Re: Please use our scripts to save time and enhance quality [Was: [Git][med-team/igdiscover][master] 2 commits: Adjusting build dependencies]
On Fri, Aug 02, 2019 at 11:43:01AM +0200, Steffen Möller wrote:
> Hi Andreas,
>
> Thank you for fixing this and for caring in the first place.
>
> igdiscover is already with us for a llooooonngggg time, just parked in
> the new queue, for so long that I just yesterday updated it to remain
> compatible with newer versions of cutadapt when I rebuilt it for the
> "Debian Med Crash Space" :) - med.functional.domains for those who have
> not followed that thread. So I don't feel too bad about IgDiscover. But
> I do feel bad for https://salsa.debian.org/med-team/python-alignlib
> which I added just a few days ago
I remember Michael has even added code to re-run the script on existing
repositories to set all missing hooks. Wanted to try this but master
branch does not even have a debian/ dir.
> - would you mind to have a pointer to
> that script right on top of https://salsa.debian.org/med-team where the
> "New project" button is?
No, sorry. I do no Salsa hacking. I trust that you will adapt to a
more easy and time saving workflow. You managed with routine-update
after about one year - so you probably will with inject-into-salsa-git.
:-P
> I must admit that I used even used the script
> once but have not really grasped / ignored the nitty-gritty technical
> infrastructure behind it and there was this overhead on getting the
> authentication done.
There is not really any overhead. It now tells you what needs to be
done if not setup properly.
> I am busy with other things, which is also why I just do what is under
> my fingertips and don't reeducate me on scripts / not even think about
> them. It all needs to become simpler.
Making things simpler is the actual point of those scripts. Put these
in your PATH and just type inject<tab> ...
> Just like allowing me to be
> sufficiently functional when my mind is not. After all, we do this
> Debian bits too often only after the working day has already ended. So,
> I propose to start with a more visible pointer to that script right on
> top of everything. And then there will more and more patches against
> that script to help adopting that in some interactive fashion. A first
> patch I would then look into is how to add hooks to projects that are
> already on salsa with the same script.
That was done by Michael as I said above ...
Further enhancements are welcome
Andreas.
> On 02.08.19 06:58, Andreas Tille wrote:
> > Hi,
> >
> > since I have realised the upload of igdiscover to new and no according
> > commit message in our commit mailing list I added the commit hooks
> > manually. There would be no need for manual intervention if everybody
> > who creates a new repository would use
> >
> > https://salsa.debian.org/med-team/community/helper-scripts/blob/master/inject-into-salsa-git
> >
> > This script should be called from a new local git repository and will
> > create the remote one on salsa.d.o and sets all the hooks we want. I
> > would be happy if everybody would prefer this over manual creation of
> > repositories since usually this will not include all settings we want
> > to have.
> >
> > Hope this helps
> >
> > Andreas.
> >
> >
> > ----- Forwarded message from Andreas Tille <gitlab@salsa.debian.org> -----
> >
> > Date: Fri, 02 Aug 2019 03:28:05 +0000
> > From: Andreas Tille <gitlab@salsa.debian.org>
> > To: debian-med-commit@lists.alioth.debian.org
> > Subject: [Git][med-team/igdiscover][master] 2 commits: Adjusting build dependencies
> >
> >
> >
> > Andreas Tille pushed to branch master at Debian Med / igdiscover
> >
> >
> > Commits:
> > 1cc07041 by Steffen Möller at 2019-02-12T09:25:15Z
> > Adjusting build dependencies
> >
> > - - - - -
> > df76dec7 by Steffen Möller at 2019-08-01T23:06:59Z
> > Fixed FTBFS with new cutadapt (>= 2.3)
> >
> > - - - - -
> >
> >
> > 4 changed files:
> >
> > - debian/changelog
> > - debian/control
> > - + debian/patches/minoverlap.patch
> > - + debian/patches/series
> >
> >
> > Changes:
> >
> > =====================================
> > debian/changelog
> > =====================================
> > @@ -1,4 +1,11 @@
> > -igdiscover (0.11-1) UNRELEASED; urgency=medium
> > +igdiscover (0.11-2) unstable; urgency=medium
> > +
> > + * Introduced patch for compatibility with later versions
> > + of cutadapt - upstream has it already in their master branch.
> > +
> > + -- Steffen Moeller <moeller@debian.org> Fri, 02 Aug 2019 00:55:02 +0200
> > +
> > +igdiscover (0.11-1) unstable; urgency=medium
> >
> > * Initial release (Closes: #919718)
> >
> >
> >
> > =====================================
> > debian/control
> > =====================================
> > @@ -38,9 +38,8 @@ Depends: ${python3:Depends},
> > python3-cutadapt,
> > python3-scipy (>=0.16.1),
> > python3-xopen (>=0.3.2),
> > + python3-seaborn (>=0.9),
> > python3-ruamel.yaml
> > -Recommends: python3-seaborn (>=0.8.1),
> > - python3-matplotlib
> > Suggests: snakemake,
> > igdiscover-doc
> > Description: analyzes antibody repertoires to find new V genes
> >
> >
> > =====================================
> > debian/patches/minoverlap.patch
> > =====================================
> > @@ -0,0 +1,14 @@
> > +Index: igdiscover/igdiscover/utils.py
> > +===================================================================
> > +--- igdiscover.orig/igdiscover/utils.py
> > ++++ igdiscover/igdiscover/utils.py
> > +@@ -330,8 +330,7 @@ def find_overlap(s, t, min_overlap=1):
> > + >>> find_overlap('ABC', 'X') is None
> > + True
> > + """
> > +- aligner = Aligner(s, max_error_rate=0)
> > +- aligner.min_overlap = min_overlap
> > ++ aligner = Aligner(s, max_error_rate=0, min_overlap=min_overlap)
> > + result = aligner.locate(t)
> > + if result is None:
> > + return None
> >
> >
> > =====================================
> > debian/patches/series
> > =====================================
> > @@ -0,0 +1 @@
> > +minoverlap.patch
> >
> >
> >
> > View it on GitLab: https://salsa.debian.org/med-team/igdiscover/compare/dc8ee7231d33b6e526845fcb32a407ead20cb1a1...df76dec71814551cc3d7914005f4a2e15635b7be
> >
>
>
--
http://fam-tille.de
Reply to: