(... because it turned out that the performance bottleneck for an algorithm-focused implementation of bottom-sketches (MinHash sketches) was the parsing of FASTQ files).
Project web page: https://lgautier.github.io/fastq-and-furious/
Efficient and flexible parsing of FASTQ files(*) from Python. It is about as fast as the fastest parser with C extensions, but it is substantially more flexible.