pond-ts API Reference (core)
    Preparing search index...

    Class ChunkedStringColumn

    String chunked column. Same shape contract as the other chunked variants; chunks are plain StringColumn and can have per- chunk dictionaries that differ from one another. read and scan route through each chunk's own read so the per-chunk encoding (dict vs fallback) is handled transparently — no cross-chunk dictionary unification at construction. See the class-level comment below for why that's deferred.

    Index

    Constructors

    constructor

    Properties

    chunkOffsets chunks kind length storage validity?

    Methods

    at first firstDefined hasMissing last lastDefined nullCount read scan slice sliceByIndices sliceByRange uniqueCount

    Constructors

    Properties

    chunkOffsets: Int32Array
    chunks: readonly StringColumn[]
    kind: "string" = ...
    length: number
    storage: "chunked" = ...
    validity?: ValidityBitmap

    Methods

    • Gather → plain StringColumn. The result is built via stringColumnFromArray so the dict-vs-fallback heuristic runs once on the gathered values — matching what materialize's compact step would produce.

      Parameters

      • indices: Int32Array

      Returns StringColumn

    Settings

    Member Visibility

    On This Page

    Constructors
    Properties
    Methods