You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I just noticed that it copies way too many files back than I wanted.
I agree that it can end up copying a lot (especially the lib and rlib artifacts). I have updated the PR to include only bin, dylib and cdylib by default. It would be nice to also allow this to be overridden by a config option that explicitly specifies all of the kinds of artifacts to include (if the config option is missing, the default is used, otherwise whatever is in the config option is used). It also might make sense to provide a config setting where the user can (optionally) provide a regexp to match artifact filenames that should be copied (this would be checked after the initial artifact-kind filter has been applied). Does that sound reasonable?
Would you want to implement the option? Or shall I do it after this PR?
I could implement it, but might not get to it right away. I'm fine with taking either path, but probably the new PR can be merged in any case.
With more artefact files being copied back to the project, it might be more suitable to make this configurable.
In first iteration wargo only copied back the executable binaries, which was sufficient for my own use cases.
But this doesn't solve all cases, some people need library files for example.
From #110:
I agree that it can end up copying a lot (especially the
lib
andrlib
artifacts). I have updated the PR to include onlybin
,dylib
andcdylib
by default. It would be nice to also allow this to be overridden by a config option that explicitly specifies all of the kinds of artifacts to include (if the config option is missing, the default is used, otherwise whatever is in the config option is used). It also might make sense to provide a config setting where the user can (optionally) provide a regexp to match artifact filenames that should be copied (this would be checked after the initial artifact-kind filter has been applied). Does that sound reasonable?I could implement it, but might not get to it right away. I'm fine with taking either path, but probably the new PR can be merged in any case.
Originally posted by @edgimar in #110 (comment)
The text was updated successfully, but these errors were encountered: