That was useless
Diff Detail
Diff Detail
- Repository
- rDCIDX Metadata indexer
- Branch
- unmock-libmagic
- Lint
No Linters Available - Unit
No Unit Test Coverage - Build Status
Buildable 3280 Build 4226: tox-on-jenkins Jenkins Build 4225: arc lint + arc unit
Event Timeline
Comment Actions
Build is green
See https://jenkins.softwareheritage.org/job/DCIDX/job/tox/188/ for more details.
Comment Actions
The idea was to avoid calling side-effect function and depend on the actual call.
In the long run, i wanted to avoid failures here because for example, the inner implementation changed their detection mechanism.
Comment Actions
I don't think there's any side-effect
In the long run, i wanted to avoid failures here because for example, the inner implementation changed their detection mechanism.
In which case we'll just get a test failure.
On the other hand, mocking libmagic means we don't test that we use libmagic properly.
Comment Actions
I don't think there's any side-effect
reading a file is a side-effect...
wait, we are no longer reading a file (first implementation did ;)
In which case we'll just get a test failure.
Yes, but that's not something i wanted. That's libmagic's job to properly test itself.
On the other hand, mocking libmagic means we don't test that we use libmagic properly.
Yes, that was the goal at the time.
To be clear, i'm not fighting this, just mentioning the initial idea.