Repository navigation
IMP: recognize type-mapped Bool variables as Bool variables - #401
Conversation
|
|
ebolyen
left a comment
ebolyen
left a comment
There was a problem hiding this comment.
I don't think you actually need this PR for what you wanted to do, but I have no objection to making it so that TypeVar where all branches are Bool is treated the same as Bool % TypeVar.
|
What's your preference re keeping this? Indeed your changes in |
|
I think there's no harm in making this work. I guess the only difficulty is testing and if it's worth adding another action to dummy_plugin for it. I'm also fine with putting it on the back-burner. There are perhaps other ways to handle this directly, such as distributing the TypeVar branches over the predicate so that it just always normalizes. Although it may be tricky to ensure that it still looks correct at different call-sites. |
|
|
||
| def test_higher_union_bool_typemap_parameter_is_flag(self): | ||
| plugin = _register_higher_union_bool_flag_action() | ||
| command = PluginCommand(get_plugin_state(plugin), 'dummy-plugin') |
There was a problem hiding this comment.
Well that looks pretty good to me actually. I wonder if we could transition more tests this way? Then q2cli isn't so coupled to a test plugin in another package.
3 participants
Description
Fixes an issue where type-mapped
Boolvariables were not recognized asBools and rendering on the cli fell back to the standard--p-param VALUEformat.AI Disclosure
AI Usage Details
Generated by Codex while working on qiime2/q2-dada2#190