in_toto_verifylib: fix require/disallow queue

[SantiagoTorres]

Fixes issue #: None

Description of the changes being introduced by the pull request:

This reverts commit 21129a6 and
replaces the consumed = artifacts_queue with consumed = set(), the
reason as to why is that a require rule passing overwrites all the queue
with a set() call...

Please verify and check that the pull request fulfills the following
requirements:

• The code follows the Code Style Guidelines
• Tests have been added for the bug fix or new feature
• Docs have been added for the bug fix or new feature

[coveralls]

Coverage remained the same at 100.0% when pulling 5056a13 on fix-require-queue-modification into cc73474 on develop.

[lukpueh]

Thanks for the fix, @SantiagoTorres! I referenced the PR in the list of artifact rule related PRs (in-toto/specification#4).

What do you think about defining the empty set only once above the if/else cascade? Similar to how we do it in our in-toto-golang (where the compiler forced us to do this right in the first place).

[SantiagoTorres]

I was hesitant to do that just to avoid it to allocate a new set() for every iteration. I wonder if what we'd like instead is to declare it as an empty set outside the for?

[lukpueh]

Okay, fair enough. Not sure I understand your alternative proposition. We can't declare consumed outside the loop as it gets overridden by match/create/delete/modify/allow rules. Or did you mean declare an empty_set = set() and in disallow/require do something like consumed = empty_set?

I think my preferred solution would be one of,

• as it is now,
• what I suggested above (despite the unnecessary allocations),
• in match/create/delete/modify/allow replace
  consumed = verify_*_rule(... with artifacts_queue -= verify_*_rule(...,
• modify verify_disallow_rule and verify_require_rule to return an empty set too and also assign it to consumed

[lukpueh]

Let's merge this soon. Let me know what you think of my suggestions (above). If none if it seems preferable we can merge as is.

[SantiagoTorres]

Let's do what you suggested above despite the allocations.

[lukpueh]

@SantiagoTorres, I took over your branch. And made the small modification we talked about. I hope that's okay? I also fixed the outdated docstring. Can you take a quick look at it and greenlight?

[SantiagoTorres]

LGTM. Thanks!