HTML to NSAttributedString

gistfile1.m

//
//  NSAttributedString+fromHTML.m
//  Berlingske
//
//  Created by Taras Kalapun on 18.08.10.
//  Copyright (c) 2010 Ciklum. All rights reserved.
//

#import "NSAttributedString+fromHTML.h"

@implementation NSAttributedString (fromHTML)


static BOOL underlineFlag = NO;
static BOOL boldFlag = NO;
static BOOL italicFlag = NO;

static CGFloat size = 18.0;
static NSString *linkString;

- (void)setCurrentAttributes:(NSMutableAttributedString *)attrString;
{
    
    UIFont *nFont  = [UIFont fontWithName:@"LucidaBright" size:size];
    UIFont *bFont  = [UIFont fontWithName:@"LucidaBright-Demi" size:size];
    UIFont *iFont  = [UIFont fontWithName:@"LucidaBright-Demi" size:size];
    UIFont *biFont = [UIFont fontWithName:@"LucidaBright-Demi" size:size];
    
    NSRange range;
    
    range.location = 0;
    range.length = [attrString length];
    
    UIFont *font;
    
    //NSLog(@"setting attributes - b:%d, i:%d, u:%d", boldFlag, italicFlag, underlineFlag);
    
    if (boldFlag && italicFlag) {
        font = biFont;
    } else if (boldFlag) {
        font = bFont;
    } else if (italicFlag) {
        font = iFont;
    } else {
        font = nFont;
    }
    
    //font = [font fontWithSize:size];
    
    CTFontRef ctFont = CTFontCreateWithName((CFStringRef)font.fontName, font.pointSize, NULL);
    /*
     [attrString addAttribute:(NSString *)kCTFontAttributeName
     value:(id)ctFont
     range:range];
     */
    
    NSMutableDictionary *attrDict;
    
    attrDict = [NSMutableDictionary dictionaryWithCapacity:0];
    [attrDict setObject:(id)ctFont forKey:(NSString *)kCTFontAttributeName];
    [attrDict setObject:[NSNumber numberWithBool:underlineFlag] forKey:(NSString *)kCTUnderlineStyleAttributeName];
    //if (linkString) [attrDict setObject:linkString forKey:NSLinkAttributeName];
    
    [attrString addAttributes:attrDict range:range];
}

- (NSAttributedString *)initWithHTML:(NSString *)htmlString;
{
    //NSLog(@"initWithHTML: %@", htmlString);
    
    NSMutableAttributedString *attributedString;
    NSRange range;
    NSScanner *htmlScanner;
    NSMutableString *strippedString;
    
    boldFlag = NO;
    italicFlag = NO;
    underlineFlag = NO;
    
    NSDictionary *keywordDictionary = [[NSDictionary alloc] initWithObjectsAndKeys:
                                       @"&quot", @"\"",
                                       @"&amp", @"&",
                                       @"&lt", @"<",
                                       @"&gt", @">",
                                       nil];
    
    strippedString = [NSMutableString stringWithCapacity:[htmlString length]];
    htmlScanner = [NSScanner scannerWithString:htmlString];
    while (![htmlScanner isAtEnd]) {
        NSString *token;
        
        [htmlScanner scanUpToCharactersFromSet:[NSCharacterSet whitespaceAndNewlineCharacterSet] intoString:&token];
        [htmlScanner scanCharactersFromSet:[NSCharacterSet whitespaceAndNewlineCharacterSet] intoString:NULL];
        [strippedString appendFormat:@"%@ ", token];
    }
    
    attributedString = [[[NSMutableAttributedString alloc] init] autorelease];
    
    htmlScanner = [NSScanner scannerWithString:strippedString];
    while (![htmlScanner isAtEnd]) {
        NSString *firstPass, *tagString, *newString;
        NSScanner *keywordScanner;
        NSCharacterSet *openTagSet;
        NSCharacterSet *closeTagSet;
        NSMutableAttributedString *newAttributedString;
        
        openTagSet = [NSCharacterSet characterSetWithCharactersInString:@"<"];
        closeTagSet = [NSCharacterSet characterSetWithCharactersInString:@">"];
        newAttributedString = [[[NSMutableAttributedString alloc] init] autorelease];
        
        if ([htmlScanner scanUpToCharactersFromSet:openTagSet intoString:&firstPass]) {
            keywordScanner = [NSScanner scannerWithString:firstPass];
            while (![keywordScanner isAtEnd]) {
                NSString *keyword = nil;
                BOOL knownTag = NO;
                NSCharacterSet *keywordTag;
                NSEnumerator *keyEnum;
                
                keywordTag = [NSCharacterSet characterSetWithCharactersInString:@"&"];
                keyEnum = [[keywordDictionary allKeys] objectEnumerator];
                [keywordScanner scanUpToCharactersFromSet:keywordTag intoString:&newString];
                [newAttributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:newString]autorelease]];
                
                while (![keywordScanner isAtEnd] && (keyword = [keyEnum nextObject]))
                    if ([keywordScanner scanString:keyword intoString:NULL]) {
                        [newAttributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:[keywordDictionary objectForKey:keyword]]autorelease]];
                        knownTag = YES;
                    }
                if (!knownTag && [keywordScanner scanCharactersFromSet:keywordTag intoString:&newString])
                    [newAttributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:newString]autorelease]];
            }
            
            [self setCurrentAttributes:newAttributedString];
            [attributedString appendAttributedString:newAttributedString];
        }
        
        // Either we hit a '<' or we're at the end of the text.
        if (![htmlScanner isAtEnd]) {
            [htmlScanner scanCharactersFromSet:openTagSet intoString:NULL];
            [htmlScanner scanUpToCharactersFromSet:closeTagSet intoString:&tagString];
            [htmlScanner scanCharactersFromSet:closeTagSet intoString:NULL];
            tagString = [tagString lowercaseString];
            
            
            if ([tagString hasPrefix:@"a "]) {
                range = [tagString rangeOfString:@"\""];
                
                if (range.length) {
                    tagString = [tagString substringFromIndex:NSMaxRange(range)];
                    range = [tagString rangeOfString:@"\""];
                    linkString = [tagString substringToIndex:range.location];
                } else if ((range = [tagString rangeOfString:@"="]).length) {
                    linkString = [tagString substringFromIndex:NSMaxRange(range)];
                } else
                    linkString = nil;
                [linkString retain];
            } else if ([tagString isEqual:@"/a"]) {
                [linkString release];
                linkString = nil;
            } else if ([tagString isEqual:@"b"] || [tagString isEqual:@"strong"])
                boldFlag = YES;
            else if ([tagString isEqual:@"/b"] || [tagString isEqual:@"/strong"])
                boldFlag = NO;
            else if ([tagString isEqual:@"h2"] || [tagString isEqual:@"h3"]) {
                boldFlag = YES;
                [attributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:@"\n\n"]autorelease]];
            } else if ([tagString isEqual:@"/h2"] || [tagString isEqual:@"/h3"]) {
                boldFlag = NO;
                [attributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:@"\n\n"]autorelease]];
            } else if ([tagString isEqual:@"i"] || [tagString isEqual:@"em"] || [tagString isEqual:@"cite"])
                italicFlag = YES;
            else if ([tagString isEqual:@"/i"] || [tagString isEqual:@"/em"] || [tagString isEqual:@"/cite"])
                italicFlag = NO;
            else if ([tagString isEqual:@"u"])
                underlineFlag = YES;
            else if ([tagString isEqual:@"/u"])
                underlineFlag = NO;
            else if ([tagString isEqual:@"p"])
                [attributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:@"\n\n"]autorelease]];
            else if ([tagString isEqual:@"br"] || [tagString isEqual:@"br /"])
                [attributedString appendAttributedString:[[[NSAttributedString alloc] initWithString:@"\n"]autorelease]];
            else if ([tagString isEqual:@"/font"])
                size = 12.0;
            else if ([tagString hasPrefix:@"font size="] || [tagString hasPrefix:@"font size ="]) {
                float foo;
                if (sscanf([tagString UTF8String], "font size = +%f", &foo) == 1)
                    size += foo + 9;
                else if (sscanf([tagString UTF8String], "font size = %f", &foo) == 1)
                    size = foo + 9;
            }
        }
    }
    
    range.location = 0;
    range.length = [attributedString length];
    
    [keywordDictionary release];
    
    return [self initWithAttributedString:attributedString];
    //return attributedString;
}



@end