Create artefact store object #301
Conversation
…`pytest tests` at source directory
Minor documentation update for #277.
278 absolute paths in compilation
…ed, as reported by CI pipeline on Github
#5 Fix test warnings
There was a problem hiding this comment.
No major issues here.
I imagine our future contains a long and involved discussion about redesigning and refactoring the way artefacts are stored and managed but in the mean time this tidies up the current implementation a little.
It also starts the move away from random anonymous data structures and towards properly structured objects.
|
OK, I don't know what to do about the type information, please advise ;) |
Hmm - not really, it seems to work :) The only question I have is in #300: for the follow-up PR, do you prefer to have methods for certain standard artefacts (e.g. all f90 files, i.e. all files to be compiled by Fortran, ...), or do you prefer to have some constants exposed and use them in the standard way? Otherwise ready for next review. |
|
Gee, python-specific error messages - annoying. |
| # 3.8: ... with abstract method | ||
| # 3.12: ... without an implementation for abstract | ||
| # so we only test for the begin which is identical: | ||
| assert "Can't instantiate abstract class MyClass with" in str(err.value) |
There was a problem hiding this comment.
I don't mind the "belt and braces" approach here but I think the actual error message is of secondary importance. The key thing is that the exception is thrown.
This fixes a #TODO, and prepares for #300.