Uh oh!
There was an error while loading. Please reload this page.
Miscellaneous cleanups - #17189
Conversation
It became redundant when `input_expr_types: &[DataType]` was replaced with `input_expr_fields: &[Field]`, i.e. in 577c424 commit. The benefit of passing in a "field" is that it combines data type and nullable attribute.
7c383fc to
dbf832aCompareIn `UserDefinedLogicalNode::check_invariants`, the actual plan to check for invariants is `self`. The `plan` is always `LogicalPlan::Extension` and provides no further information. It's confusing.
timsaucer
left a comment
There was a problem hiding this comment.
These all look like very nice cleanups. Thank you for the PR!
Uh oh!
There was an error while loading. Please reload this page.
| field_qualifiers: vec![None; field_count], | ||
| functional_dependencies: FunctionalDependencies::empty(), | ||
| }; | ||
| dfschema.check_names()?; |
There was a problem hiding this comment.
Context is that checking names here makes this consistent with the other (fallible) ways to create a DFSchema from a Schema
There was a problem hiding this comment.
FWIW it seems this has caused a regression for some users in 50.0.0 (there are some code paths that violate an invariant). See #17706
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
This reverts commit c31ab2e.
alamb
left a comment
There was a problem hiding this comment.
Thanks @findepi and @timsaucer
Uh oh!
There was an error while loading. Please reload this page.
See individual commits