Partial Interval Selection - what to do

If you slice the continuous index to include only part of an interval what should we do with an interval. The current behavior:

```python
from linked_indices.interval_index import DimensionInterval
from linked_indices.util import interval_dataset
ds = interval_dataset(interval_dim="intervals")
ds = (
ds.drop_indexes(["time", "intervals"])
.set_xindex(["time", "intervals"], DimensionInterval)
.drop_vars("word")
)
print(ds.sel(time=slice(500,1500)))
```

```
Size: 50kB
Dimensions: (C: 10, time: 201, rgb: 3, intervals: 2)
Coordinates:
* time (time) int64 2kB 500 505 510 515 520 ... 1480 1485 1490 1495 1500
* intervals (intervals) interval[int64, left] 32B [0, 1000) [1000, 2500)
Dimensions without coordinates: C, rgb
Data variables:
data (C, time, rgb) float64 48kB 1.0 0.0 0.0 1.0 ... 1.0 0.2806 0.0
Indexes:
┌ time DimensionInterval
└ intervals
```

makes decent sense. But you could image that the interval should maybe be truncated to match the actual time values