fn generate_split_method(
row_name: &Ident,
variant_infos: &[VariantInfo],
impl_generics: &ImplGenerics<'_>,
ty_generics: &TypeGenerics<'_>,
where_clause: Option<&WhereClause>,
) -> TokenStreamExpand description
Generate the split representation for an enum DataFrameRow: the hidden
to_dataframe_split associated method (body holder) plus the
DataFrameRowSplit bridge impl that surfaces the public
rows.into_dataframe_split() verb (IntoDataFrameSplit).
Returns a DataFrameShape whose partitions are rooted BuiltDataFrames
(#1750): Bare for a single-variant enum, PerVariantList (one entry per
variant, named with the snake_case variant name) otherwise. Each partition
data.frame has only that variant’s columns (non-optional types — no NA fill
from other variants).