solution for https pages?

[keith523]

Hi, I found your script from just a little searching (it was actually apart of a helpful "what are my options" type thing that lead to me just what I needed.) and had minimal trouble configuring

sidenote: i set it up under easyphp 1.8, a local apache webserver and mysql db server.  Only hitch was to change the thumbnails folder to my www root path and escape the backslashes ("c:\\program files\\...")

IECapt.exe is a handy tool that seems to be the backbone, but from what I can tell, it doesn't support https protocol.  I'm trying to develop a solution for visually checking a few thousand webpages but they're all behind secure http.

Any ideas on a workaround or if I maybe still have something misconfigured?  the images generated by IECapt are blank white, no matter the format (jpg, gif, png)

Thanks again for all your hard work, and to anyone who might be able to help!
-Keith

[zubrag]

IECapt and Website thumb creator should have no problems with https. Tested on Windows XP SP1, IE 6.0.

I suspect the problem in your case is somehow related to security dialogs which IE may show when connecting to https website (or maybe service pack 2 somehow blocks it).

Try running IECapt manually, like this. Does it show some prompts/dialogs? Does it generate thumb.png?

IECapt.exe  https://adwords.google.com/select/Login thumb.png

[keith523]

zubrag, thanks so much for your dedication to these boards!

I thought I had tried the command line with an https page, but I guess not successfully.  The link you suggested worked fine from both the command line as well as the script through my browser.

The host I'm trying to get working does require authentication (web-form based) but I'm pretty sure I was attempting from an authenticated session.

I'm home now, but will try tomorrow afternoon.  Cheers!

[keith523]

I can, however, give a little bit more diagnostics about what's going on.

I tried the script against my own webserver's admin login over https.  The script works fine for port 80, but unfortunately is not where the login form is located, which is 8443.  When I use that port in the url= parameter, I get a blank image, just like I was getting before with my target host.

The command line of iecapt produces the same, blank white image file for port 8443.

Maybe that helps flesh out the issue a little bit.  My target host is on port 443 I believe, and my goal is to be able to use this with other get parameters, which themselves might gum up the works.  but we'll cross that bridge later.

Thanks a million
-Keith

[zubrag]

I see. Yes, custom port may be causing problem. I did not test it with custom ports. It all depends on iecapt.exe. If it cannot grab image via 8443 then I can do nothing. Unfortunately. But seems to me I've seen another exe program which does the same thing as iecapt (capture website image from the browser). Cannot remember the name now, but i'll do some research on this subject.

[keith523]

The test https site I was trying at home is port 8443, but the target site I believe is the normal 443.  And actually, if I point iecapt directly at the exact login page, it captures correctly!

This is a good step forward, but I'm still left with my authentication problem.  IECapt works as if it's opening a new instance of IE, surfs to the host provided, and grabs the image.  I was hoping that if I logged into the site through IE, it would register my session for all instances of IE.  Unfortunately, if I open a new window from my valid session, and try to surf directly to the page I'd like to capture, I get redirected to the login, which would probably explain the blank image that's generated.  IECapt sees the target just fine, the only thing is that it's a 302 status code redirecting hte browser to the login.

Basically, I need IECapt to use an authenticated instance of IE, or somehow set a session cookie to a valid one right before surfing to the page.  I'll try inserting some javascript and see how IECapt reacts.

-Keith

[zubrag]

Seems to me IE5 was copying session data to new windows. Maybe Microsoft changed that behavior on IE6 and IE7 for some security reasons. I'll check on whether it is possible to have iecapt.exe (or maybe some other capture program) to use existing IE instance.