Class BarrageCopyKernel.Runs
java.lang.Object
io.deephaven.extensions.barrage.chunk.BarrageCopyKernel.Runs
- Enclosing interface:
BarrageCopyKernel
Where every row of one output column comes from, as stretches that are contiguous in both the output and the
delta chunk they come from: for each, the first output position, the encoded origin of its first row (see
BarrageCopyKernel.originOffset(int, boolean)) and how many rows follow contiguously.
No column data lives here, only the positions that say where to find it. The caller appends the runs in output order and a kernel reads them; nothing consumes or rewrites them, which is what lets columns whose rows come from the same places share one. Row sets are range-compressed and updates arrive in ranges, so a run usually covers many rows and building this costs proportionally to ranges rather than rows.
-
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptionvoidadd(long destination, long origin, long length) Append one run:lengthrows starting at output positiondestination, whose first row is at encoded originoriginand whose remaining rows follow it contiguously.longHow many output rows these runs account for between them.
-
Constructor Details
-
Runs
public Runs()
-
-
Method Details
-
totalRows
public long totalRows()How many output rows these runs account for between them. -
add
public void add(long destination, long origin, long length) Append one run:lengthrows starting at output positiondestination, whose first row is at encoded originoriginand whose remaining rows follow it contiguously.
-