“try { if(eval(‘window.’+'winName’+’.document’)) { alert(‘YES’); } } catch(e) { alert(‘NO’); }”

If you need to catch user actions of these links, you have to install your own onclick handlers which will call your own javascript function whenever the link is clicked.

Thanks for your awesome blog, it is very helpful!
I am wondering how to catch the url of links like <a href=”#content” rel=”nofollow”>Skip to this page’s content</a>. shouldStartLoadWithRequest would not be called in this case.

Thanks for your help!

Ok, back to the drawing boards. I downloaded your app and it works fine on the page as is. Very nice app by the way. I’m going to keep fooling around and see if I can’t get things working.

<a href=”javascript:void()” onclick=”location.href=url”>

which should be just be done this way:

<a href=”url”>

Or if they need the onclick hander, they should return the value “false” for the onclick handler to tell the browser that the standard behavior of the browser (open the page from the HREF attribute) should not be done. Then they can just put the real URL into the HREF attribute and open the URL via onclick handler. Then the links would also work when javaScript is switched off, then the link your also work when it is used outside of the context of the web page, like when a browser reads the link via contextual menu to open it in a new tab or the user wants to copy & paste the like etc.

Thanks. If I have access to the web developer, what do I have to get him to do to implement a better practice so that I could grab the link? Just use the HREF attribute? Is that something that could be done if the source files are being pulled from a database and those are the pdf links that are being opened? I.e. this is a court system that pulls records from an archaic database and displays the PDFs from the links.

Also, you’re making me feel better about my limited programing as I can get webkit to open links that usually open in new windows, but I wasn’t able to get those to work.

I guess the web page is using JavaScript to open the PDF files via onClick handler and uses these these handlers in “A” tags. And because the onClick handlers are already opening the PDF file, the HREF attribute seems to be useless and so the web developers inserted “javascript:void(0)” as URL in these HREF attributes, which is definitely not what should be done in this case. This is just a bad practice and also introduces usually many issues and should be avoided. One issue that you can no longer copy the link URL, because there’s no link URL anymore. And for UIWebViews this means you can’t directly control where these links will open, because the “links” itself are just “dummies”.

I understand and have seen UIWebView methods implemented in the controller file, correct? Specifically, in my app “Story_2_testViewController.m” is where I’m placing your method to access the ModifyLinkTargets.js file. Right before my viewDidLoad method.

The delegate methods are those from the UIWebView class. You have to implement some of the delegate methods from your UIWebView class to be notified when the page has finished loading and you can execute the JavaScript code.

very nice tutorial. I’m following what you’re doing, just not how.

Awesome blog and app.

I’m wondering if it’s possible with UIWebview on iOS to intercept the web page request so as to re-direct to a proxy and feed the request there. i.e. http://www.bbc.com should go to proxy 192.168.1.10:1234 and use NTLM authentication.

I’ve tried by doing this in shouldStartLoadWithRequest using ASIHTTPRequest & loadHTMLString but that doesn’t bring back resources (images, css etc).

Is this even possible with UIWebView on iOS?

But the demo page of this rich text editor is able to detect an iOS browser and then simply falls back to a regular “textarea” element of HTML, so you can enter text on iPad/iPhone as well, but here it’s a standard text field, not a rich text field.

The web page probably uses the “UserAgent” information to find out if the page is running on and iPad/iPhone and maybe other touch-based devices.

The problem with rich text fields on touch.based devices with a virtual keyboard is that on these devices the keyboard is not attached and available all the time. The keyboard only opens when you actively activate a text field. But “rich text editors” on web pages are not “real” text fields, these are in Javacript programs which build up some HTML and CSS code and render it in an iframe so this whole thing looks like a text field with lots of properties and settings for all the features. But it’s not a “real” text field the iOS would know of. The Rich text editor is then listening for keyboard events on document or window level, which is the main issue on touch-based devices with virtual keyboards. On a desktop computer where the keyboard is available all the time, you can create keyboard events all the time by pressing a key, even if no text field is “listening”. The OS delivers these keyboard events also to GUI elements or objects which are not textfields. But on devices with virtual keyboards this does not work, because the keyboard is only activated when a textfield is activated, and in order to make rich text editors work, the virtual keyboard would have to be available all the time, even when no textfield is active.

Thanks in advance!