Project_Scanner: PythonScanner fixes and testing

[FrancescoPezzella]

No longer hard-locked to only look for src/lib/test directories.

Instead of having duplicate Scanner methods and old broken code, I decided to make a PR addressing this issue. To be clear, other methods like JavaScanner also have their own issues and that has been noted with a TODO comment.

(#198)

[FrancescoPezzella]

The CI/CD failing has nothing to do with this branch. When pymarkdown was updated to 0.9.40, it added new doc requirements.

[pjljvandelaar]

We will merge after #205
As tests should pass afterwards.

[pjljvandelaar]

Could you explain why you opted for using str instead of Path?

When using str don't you have the risk that it isn't a valid Path? If so, shouldn't an exception be captured and rethrown in that case - to ensure a clear and correctly localized exception to the user?

[FrancescoPezzella]

Could you explain why you opted for using str instead of Path?

I did it for the sake of consistency. Changing it to Path would have been a bigger change, since other methods don't use it. Find_Source methods in all the other Scanners return a list[str]. I didn't want PythonScanner to be exception.

    def find_sources(self) -> list[str]:
        """Return every file entry listed in compile_commands.json, sorted and deduplicated."""
        if not Path(self.compile_commands_path).exists():
            message = "compile_commands.json not found"
            raise FileNotFoundError(message)
        with Path(self.compile_commands_path).open() as f:
            commands = json.load(f)
        return sorted({entry["file"] for entry in commands if "file" in entry})

    def find_sources(self) -> list[str]:
        """Return every .java file under root_dir, sorted."""
        # TODO: is this correct? does this filter out files correctly?
        # See https://github.com/TNO/Renaissance.Py/issues/200

        java_files = Path(self.root_dir).rglob("*.java")
        return sorted(str(f) for f in java_files)

    def find_sources(self) -> list[str]:
        """Generate compile_commands.json via Bear if missing, then return its listed sources."""
        if not Path(self.compile_commands_path).exists():
            self.run_bear()
        return super().find_sources()

When using str don't you have the risk that it isn't a valid Path? If so, shouldn't an exception be captured and rethrown in that case - to ensure a clear and correctly localized exception to the user?

I will correct that in the commit.

[FrancescoPezzella]

Rebased and fixed all changes