Closed luizirber closed 4 years ago
Wow, I was just going to suggest this (I am trying out nibbler for some bioinformatics applications), and ran into this issue while testing it. Please do!
Wow, I was just going to suggest this (I am trying out nibbler for some bioinformatics applications), and ran into this issue while testing it. Please do!
Hah, cool! Do you want to test the multigz
branch from #28 and see if it works for you? Or, given that you have more experience with BGZF, do you see any drawbacks of using MultiGzDecoder
by default?
I do not see any or experienced any drawbacks of using MultiGzDecoder
by default. I haven't done any serious bench marking, but if you work with Illumina data fresh from the machine MultiGzDecoder
is your only option. I will test the multigz
branch and will let you know what I find.
The multigz
branch works great for me.
2.1.1 released with this change, thanks for trying it @hcdenbakker!
Should we replace the
GzDecoder
from flate2 withMultiGzDecoder
(which would also support BGZF)?Are there other consequences to use
MultiGzDecoder
with regulargzip
files? (apparently not, it didn't brake any of our tests)from https://github.com/onecodex/needletail/pull/45#issuecomment-651235214
Also connected to #24, since
niffler
users might want to use specific backends instead of our default choices.