Strip down webpage to just text on iOS (Objective C) - html

Question

My main goal here is to achieve an effect like Readability or Safari's Reader service where the main content of the webpage is converted to text. I don't actually want to display any images, just get all the webpage's important text. I am currently using some pretty long self-built code to parse the webpage for the s to find out what the heading may look like, and I am also parsing the s that I hope contain the majority of the page's content.
-(void)interpretAndDisplay {
NSURL *URL = [NSURL URLWithString:self.url];
NSData *data = [NSData dataWithContentsOfURL:URL];
NSString *html = [NSString stringWithUTF8String:[data bytes]];
//Getting the H1s
NSMutableArray *h1Full = [[NSMutableArray alloc] init];
h1Full = [self stringsBetweenString:#"<h1" andString:#">" andText:html];
if ([h1Full count] > 0) {
NSMutableArray *h1Content = [[NSMutableArray alloc] init];
h1Content = [self stringsBetweenString:[NSString stringWithFormat:#"<h1%#>",[h1Full firstObject]] andString:#"</h1>" andText:html];
NSMutableArray *h1Sanitize = [[NSMutableArray alloc] init];
h1Sanitize = [self stringsBetweenString:#"<" andString:#">" andText:html];
if ([h1Content count] > 0) {
NSString *finalTitle = [h1Content firstObject];
for (int i = 0; i < [h1Sanitize count]; i++) {
NSString *toRemove = [NSString stringWithFormat:#"<%#>",[h1Sanitize objectAtIndex:i]];
finalTitle = [finalTitle stringByReplacingOccurrencesOfString:toRemove withString:#""];
finalTitle = [finalTitle stringByReplacingOccurrencesOfString:#"\n" withString:#""];
}
finalTitle = [self sanitizeString:finalTitle];
[self.titleLabel setText:finalTitle];
}
}
//Now for the body!
NSMutableArray *pTag = [[NSMutableArray alloc] init];
pTag = [self stringsBetweenString:#"<p" andString:#">" andText:html];
if ([pTag count] > 0) {
NSMutableArray *pContent = [[NSMutableArray alloc] init];
pContent = [self stringsBetweenString:[NSString stringWithFormat:#"<p%#>",[pTag firstObject]] andString:#"</p>" andText:html];
NSMutableArray *pSanitize = [[NSMutableArray alloc] init];
pSanitize = [self stringsBetweenString:#"<" andString:#">" andText:html];
if ([pContent count] > 0) {
for (int i = 0; i < [pContent count]; i++) {
NSString *pToEdit = [pContent objectAtIndex:i];
for (int i = 0; i < [pSanitize count]; i++) {
NSString *toRemove = [NSString stringWithFormat:#"<%#>",[pSanitize objectAtIndex:i]];
pToEdit = [pToEdit stringByReplacingOccurrencesOfString:toRemove withString:#""];
}
[pContent replaceObjectAtIndex:i withObject:pToEdit];
}
for (int i = 0; i < [pContent count]; i++) {
NSString *pToEdit = [pContent objectAtIndex:i];
pToEdit = [pToEdit stringByReplacingOccurrencesOfString:#"\n" withString:#""];
[pContent replaceObjectAtIndex:i withObject:pToEdit];
}
NSString *finalBody = #"";
for (int i = 0; i < [pContent count]; i++) {
if ([finalBody isEqualToString:#""]) {
finalBody = [NSString stringWithFormat:#"%#",[pContent objectAtIndex:i]];
}
else {
finalBody = [NSString stringWithFormat:#"%#\n\n%#",finalBody,[pContent objectAtIndex:i]];
}
}
finalBody = [self sanitizeString:finalBody];
[self.textLabel setText:finalBody];
}
}
}
The above code extracts all of the elements just fine and sanitizes them with a method that I created, but the problem is that just analyzing the P tags sometimes completely fails to simplify the content, and analyzing all possible content tags could mess with the content's order and layout.
Is there a better way or some framework that converts all the text into a nice string?
EDIT
Searching around, I found a Boilerpipe project that can extract text extremely easily (https://github.com/k-bx/boilerpipe/wiki/QuickStart). It looks as easy as this: String text= ArticleExtractor.INSTANCE.getText(url);
Can I do this on Objective C?
Edit 2
It appears that there is a boilerpipe API, but this has limited requests. I am mostly looking for a user-side solution.

Answer 1

Reggie is not the most tolerant approach in my opinion.
I'd try to find an existent open source (i.e. https://github.com/Kerrick/readability-js) and use WebKit to inject JS into a web page once loaded.
After that you can inject another JS, extracting the processed content (using the appropriate class from the source)
Then, using JavaScriptCore you can pass div's content to Objective-C (JS offers many ways of doing that)