[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#66782: 29.1; ERT tests report test redefined depending on loading se
From: |
Mattias Engdegård |
Subject: |
bug#66782: 29.1; ERT tests report test redefined depending on loading sequence |
Date: |
Thu, 2 Nov 2023 19:56:41 +0100 |
2 nov. 2023 kl. 18.17 skrev Xiyue Deng <manphiz@gmail.com>:
>> I understand if upstream don't want to complicate `require' logic too
>> much. However I wonder whether it's OK to add warning if a required
>> module has `ert-deftest' in it, so that it can help people identify that
>> a `Test "foo" redefined' error is due to requiring other module instead
>> of an actual duplicated test name. How does this sound?
>
> As I didn't get an answer I assume this was a no-go.
No, please don't make such an assumption -- I was just busy elsewhere and
hadn't given your message the attention it deserved. Sorry about that.
That said, in this case I'm not sure how to implement your suggestion in a
clean way and if all that effort is really worth the trouble, so perhaps the
answer would be the same anyway. And we probably don't want to prohibit
`ert-deftest` in required modules in general for reasons mentioned -- they
could be used with perfectly fine discipline elsewhere.
> So instead I'd
> like to propose a slight change to the error message to mention that it
> may also be caused by an ert test being loaded multiple times. Patch is
> attached, please let me know whether this works.
I wouldn't mind such a change if it really would help. Would it? Isn't it just
restating the problem in other words?