MSHTML AccessViolationException error??!

Bob, thanks for looking into this for me, but the issue seems to lie in MSHTML refusing to work properly for me. No matter how I have tried calling createDocumentFromUrl, it always returns an AccessViolationException error. So, I was wondering, could it be I have a strange version of MSHTML? The version # I have is 7.0.3300.0.

Is this maybe not a typical version #?

Also, the real reason I wanted to use MSHTML is because I want to be able to programmatically log-in to a web site and then be able to navigate to other pages on that site while remaining logged in. So, I might be able to sidestep the issue completely if you can instead tell me (or point me int he right direction to find out) how to use other namespaces & objects to accomplish this. Ideally, I would love to work from the system.net namespace, since I am informed that this carries a much smaller footprint. Most of what I will be doing is programmatic and does not require a UI or any reason to actually display web pages. Rather, I am trying to log in to a site and then provide methods for programmatically navigating to various pages, extracting the desired data (as the logged-in user) and then present the extracted data on its own.

I have yet to find a way to get createDocumentFromUrl to work properly, so I chose a different direction.  What did you think about the web browser control idea?

Bob

Well, it looked like it still relied on the mshtml library, which for some reason won't function correctly for me. Unless I didn't understand what it was that you wanted me to look at... I would prefer to avoid the Webbrowser control altogether though, if possible. What I really need is to be able to maintain log-in credentials whenever I navigate to a new page on the same site. Sorta like the way one would log-in here to Experts Exchange and then even if you opened a new window to a new page, you would remain logged in. I am looking for a way to do that programmatically.

Experts-Exchange uses cookies, so does the site that you are connecting to use cookies for credentials?  

There are tests that you can do to see if MSHTML is properly running.

Bob

I'm not sure if they use cookies, but I would think so. The site I'm trying to access is MySpace, if that helps. And I would like to be able to automatically log-in and then use my log-in to be able to navigate to certain pages, while still logged in, and download the pages and parse out various elements. All this being done programmatically, rather than by the user.

The stumbling block is figuring out how to log-in and then maintain that logged-in state while navigating MySpace. Ideally, I would like to even be able to maintian several log-in credentials, allowing each to perofrm their own programmatic navigation with separate log-ins (for instance, for logging into both my personal and professional profiles).

Also, just now I did look a little more closely at the example you mentioned, using the MS Explorer control and am appearing to have success there. I'll have to play with it more to see if it'll do what I want. I still would prefer to not have to use this control at all. Some functionality i have written that does not require the user to be logged into his/her profile uses the WebClient and HTTPRequest objects fomr the system.net namespace. Is it possible to accomplish what I want using system.net objects?

Sorry, I also forgot to ask, how do I test my MSHTML to see if it's working?

nd you mentioned that you also have not been able to get the createDocumentFromUrl method to work - what exception do you receive when you try it?

Bob, I think I may have actually just stumbled on a workable way to use createDocumentFromUrl. Somehow, I found this article and when I copied the code directly into my IDEa, voila, it worked! I don't understand completely the why or how, but at least now I can call createDocumentFromUrl without it throwing an exception.

http://radio.javaranch.com/balajidl/2006/01/18/1137606354980.html

I still would like to know how to maintain a log-in (or even better, several log-ins) for navigation. So, any input you have on that would be appreciated. Thanks. :-)

1) That code mixes IHTMLDocument and IHTMLDocument2 to achieve the desired result.

2) You didn't answer the question about how the site handles log-ins?  Does it use cookies?

Bob

I don't know if it uses cookies. I was wondering if you could maybe help me figure out if they do. I did look in my cookies folder under c:\Documents and Settings\my_desktop_name\cookies to see if anything labelled as myspace seemed to update while I logged in and I also looked at the most recently modified cookie. Neither appeared to be affected from me logging in. But I don't honestly know what I am looking for.

And, it's ok with me if it mixes the document interfaces. In fact, I think I am going to create a new document object that implements all of the interfaces and wraps that code I found so I can simplify accessing web DOMs.

1) Is it an https:// address?

2) Have you ever heard of Fiddler?

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnwebgen/html/ie_introfiddler2.asp

Bob

I had not heard of Fiddler before, but that sure does sound like it would do the trick. :-)

And no, it is not an https:// address. The address I use is http://login.myspace.com.

I would like to know more about ways that a website can track login information. Previously, I was completely in the dark, now I know that cookies can be used for that. Could you tell me more about how web sites do that? Also, I'm going to give you the points on the next post. You were really helpful and your posts helped move me along in the right direction.

Thanks!
Beau

Fiddler is a pretty awesome tool that you can use to trace HTTP requests.  When you are going against an HTTPS URL, it is encrypted, so you can't get very much useful information.  

Any cookies that are associated with a web site are sent with the Request object on every request, so it is pretty easy to get the cookie and examine it to see if you are currently logged in or not.  Each cookie has an expiration time, and can be refreshed to keep it current.

If that doesn't make any sense, I can try explaining that differently.

Bob

It doe smake sense, but I would be very grateful for any additional information/guidance you can lend on the subject, either broadly or specifically (or both?).

In either case, your help has been, and is, much appreciated, so I'm more than glad to award you the points.