Interface ResultToColumnarBatchOptionsExperimental

Explicit conversion window plus the batch identity an object-path Result cannot supply on its own. Every field except maxFeatures defaults from ColumnarResult.columnar when the Result came from columnarBatchToResult.

interface ResultToColumnarBatchOptions {
    attributes?: ResultToColumnarBatchAttributeOptions;
    geometry?: ResultToColumnarBatchGeometryOptions;
    id?: string;
    identity?: ColumnarBatchIdentityV1;
    limit?: number;
    limits?: GeoArrowConversionLimits;
    maxFeatures: number;
    offset?: number;
    rowOffset?: number;
    schemaId?: string;
    sequence?: number;
    unmappedAttributes?: "fail" | "drop";
}

Properties

id?: string
limit?: number

Features to convert from offset. Defaults to the rest of the array.

maxFeatures: number

Hard ceiling on converted rows. Required, for the same reason as on ColumnarBatchToResultOptions.

offset?: number

Zero-based first feature of the window. Defaults to the first feature.

rowOffset?: number
schemaId?: string
sequence?: number
unmappedAttributes?: "fail" | "drop"

What to do with an attribute no column binding covers.

A normative GeoArrow batch holds exactly a geometry column plus optional temporal, dictionary, and feature-id columns, so an object Result can always carry attributes the batch cannot. "fail" — the default — refuses with unsupported-layout naming the attribute, because dropping a value the caller did not agree to lose is exactly the silent coercion this module exists to prevent. "drop" is the explicit opt-in, and even then every dropped attribute name is reported back on ResultToColumnarBatchResult.droppedAttributes so the loss is stated rather than assumed.