Re: Packaging Hybrid R/Python Software
Hi Ben,
On Wed, Nov 18, 2020 at 06:38:00AM +0000, Morledge-Hampton, Benjamin Stuart wrote:
> This is extremely helpful! Thank you so much. I've had a chance to work with some of the tools you described, and I do have a couple of quick questions:
You are welcome.
> I have run the function you described and noticed that it performed similarly to the dh-make-R command I was using previously. Is this command invoking prepare_missing_cran_package at some point? Is this an acceptable way to discover and package dependencies?
Yes, prepare_missing_cran_package is a wrapper around dh-make-R.
> Also, I have noticed that this process creates two directories in the ../CRAN_prospective directory for each dependency. One is named simply after the dependency (e.g. "lomb") while the other seems to be named using the debian standard for R packages (e.g. "r-cran-lomb").
That's strange. I tried to reproduce this here but I get only
r-cran-lomb (which is the right one and should contain a Git repository
including the debian/ dir).
> Adding to my confusion, the directory the simpler name contains what I would expect to be standard for building a debian package (i.e. the debian directory along with the usual R package structure), but the other directory only contains a .git directory. What is the rationale behind the creation of both of these directories? My inclination is to only use the one with the created debian directory and ignore the other, but I am concerned by the presence of this second directory. Am I meant to combine the directories somehow before running the script to push the repository to salsa? My brief read-through of the script you sent me seems to indicate that this is the case. (Though admittedly, I'm a bit of a novice when it comes to bash scripting, so I may be missing something)
I tried to reproduce this an all went well. I pushed the result to
https://salsa.debian.org/r-pkg-team/r-cran-lomb
You can see an example what needs to be changed in a manual commit.
Feel free to add your user ID to the Uploaders field and send an
ITP for that package.
Since you've read the script I would be really happy if you could
add some `set -x` and check where the problem might occure on your
side. There are no reports about that kind of failure before.
> Thanks again for all your help. It means a lot to me. I feel like I'm making real progress now on this project, and it's very exciting!
You are welcome
Andreas.
--
http://fam-tille.de
Reply to: