input: support for variable depends #510
Conversation
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## master #510 +/- ##
==========================================
- Coverage 88.26% 87.90% -0.36%
==========================================
Files 46 46
Lines 14370 14400 +30
==========================================
- Hits 12683 12658 -25
- Misses 1687 1742 +55
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
There was a problem hiding this comment.
I think there is a fundamental problem with the handling of provideDeps. With the current approach the variable expansion will be done after the pattern expansion. This is exactly the opposite order in which shells are doing it. Think about patterns like *-${FOO}-dev. This won't ever match anything. I think the order has to be reversed to be useful...
rhubert
force-pushed
the
317
branch
3 times, most recently
from
48383c4 to
9516691
Compare
|
At least |
|
I think the |
|
I think the code works as intended. However, I took myself the freedom to optimize it a bit so that the performance overhead is not noticeable. Previously the recipe parsing time was ~8% longer, even if the feature was not used. Now it's somewhere at 0.3%. Once the documentation is updated it's ready for merging I guess... |
|
Thanks!!! I did some testing this morning and it's still working for my use-cases ;) Added a few words to the documentation + squashed my commits. |
Apply string substitution to the names of dependencies. This avoids the need of duplication and/or use of `if`.
Most likely no string substitution is required on provideDeps. Wrap the entries into a specialized class that handles only what is really necessary.
|
Thanks! |
2 participants
No description provided.