Possible API improvements after implement custom database

[harrison314]

Hi, I created a time series database prototype using ZoneTree (it's not ready for publication yet, but I can provide the source codes using some private canal) and I have to say that it surprised me with its speed. Great job.

I have some ideas for adding an API that would be useful in implementing a database built on ZoneTree.

Range iterator

Description: Enable iteration in the range of entered keys from-to. This functionality is going to be implemented, but I have to save somewhere the correct comparator that ZoneTree has in the configuration, it would be more convenient to extend the API.
UseCase: Insertion by secondary index, or iteration in a range of values. I assume that this functionality would also be useful for people who do not implement a database.

Gets or estimate key count between two keys

Description: If it is possible to find out or estimate the number of keys between two specified keys in the time complexity $O(2.log(n))$, then an API that returned this data would be useful. I really don't know if such a functionality will be implemented, because I didn't find it even in similar databases (LevelDb, RoksDb, FoundationDb).
UseCase: When using secondary indexes, it is possible to decide which index to use based on the (also approximate) number of keys.

[koculu]

Thank you for your suggestions. We can discuss this on discord channel:

[koculu]

You can access comparer, key and value serializers from IZoneTree interface now:
Expose comparer, key and value serializers on ZoneTree API

You can implement your own wrapper to iterate your desired boundary.

[harrison314]

Here https://github.com/harrison314/YATsDb/tree/main/src/Src/YATsDb.Core is an implementation of the Time Series database I mentioned above.