarati, Gurmukhi, Kannada, Malayalam, Oriya, Tamil, & Telegu), Jamo, Katakana, Syriac, Thaana, & Thai.

Normalize User-Generated Content

One of the more practical applications for string transformation is to normalize unpredictable user input. Even if your application doesn't specifically deal with other languages, you should be able to intelligently process anything the user types into your app.

For example, let's say you want to build a searchable index of movies on the device, which includes greetings from around the world:

var mutableString = NSMutableString(string: "Hello! こんにちは! สวัสดี! مرحبا! 您好!") as CFMutableStringRef
  • First, apply the kCFStringTransformToLatin transform to transliterate all non-English text into a Latin alphabetic representation.
CFStringTransform(mutableString, nil, kCFStringTransformToLatin, Boolean(0))

Hello! こんにちは! สวัสดี! مرحبا! 您好! → Hello! kon'nichiha! s̄wạs̄dī! mrḥbạ! nín hǎo!

  • Next, apply the kCFStringTransformStripCombiningMarks transform to remove any diacritics or accents.
CFStringTransform(mutableString, nil, kCFStringTransformStripCombiningMarks, Boolean(0))

Hello! kon'nichiha! s̄wạs̄dī! mrḥbạ! nín hǎo! → Hello! kon'nichiha! swasdi! mrhba! nin hao!

  • Finally, downcase the text with CFStringLowercase, and split the text into tokens with CFStringTokenizer to use as an index for the text.
let tokenizer = CFStringTokenizerCreate(nil, mutableString, CFRangeMake(0, CFStringGetLength(mutableString)), 0, CFLocaleCopyCurrent())

var mutableTokens: [String] = []
var type: CFStringTokenizerTokenType
do {
    type = CFStringTokenizerAdvanceToNextToken(tokenizer)
    let range = CFStringTokenizerGetCurrentTokenRange(tokenizer)
    let token = CFStringCreateWithSubstring(nil, mutableString, range) as NSString
    mutableTokens.append(token)
} while type != .None

(hello, kon'nichiha, swasdi, mrhba, nin, hao)

By applying the same set of transformations on search text entered by the user, you have a universal way to search regardless of either the language of the search string or content!


CFStringTransform can be an insanely powerful way to bend language to your will. And it's but one of many powerful features that await you if you're brave enough to explore outside of Objective-C's warm OO embrace.