Luxar Viewer API Documentation - v2026.10.19
    Preparing search index...

    Module data/loaders/spatial-query/prefetch-ranges

    Shared cache-warming read for the Points / Lines / GSplats spatial-index loaders' prefetchChunks.

    For every array, computes the DISTINCT chunk coordinates the first-axis ranges touch (all chunks over the trailing axes) and warms each one once via the L0 proxy's warmChunk: the chunk is fetched (L1/L2) and decoded into L0, and nothing else happens — no zarrita get(), so no output selection is allocated and filled only to be discarded, and no read of the owning loader's demand-load hooks (its per-update abort signal and residency probe stay untouched by prefetch traffic). An array not wrapped by L0 falls back to a bare getChunk (fetch + decode, still no output assembly).

    Deliberately NOT a RangeLoader.loadDirectTyped call. Prefetch warms FUTURE frames. Callers may supply their own speculative-work signal (for example a ladder lookahead cancelled by a view change); it is intentionally distinct from the active demand update's signal. Miss decodes count under decode.count.prefetch unless that signal is tagged with another origin (e.g. 'lookahead').

    chunkCoordsForRange
    prefetchRangesIntoCache