VCTitleCase.m

//
//  VCTitleCase.m
//  Title Case extension for NSString
//
//  Based on titlecase.pl by:
//    John Gruber
//    http://daringfireball.net/
//    10 May 2008
//
//  Cocoa Foundation version by:
//    Marshall Elfstrand
//    http://vengefulcow.com/
//    24 May 2008
//
//  License: http://www.opensource.org/licenses/mit-license.php
//

#import "VCTitleCase.h"

NSString * VCTitleCaseString(NSString *input) {
    static NSArray *shortWords;
    static NSMutableCharacterSet *wordStartCharacterSet;
    static NSMutableCharacterSet *wordMiddleCharacterSet;
    static NSMutableCharacterSet *wordEndCharacterSet;
    static NSMutableCharacterSet *wordIgnoreCharacterSet;
    
    // Initialize the list of "short" words that remain lowercase.
    if (!shortWords) {
        shortWords = [[NSArray alloc] initWithObjects:
            @"a", @"an", @"and", @"as", @"at", @"but", @"by", @"en", @"for",
            @"if", @"in", @"of", @"on", @"or", @"the", @"to", @"v", @"via",
            @"vs", nil];
    }
    
    // Initialize the set of characters allowed at the start of words.
    if (!wordStartCharacterSet) {
        wordStartCharacterSet = [[NSCharacterSet uppercaseLetterCharacterSet] mutableCopy];
        [wordStartCharacterSet formUnionWithCharacterSet:[NSCharacterSet lowercaseLetterCharacterSet]];
    }
    
    // Initialize the set of characters allowed in the middle of words.
    if (!wordMiddleCharacterSet) {
        wordMiddleCharacterSet = [[NSCharacterSet uppercaseLetterCharacterSet] mutableCopy];
        [wordMiddleCharacterSet formUnionWithCharacterSet:[NSCharacterSet lowercaseLetterCharacterSet]];
        [wordMiddleCharacterSet addCharactersInString:@".&'’"];
    }
    
    // Initialize the set of characters allowed at the end of words.
    if (!wordEndCharacterSet) wordEndCharacterSet = wordStartCharacterSet;
    
    // Initialize the set of characters that cause a word to be ignored
    // when they appear in the middle.
    if (!wordIgnoreCharacterSet) {
        wordIgnoreCharacterSet = [[NSCharacterSet uppercaseLetterCharacterSet] mutableCopy];
        [wordIgnoreCharacterSet addCharactersInString:@"."];
    }
    
    // Create a mutable copy of the string that we can modify in-place.
    NSMutableString *newString = [input mutableCopy];

    // Create a scanner that we can use to locate words in the string.
    NSScanner *scanner = [NSScanner scannerWithString:input];
    [scanner setCaseSensitive:YES];

    // Begin scanning for words.
    NSRange currentRange = {};   // Range of word located by scanner
    NSString *word;              // Extracted word
    NSString *lowercaseWord;     // Lowercase version of extracted word
    NSRange ignoreTriggerRange;  // Range of character causing word to be ignored
    BOOL isFirstWord = YES;      // To determine whether to capitalize small word
    while (![scanner isAtEnd]) {
    
        // Locate the beginning of the next word.
        [scanner scanUpToCharactersFromSet:wordStartCharacterSet
                                intoString:NULL];
        if ([scanner scanLocation] >= [input length]) continue;  // No more words
        currentRange = NSMakeRange([scanner scanLocation], 1);

        // Check to see if we stopped on whitespace and advance to the
        // actual beginning of the word.
        if (![wordStartCharacterSet characterIsMember:[input characterAtIndex:[scanner scanLocation]]]) {
            [scanner setScanLocation:[scanner scanLocation] + 1];
            currentRange = NSMakeRange([scanner scanLocation], 1);
        }

        // Advance to the next character in the word.
        [scanner scanString:[input substringWithRange:currentRange]
                 intoString:NULL];
                 
        // See if the next character is a valid word character, and if so,
        // scan through the end of the word.
        if (input.length > scanner.scanLocation && [wordMiddleCharacterSet characterIsMember:[input characterAtIndex:[scanner scanLocation]]]) {
            [scanner scanCharactersFromSet:wordMiddleCharacterSet
                                intoString:NULL];
            currentRange.length = [scanner scanLocation] - currentRange.location;
        }
        
        // Back off the word until it ends with a valid character.
        unichar lastCharacter = [input characterAtIndex:(NSMaxRange(currentRange) - 1)];
        while (![wordEndCharacterSet characterIsMember:lastCharacter]) {
            [scanner setScanLocation:[scanner scanLocation] - 1];
            currentRange.length -= 1;
            lastCharacter = [input characterAtIndex:(NSMaxRange(currentRange) - 1)];
        }

        // We have now located a word.
        word = [input substringWithRange:currentRange];
        lowercaseWord = [word lowercaseString];

        // Check to see if the word needs to be capitalized.
        // Words that have dots in the middle or that already contain
        // capitalized letters in the middle (e.g. "iTunes") are ignored.
        ignoreTriggerRange = [input
            rangeOfCharacterFromSet:wordIgnoreCharacterSet
                            options:NSLiteralSearch
                              range:NSMakeRange(currentRange.location + 1, currentRange.length - 1)
        ];
        if (ignoreTriggerRange.location == NSNotFound) {
            if ([word rangeOfString:@"&"].location != NSNotFound) {
                // Uppercase words that contain ampersands.
                [newString replaceCharactersInRange:currentRange
                                         withString:[word uppercaseString]];
            } else {
                if ((!isFirstWord) && [shortWords containsObject:lowercaseWord]) {
                    // Lowercase small words.
                    [newString replaceCharactersInRange:currentRange
                                             withString:lowercaseWord];
                } else {
                    // Capitalize word.
                    [newString replaceCharactersInRange:currentRange
                                             withString:[word capitalizedString]];
                }
            }
        }

        isFirstWord = NO;
    }

    // Make sure the last word is capitalized, even if it is a small word.
    if (lowercaseWord && [shortWords containsObject:lowercaseWord]) {
        [newString replaceCharactersInRange:currentRange
                                 withString:[lowercaseWord capitalizedString]];
    }
    
    return [newString copy];
}