[TimeSeries] split handle non-successive chunks inconsistently #3941

Describe the current behavior

split() works for some chunk size but fails for others when chunks are not successive.

Example

chunk1: offset = 0, values = [1, 2, 3]
chunk2: offset = 6, values = [7, 8]
# timeSeries
timeSeries (chunk1, chunk2): [1d, 2d, 3d, NaN, NaN, NaN, 7d, 8d]

• split(2) works and returns non successive chunks:

offset = 0, values = [1, 2]
offset = 2, values = [3]
offset = 6, values = [7, 8]

These chunks are not successive: after offset 2 with size 1, the next expected offset should be 3, not 6.

• split(3) works also.

• but split(4) fails with:

java.lang.IllegalArgumentException: Chunks are not successive. First offset is 0 and first size is 3; second offset should be 3but is 6

Describe the expected behavior

The behavior should be consistent.

Either:

• non successive chunks are supported, and split() should work for all valid split sizes, or
• non successive chunks are not supported, and split() should fail consistently.

Describe the steps

1. Create a time series with two non successive chunks:
   Example

DoubleDataChunk chunk1 = new UncompressedDoubleDataChunk(0, new double[]{1d, 2d, 3d});
DoubleDataChunk chunk2 = new UncompressedDoubleDataChunk(6, new double[]{7d, 8d});
StoredDoubleTimeSeries timeSeries = new StoredDoubleTimeSeries(metadata, chunk1, chunk2);

2. Call split(2)
   this works
3. Call split(3)
   this works also
4. Call split(4)
   this fails

Environment

Version: 7.2.1

Relevant Log Output

No response

Extra Information

No response