Struct CategoricalColumn 

pub struct CategoricalColumn { /* private fields */ }

A categorical column: a vector of codes pointing into a shared ByteDictionary, plus an optional null bitmap.

Invariant

codes.len() == nrows. If nulls is present, nulls.nrows() == nrows. A row is valid iff nulls.is_none() || nulls.unwrap().get(i). (A 1 bit means valid — same convention as BitMask elsewhere in cjc-data.)

Equality

Two CategoricalColumns are deep-equal iff they have the same codes, same dictionary contents (including ordering), and same null pattern.

Implementations

impl CategoricalColumn

pub fn new() -> Self

Empty column with a fresh FirstSeen dictionary.

pub fn with_dictionary(dictionary: ByteDictionary) -> Self

Empty column with the given dictionary (used to share a dictionary across columns or to seed with Explicit ordering).

pub fn len(&self) -> usize

Number of rows.

pub fn is_empty(&self) -> bool

True if no rows.

pub fn dictionary(&self) -> &ByteDictionary

Read-only access to the dictionary.

pub fn codes(&self) -> &AdaptiveCodes

Read-only access to the codes.

pub fn nulls(&self) -> Option<&BitMask>

Read-only access to the null bitmap if present.

pub fn is_null(&self, i: usize) -> bool

True if the row at index i is null.

pub fn push(&mut self, bytes: &[u8]) -> Result<u64, ByteDictError>

Append a non-null value, interning it in the dictionary.

Returns the assigned code.

pub fn push_null(&mut self)

Append a null. Allocates the null bitmap on first call if absent.

pub fn push_with_policy( &mut self, bytes: &[u8], policy: &UnknownCategoryPolicy, ) -> Result<(), ByteDictError>

Append a value with explicit unknown-policy handling. If the policy returns InternedCode::Null, the row is recorded as null.

pub fn get(&self, i: usize) -> Option<&[u8]>

Resolve the bytes for row i. Returns None if the row is null.

pub fn iter(&self) -> impl Iterator<Item = (usize, Option<&[u8]>)> + '_

Iterate (row_index, Option<bytes>). Deterministic.

pub fn seal_lexical(&mut self)

Run seal_lexical on the dictionary and rewrite the codes. After this call the dictionary’s iteration order is byte-lex; the codes are remapped accordingly. Bit-identical results regardless of insertion order.

pub fn profile(&self) -> CategoricalProfile

Build a CategoricalProfile summarising this column. Internal stats only; not exposed as a language-level builtin in Phase 1.

Trait Implementations

impl Clone for CategoricalColumn

fn clone(&self) -> CategoricalColumn

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CategoricalColumn

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for CategoricalColumn

fn default() -> Self

Returns the “default value” for a type.