Description
Either a fork is required, or an alternate solution.
The two things we need is:
-
Preload the eval context with values. I intend to load in terraform plan output, and possibly even tfstate. This is so I can reference resolved data blocks if I have this information.
-
Apply semantics to arbitrary data blocks. This effectively "simulates" what the terraform provider would do for a given block, when the semantics can be done locally (without some external api or something). This is very similar to the presets you referred to in how some of your aws blocks are handled today.
Open PR here: aquasecurity/trivy#8302
This PR will likely not be accepted. If a fork is created, this fork needs to be maintained, ideally with patches to keep tracking the upstream simple.