Using WebScarab

WebScarab can be used stand-alone, or in conjunction with a browser. If you
plan to use a browser, you will need to set the upstream proxy to point to 
the computer on which WebScarab is running. WebScarab's proxy can intercept
SSL as well as unencrypted connections. If necessary, you can configure 
an upstream proxy in WebScarab, effectively placing WebScarab into a proxy
"chain". WebScarab only supports Basic Authentication, so if your upstream
proxy requires NTLM authentication, you will not be able to use WebScarab.
We would appreciate assistance in implementing this feature.

Once you have configured your browser and WebScarab, you should be able to
simply browse the web site under test. 

SECURITY NOTE: If it is an HTTPS site, you will be prompted by your 
browser, since the certificate WebScarab uses to masquerade as the 
destination web site will obviously not be the correct certificate 
for the site. We recommend that you do not accept the certificate
permanently, since that could lead you to trust a site that you should not,
even when you are no longer using WebScarab. Rather accept the certificate 
for the duraction of the session only.

NOTE:

When you start WebScarab, it will run with reduced functionality until
you start a new session, or open an old one. In particular, you will be unable
to see the details of the requests and responses in the conversation tab.
This is because the requests and responses can take up a LOT of memory very
quickly, and lead to exhaustion of available resources. Consequently, the
requests and responses are discarded immediately after processing, and
read back from disk if requested. If no session has been created, those
requests and responses cannot be retrieved.

While you are navigating the site using your browser, WebScarab will be 
recording information about the requests and responses that it sees. This 
information includes the request method, the URL, any cookies sent, and
any parameters or entity body. It also records the response code and message, 
any cookies set by the server, and other information extracted by the various
plugins. Some of this information can be seen in the Conversation log, but
the rest should be visible in the various plugins. 

For example, the Spider plugin maintains a list of links that have been 
seen, and not yet visited, arranged in both a list and tree view. These 
links can be selected by the operator, and requested from the server, 
using up to 4 (hard-coded) concurrent threads. If desired, the Spider can
be instructed to retieve pages recursively, queueing new links as soon as it
sees them. There is a method of controlling which pages may be fetched in
this fashion, using regular expressions for the sites which may be spidered,
as well as a regular expression for paths on those sites which should not
be spidered. This can be used to prevent fetching binary files such as PDF, or
following a URL that would invalidate a session cookie.

