The 2 are different, let's start by defining a couple of things:
Snapshot: Page + DOM state (starts as no DOM)
Page: Resource per URL
You may see the same "page" being audited, but actually it's a different snapshot (DOM state) of the page, for example, the list of transitions will probably be different -- not necessarily, but usually new states come after DOM events which are recorded as page DOM transitions.
Also, the system doesn't crawl first and keep a list (or map or tree) of resources to audit later, it's all on the fly and in a way interconnected in a feedback-loop -- the crawl and audit are basically complementary processes.
Not sure if the above makes sense, I haven't had my coffee yet and the first draft of my reply really did not make any.
Hi Tasos, we finally succeded to scan the entire website (2.5 millions lines of code) :) .
We found a workaround by downgraded phantomjs version. From what we understood, it was a shared port issue by phantomjs and selenium.
Unfortunatly, we had turn off scans : xss_dom* and unvalidated_redirect*
From what I saw, with dom redirects it stayed stuck in loop.
I can't give you access to our code/website, but if you like we could help you for debugging (beta test).
today issue is in protocol.rb:158:in 'rescue in rbuff_fill'
will update if we found a solution for it.