Reading HTML content from a UIWebView

To get the whole HTML raw data (with <head> and <body>):

NSString *html = [webView stringByEvaluatingJavaScriptFromString:@"document.documentElement.outerHTML"];

The second question is actually easier to answer. Look at the stringWithContentsOfURL:encoding:error: method of NSString - it lets you pass in a URL as an instance of NSURL (which can easily be instantiated from NSString) and returns a string with the complete contents of the page at that URL. For example:

NSString *googleString = @"http://www.google.com";
NSURL *googleURL = [NSURL URLWithString:googleString];
NSError *error;
NSString *googlePage = [NSString stringWithContentsOfURL:googleURL 
                                                encoding:NSASCIIStringEncoding
                                                   error:&error];

After running this code, googlePage will contain the HTML for www.google.com, and error will contain any errors encountered in the fetch. (You should check the contents of error after the fetch.)

Going the other way (from a UIWebView) is a bit trickier, but is basically the same concept. You'll have to pull the request from the view, then do the fetch as before:

NSURL *requestURL = [[yourWebView request] URL];
NSError *error;
NSString *page = [NSString stringWithContentsOfURL:requestURL 
                                          encoding:NSASCIIStringEncoding
                                             error:&error];

EDIT: Both these methods take a performance hit, however, since they do the request twice. You can get around this by grabbing the content from a currently-loaded UIWebView using its stringByEvaluatingJavascriptFromString: method, as such:

NSString *html = [yourWebView stringByEvaluatingJavaScriptFromString: 
                                         @"document.body.innerHTML"];

This will grab the current HTML contents of the view using the Document Object Model, parse the JavaScript, then give it to you as an NSString* of HTML.

Another way is to do your request programmatically first, then load the UIWebView from what you requested. Let's say you take the second example above, where you have NSString *page as the result of a call to stringWithContentsOfURL:encoding:error:. You can then push that string into the web view using loadHTMLString:baseURL:, assuming you also held on to the NSURL you requested:

[yourWebView loadHTMLString:page baseURL:requestURL];

I'm not sure, however, if this will run JavaScript found in the page you load (the method name, loadHTMLString, is somewhat ambiguous, and the docs don't say much about it).

For more info:

  • UIWebView class reference
  • NSString class reference
  • NSURL class reference

if you want to extract the contents of an already-loaded UIWebView, -stringByEvaluatingJavaScriptFromString. For example:

NSString  *html = [webView stringByEvaluatingJavaScriptFromString: @"document.body.innerHTML"];

Note that the NSString stringWithContentsOfURL will report a totally different user-agent string than the UIWebView making the same request. So if your server is user-agent aware, and sending back different html depending on who is asking for it, you may not get correct results this way.

Also note that the @"document.body.innerHTML" mentioned above will only display what is in the body tag. If you use @"document.all[0].innerHTML" you will get both head and body. Which is still not the complete contents of the UIWebView, since it will not get back the !doctype or html tags, but it is a lot closer.