Check whether this behavior is needed or not. If yes, provide more information in form of inline comments, maybe referring to this issue, so that we don't ask ourselves the same questions in a years time. If not, change the code and remove the workarounds as applicable.
Description
https://github.com/NCAR/ccpp-framework/blob/e86d0a7681ed96c206a44ba26bf9fd4cd302058f/scripts/metavar.py#L303
We should check why the
ddt_type
attribute is copied into thekind
attribute and then deleted. Maybe that is not necessary? Because of this behavior, PR https://github.com/NCAR/ccpp-framework/pull/512 is introducing a workaround inadd_var_debug_check
(around line 1257 inscripts/suite_objects.py
in the PR). Also, there is a workaround inccpp_prebuild.py
for this: https://github.com/NCAR/ccpp-framework/blob/0eca5c2c8885f3acadfed1f4945d19d1e97bb25f/scripts/metadata_parser.py#L212Solution
Check whether this behavior is needed or not. If yes, provide more information in form of inline comments, maybe referring to this issue, so that we don't ask ourselves the same questions in a years time. If not, change the code and remove the workarounds as applicable.
Alternatives (optional)
n/a
Related to (optional)
n/a