Huge thanks to @Rogdham for restoring pyzstd library development! Earlier I had almost given up hope for pyzstd survival, and therefore for further antidb maintenance. Now I guess there's no cause for concern.
- Two new built-in sorting rules. They are much more narrowly targeted, simpler, and potentially faster than the natural rule. The first one splits the line and fetches the desired cells. The second one handles values like rs12345.
- The built-in sorting rules are no longer static methods, but are now linked by some common attributes.
- Improving mem-index header with sorting settings and fixing its handling.
- Ability to pass arbitrary arguments to a function that parses lines of indexable files.
- A new test where thoroughly checked work with VCF.
- Some refactoring of the tests code.
Contributors
Rogdham