[REVIEW] libgdf_cffi renamed to libcudf_cffi #1343
Conversation
|
Can one of the admins verify this patch? |
pradghos
changed the title
|
We plan to replace CFFI with Cython in 0.7 release so this may be redundant. Suggest starting by filing an issue in the future to discuss and save effort. See #599 |
|
ok to test |
pradghos
changed the title
|
I know |
harrism
added
3 - Ready for Review
|
Please keep me posted in case of any further review comments / suggestion . Thank you! |
| # with pytest.raises(GDFError) as raises: | ||
| # libcudf.gdf_add_generic(col, col, col) | ||
|
|
||
| # raises.match("CUDA ERROR.") |
There was a problem hiding this comment.
What's the goal of this test?
There was a problem hiding this comment.
@kkraus14 : This is existing test case only ! Probably after renaming the path from libgdf_cffi to libcudf_cffi, it is showing differently. We can see there is another test_errorhandling.py being marked as deleted also.
So, there is nothing change as functionality is concern - Like all the places in code drop libcudf_cffi and libcudf is updated to pick up the intended module name. Thank you !
|
@kkraus14 Do we need to move this PR forward if we are just going to rip CFFI out in this release? If not, I recommend closing. |
Let's leave this in our back pocket in case that slips unexpectedly, so lets leave this open until then. |
kkraus14
added
0 - Blocked
|
Looks like @kkraus14 has replaced nearly all CFFI with Cython and will complete the task in this release, so closing this PR. Thanks @pradghos and sorry for the waste of your effort. To avoid that in the future, start a discussion in an issue first. I think you did file an issue, but needed to wait longer for a response before starting the work. |
This PR is addressing #1341