mkrd opened this issue 2 years ago · 1 comments
Performance profiling shows that seek_index_through_value is the bottleneck for fast partial reading. Consider rewriting the function in C, which will probably be much faster.
First testing did not yield faster performance