Travis Tests for ExoSeq#13
Conversation
| python: | ||
| - "2.7" | ||
|
|
||
| node_js: "node" |
There was a problem hiding this comment.
No :-) I'll remove that, just copied some ideas from Sarek ;-) Why do you even have it there?
There was a problem hiding this comment.
It's for a bonus package thing that I came across which prints the amount of time taken between each line of stdoutput. So a way to see how long each pipeline step is taking.
It was a bit of a faff though and made the Travis output kind of hard to read (looks lovely locally).
https://github.com/paypal/gnomon
https://github.com/SciLifeLab/Sarek/pull/572/files
There was a problem hiding this comment.
Good catch, totally forgot that I had it there.
Thanks
The other problem with gnomon is that it was hiding the fact that there were problems in the test.
There was a problem hiding this comment.
The other problem with gnomon is that it was hiding the fact that there were problems in the test.
This is fairly simple to fix with /bin/bash -euo pipefail (this is in all of our pipelines because piping to samtools etc had the same effect of changing the exit code).
| - cd /tmp/nf-core-tools | ||
| - pip install --user -e . | ||
| # Get Kit Files | ||
| - wget -O kits.tar.bz2 https://qbic-intranet.am10.uni-tuebingen.de/owncloud/index.php/s/Qvku3etEqb3PW58/download |
There was a problem hiding this comment.
Maybe time to start playing with https://github.com/nf-core/test-datasets ?
There was a problem hiding this comment.
I just wondered how we should do that. My plan was to use the smallGrch37 by @maxulysse and @szilvajuhos https://github.com/szilvajuhos/smallRef, together with the Kit files and a very tiny small dataset that I already have. So maybe we should quickly have a decision on how to structure the test-datasets repository in a ticket there?
There was a problem hiding this comment.
By the way, https://github.com/SciLifeLab/Sarek-data is now used for testing in Sarek.
cf: SciLifeLab/Sarek#577
|
|
||
| env: | ||
| global: | ||
| - NXF_VER=0.27.6 SGT_VER=2.4.2 |
There was a problem hiding this comment.
Should use the minimum version of Nextflow but also the latest, that way we catch new bugs triggered by new releases.
|
Merging this in, theres too much to do to keep opening single PRs for now... Will split up tasks in smaller pieces after this one and once the tests work properly... Also the way to do tests has changed quite dramatically and will be much cleaner now that I have more experience with this :-) |
3 participants
Started work on a separate branch to have proper Travis CI tests for ExoSeq
#4