Improvements to benchmarks about streamly-lz4 HOT 3 CLOSED

remove the compress.decompress benchmarks. Usual use case is to either compress a stream or decompress a stream, it should be enough to measure these two separately.

We can only decompress a compressed stream so a standalone decompress is not possible unless we use a compressed file as input.

Make all the input files of equal size so that we can compare the benchmarks. We can replicate the data to make them all 5MB or 10MB.

The best way to this, I feel is using a simple script after downloading the corpora.

from streamly-lz4.

Yes, we should compress and use the compressed file. Basically we want to just leave out the compress from the benchmark measurement.

from streamly-lz4.

See #18 and #23

from streamly-lz4.