[~regression fix] fix Issue 19248 and reopened 18957

[9il]

I am not sure if it is normal regression. The issue 18957 has been fixed, but partially. The example in the issue 18957 is wrong, and it looks like it caused the wrong test.

[dlang-bot]

Thanks for your pull request, @9il!

Bugzilla references

Auto-close	Bugzilla	Severity	Description
✓	19248	regression	Wrong mangle for C++ const STL classes/structs

⚠️⚠️⚠️ Warnings ⚠️⚠️⚠️

• Regression or critical bug fixes should always target the stable branch. Learn more about rebasing to stable or the D release process.

To target stable perform these two steps:

1. Rebase your branch to upstream/stable:

git rebase --onto upstream/stable upstream/master

2. Change the base branch of your PR to stable

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub fetch digger
dub run digger -- build "master + dmd#8700"

[ibuclaw]

No test case and mind the regressions introduced.

[9il]

This is not regressions, this is wrong test

[9il]

Test fixed now

[9il]

BTW, looks like it is further fix for https://issues.dlang.org/show_bug.cgi?id=18957.
Should it be targeted to stable?

[9il]

ping @ibuclaw

[andralex]

ping @ibuclaw @WalterBright

[ibuclaw]

You have not added any tests, only comments.

Copying and pasting the source code of another project is questionable also.

[ibuclaw]

The test should make sure that linking works with mixed C++ and D. Testing the mangle string is not enough, especially if you claim the old test to be erroneous.

It wasn't erroneous for whoever added the test.

[ibuclaw]

Should it be targeted to stable?

Stable should only be receiving regressions. It will eventually make its way to master in a week anyway, so up to you.

[9il]

Should it be targeted to stable?

Stable should only be receiving regressions. It will eventually make its way to master in a week anyway, so up to you.

OK, lets it be master then

Copying and pasting the source code of another project is questionable also.

removed

The test should make sure that linking works with mixed C++ and D. Testing the mangle string is not enough, especially if you claim the old test to be erroneous.

done

[jacob-carlborg]

Please squash commits when done.

[9il]

Please squash commits when done.

rebased

[TurkeyMan]

I'm no linux expert, but I just did some mangling tests with my in-progress C++ work against actual C++ on linux, and reproduced @9il's findings in this PR.
So, LGTM.

Test case added.