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
Is your feature request related to a problem? Please describe.
Since rapidsai/cudf#9715 was merged, cuDF is consistent with Python, where the '$' EOL regex pattern character (without MULTILINE set) should match at the very end of a string and also just before the end of the string if the end of that string contains a new-line.
Java has similar behavior but considers \r to be a new-line character in this context, where cuDF and Python do not.
This means that a pattern such as a$ would match both a\n or a\r with Spark on CPU, but would only match a\n on the GPU.
Describe the solution you'd like
I would like the behavior to be consistent.
We either need additional support in cuDF to emulate Java behavior here, or we need to have our own version of the regex kernels, or perhaps we can find some workaround in the plugin but I don't have any ideas yet for how to do this in a simple and low-risk way.
Describe alternatives you've considered
None
Additional context
None.
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
Since rapidsai/cudf#9715 was merged, cuDF is consistent with Python, where the '$' EOL regex pattern character (without
MULTILINE
set) should match at the very end of a string and also just before the end of the string if the end of that string contains a new-line.Java has similar behavior but considers
\r
to be a new-line character in this context, where cuDF and Python do not.This means that a pattern such as
a$
would match botha\n
ora\r
with Spark on CPU, but would only matcha\n
on the GPU.Describe the solution you'd like
I would like the behavior to be consistent.
We either need additional support in cuDF to emulate Java behavior here, or we need to have our own version of the regex kernels, or perhaps we can find some workaround in the plugin but I don't have any ideas yet for how to do this in a simple and low-risk way.
Describe alternatives you've considered
None
Additional context
None.
The text was updated successfully, but these errors were encountered: