open_or_gzopen -> compressed_open. Add support for zstd. #937
Conversation
There was a problem hiding this comment.
This is wonderful. I think the only little thing I'd ask is to add a test somewhere. The least-effort place to do this might be in TestIlluminaDir which already has gz, bz2, and lz4 tarballs, and we could just add one more (admittedly though, it's not the most unit-y place to do it).
| @@ -397,14 +443,14 @@ def read_tabfile(inFile): | |||
| ''' Read a tab text file (possibly gzipped) and return contents as an | |||
| iterator of arrays. | |||
| ''' | |||
| with open_or_gzopen(inFile, 'rU') as inf: | |||
| with compressed_open(inFile, 'rt') as inf: | |||
There was a problem hiding this comment.
Since this was originally rU, can we have the new invocation include newline=None as a kwarg? I think this is trying to defend against the incredibly common scenario of users providing SampleSheets created by MS Excel.
|
Oh one more thing, edit the description of this PR (or do it in a commit message) to include the magic words "closes (issue x) and closes (issue y)" or "fixes" or "resolves".. simply referencing the issues w/o the magic words won't do it. https://help.github.com/en/articles/closing-issues-using-keywords |
yesimon
force-pushed
the
sy-zst
branch
2 times, most recently
from
62c90a1 to
4f422c0
Compare
Zstd has a nonstandard python api which requires customization. Also allows for multithreading compression.
Zstd has a nonstandard python api which requires customization. Also allows for multithreading
compression.