Re: R package CI test failures

[Gordon Ball <gordon@chronitis.net>]

Returning a few weeks later: CRAN packages.

r-cran-adegraphics:
r-cran-surveillance:

waiting for r-cran-spdep


r-cran-afex:

run-unit-test script hides output, so you don't see what goes wrong
missing test dependency on r-cran-dplyr
missing test dependency on mlmrev (not packaged) for the "conception"
dataset


r-cran-bbmisc:
r-cran-futile.logger

These have R-level errors in the test suites which I can't obviously
understand the causes of.



r-cran-doparallel

fixed the run-unit-test in SVN not to hide output, which reveals:
missing test dependency on r-cran-caret
missing test dependency on mlbench (unpackaged)
attempts to write a summary to the (read-only) package installation
directory - this is hard-coded in doRUnit.R and probably needs a patch


r-cran-hexbin

run-unit-test tries to run R extracted from the vignette (fails due to
missing unpackaged dependency marray), but doesn't run test/*.R (which
should succeed)
the package doesn't actually install run-unit-test or test/*


r-cran-matrixstats

Vcs-Browser link is broken
Fixed in SVN an error in run-unit-test (directories in test/) - should
(hopefully) now work


r-cran-msm

the errors are fixed by upgrading to 1.6.1 (tested)


r-cran-sp

missing test dependency rgeos




Addendum:

I'm not very familiar with svn-buildpackage; I use the following
gbp-buildpackage ~/.gbp.conf to build and run tests using an schroot;
does anyone have an equivalent for svn-buildpackage?

---8<---

[buildpackage]
builder = sbuild --no-clean-source --arch-all --source --force-orig-source
export-dir = ../build-area/
posbuild = adt-run --changes $GBP_CHANGES_FILE --- adt-virt-schroot
unstable-amd64-sbuild

---8<---

(this does end up entering and installing packages in the chroot twice,
which is a bit annoying, but it at least ensures a clean environment
each time)

Gordon

On 15/05/16 13:09, Andreas Tille wrote:
> On Sat, May 14, 2016 at 03:56:40PM +0200, Andreas Tille wrote:
>>>
>>> Patch (git format-patch for the repository) is attached. It
>>> suppress_test_writing_to_usr.patch, which removed some of the pre-amble
>>> to each vignette with one which modifies the paths the vignettes check
>>> for their data files. They should now build, but the example data has to
>>> be copied to the pwd (run-unit-test has been updated to do this).
>>
>> I'll check the patch soon.  Upgrade of the package is pending anyway but
>> there are some missing preconditions for the new version.
> 
> Patch commited to Git, waiting for preconditions for new version.
>   
>>>>> r-bioc-limma
>>>>> ------------
>>>>>
>>>>> Compared stdout of the test script to a saved copy. It's probably
>>>>> sufficient to just run the script and test the exit code, since this
>>>>> seems to fail on trivial differences.
>>>>
>>>> I might have a look.  Needs to be updated anyway.
>>>
>>> Attached a replacement run-unit-test script which does this.
>>
>> I'll check this out.
> 
> Patch commited to Git.  Will upload it with new upstream version.
>   
>>> In addition, a re-do of the r-bioc-snpstats run-unit-test is attached
>>> (three vignettes should be blacklisted, but the current script only gets
>>> two of them).
>>
>> I'll check soon. 
> 
> The patch was working around a missing r-cran-hexbin.  I've just
> uploaded r-cran-hexbin and ftpmaster seems to have become very speedy
> with r-cran packages so this is available now.
> 
> So I think we could stick to the old test, right?
> 
> Kind regards
> 
>       Andreas.
>