Fixed that build outputs are infinitely recursively processed

[SanjeetBalamurugan]

What does this PR do?

Fixes infinite recursion when a custom rule generates a file with the same extension as its input. Previously, generated files could be treated as new inputs for the same rule, causing the rule to be applied repeatedly.

How does this PR change Premake's behavior?

Generated files from custom rules are now tracked with the rule that created them. Before applying a rule, Premake checks whether the file was already generated by that same rule and skips it if so.

This prevents recursive rule application while keeping existing custom rule behavior unchanged. No breaking changes are introduced.

Closes #1940

Anything else we should know?

This addresses issue 1940.

The fix was verified with a regression case where a custom rule generates another file with the same extension as its input. The generated output is created once and is not recursively processed.

Did you check all the boxes?

• Focus on a single fix or feature; remove any unrelated formatting or code changes
• Add unit tests showing fix or feature works; all tests pass
• Mention any related issues (put closes #XXXX in comment to auto-close issue when PR is merged)
• Follow our coding conventions
• Minimize the number of commits
• Align documentation to your changes

[SanjeetBalamurugan]

.

[SanjeetBalamurugan]

May I know if there is any problem with my pull request? @Jarod42

[Jarod42]

I'm a simple contributor to Premake, like you. I am not a member of Premake.
So I don't have any right.
As contributors, I also do review to spot issue. and for your PR, I don't see problem (There is the fact there are no tests though which might be problematic.)

[SanjeetBalamurugan]

@Jarod42, Sanjeet here.

Thanks for your review regarding my pull request. As for the tests, i still have doubts about how to add them. If its a really needed, I would try to add them. But i dont know how tests work in premake-core.

Thanks