Conversation
Yurlungur
left a comment
There was a problem hiding this comment.
Good improvement. Will need to add a changelog and for tests to pass.
| EOS_ERROR(msg); | ||
| } | ||
| return; | ||
| } |
There was a problem hiding this comment.
we might want to move these to spiner instead of singularity-eos. But they're fine here for now.
There was a problem hiding this comment.
I debated doing that, but convinced myself that singularity could use hdf5 irrespective of spiner. Maybe that's a philosophical distinction and not a real one in practice
There was a problem hiding this comment.
No that's a fair point... I was just thinking of code reuse. We can leave it here for now.
|
@adamdempsey90 did tests pass on re-git? If so, I think this is good to go. |
They did not. Still investigating. |
… the return status
The issue was checking for the attributes in the wrong group. The check before trying to access an attribute is redundant, though. Now I just check the return status. Tests pass now. |
PR Summary
This hardens how the spiner EOS reads the hdf5 files, e.g., check the file exists, check groups exist, etc.
This could also be applied in Spiner when the databox reads from the file.
Partially address #592
PR Checklist
make formatcommand after configuring withcmake.If preparing for a new release, in addition please check the following: