fix: defer validation for unresolved conditional branches

[Stefanqn]

No description provided.

[Tishka17]

Actually, I would check as much as possible at graph building stage. Probably, we should have a toggle to enable this behavior

[Stefanqn]

both is fine for me

[Tishka17]

Looks like we have 2 options here:

1. validate disabled
2. validate dynamically calculated

[Stefanqn]

like that?

[Stefanqn]

I also added a toggle per factory in my latest commit

Should alias(...) and from_context(...) support the same override as well?

[Tishka17]

Isn't per factory setting too much? Do we have any UseCase for that?

[Stefanqn]

if it's too much, drop the last commit :)
I think it's kinda useful if you'd like to keep strict / fail-fast behavior and want to explicitly mark or patch a certain case.

[Stefanqn]

Can I accelerate merging any kind of bugfix for when=Has()? I'm really looking forward to use this feature

[Tishka17]

Do you have any particular usecase for this specific PR? I am not sure if I am satisfied with the API changes here though it has some sense here. Do you need this together with #696 at this moment?

[Tishka17]

In terms of configuring validation is see these cases:

1. Factory is overridden
2. Factory is just disabled
3. Factory is disabled by static evaluation
4. Factory can be dynamically disabled
5. Factory is enable by static evaluation
6. Factory is always enabled

For me it is not clear how should we let user decide whether to skip checks. In #696 I went with simple logic: cases 1,2,3 disable validation, others are always validated.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud