Fix AssemblyTests on Windows

[lefou]

Depends on #3285

Windows has an env var called PATHEXT whish is usually something like this: .COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH;.MSC
Basically a list of extensions considered as executable.
So the file in Mill's case needs to be a .bat file.

Issue analyzed and fixed by @sake92

Fix #3283

[lefou]

@sake92 Thank you for the fix! Can you please format it and also bump the Mill version to 0.11.9 so we see it succeeds? Thank you

[sake92]

@lefou where do I bump the mill version, in .config\mill-version ?

[lefou]

@lefou where do I bump the mill version, in .config\mill-version ?

Yes.

[sake92]

Seems like scalalib tests are green now.
https://github.com/com-lihaoyi/mill/actions/runs/10060348007

[lefou]

Looks like there are some new unrelated test failures for Windows now. Since we now know, the assembly tests are fixed by this PR, we could remove the Mill bump, and merge this PR (and close the bounty, I guess).

[sake92]

Looks like there are some new unrelated test failures for Windows now. Since we now know, the assembly tests are fixed by this PR, we could remove the Mill bump, and merge this PR (and close the bounty, I guess).

I guess it would be fair to make the CI green first, before closing the bounty. 😁

[lefou]

Looks like there are some new unrelated test failures for Windows now. Since we now know, the assembly tests are fixed by this PR, we could remove the Mill bump, and merge this PR (and close the bounty, I guess).

I guess it would be fair to make the CI green first, before closing the bounty. 😁

Yeah. That was the idea by not bumping Mill to cause unrelated issues.

[lefou]

I think the other test failures are valid but unrelated. They probably went under the radar since our Windows CI wasn't fully functional before #3285 .

@sake92 Looks like you are on Windows, so any further help to fix these would be much appreciated. 🙏‍

[sake92]

@lefou we can merge this one?
It is easier to fix them one by one I think.