[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Improve autodep8 for R packages



Hi Dirk,

On Tue, Mar 13, 2018 at 03:49:09PM -0500, Dirk Eddelbuettel wrote:
> 
> I looked into this once, and unless I mistaken then
> 
>   -- packages using (CRAN package) testthat do NOT allow running tests once
>      the R package is installed (as they store tests in tests/ which does not
>      migrate)
> 
>   -- packages using (CRAN package) RUnit do (as they store tests in
>      inst/tests or inst/unitTests and everything below inst/ goes into the
>      installed package
> 
> where when I say 'installed' I refer to R CMD INSTALL ... which is of course
> also part of how we build the .dep.

Without having checked a very simple workaround comes to my mind.  We could
install the tests to /var/lib/R/<package>/tests and symlink from /usr/...
to this location.  Thus the directory structure in /usr/... looks the same
but we could write to /var/lib/...
 
> Autotests would be cool.  AFAIK the R community does not (yet) support is
> consistently enough.  Can you point to a r-cran-* package leaning on
> r-cran-testthat that does?   I may have started my analysis with random /bad/
> spot checks after all...

See at the end of this mail the output of

    grep r-cran-testthat r-*/debian/tests/control | sed 's#/.*##' | sort | uniq

on my local clones.  In (probably) all cases where I've written the
tests the trick was to keep a copy of the tests dir in
/usr/share/doc/PACKAGE and use a wrapper script (also there) that is
copying the tests to a temporary dir.

> (There is certainly no use pattern in R land for testing once installed.
> Testing while writing / submitting package is common but needs source.)

In autopkgtests per definition the source is available.

Kind regards

       Andreas.

 
r-bioc-annotationfilter
r-bioc-biomformat
r-bioc-metagenomeseq
r-bioc-phyloseq
r-bioc-savr
r-cran-adegenet
r-cran-adegraphics
r-cran-adephylo
r-cran-afex
r-cran-assertthat
r-cran-batchjobs
r-cran-bayesfactor
r-cran-bbmisc
r-cran-bbmle
r-cran-bindr
r-cran-bindrcpp
r-cran-blob
r-cran-bold
r-cran-broom
r-cran-checkmate
r-cran-cli
r-cran-crayon
r-cran-crul
r-cran-curl
r-cran-data.table
r-cran-dbitest
r-cran-dbplyr
r-cran-desc
r-cran-devtools
r-cran-dimred
r-cran-dotcall64
r-cran-downloader
r-cran-dplyr
r-cran-evaluate
r-cran-fail
r-cran-fauxpas
r-cran-fitcoach
r-cran-futile.logger
r-cran-ggplot2
r-cran-ggvis
r-cran-glue
r-cran-gower
r-cran-gtable
r-cran-hms
r-cran-htmltools
r-cran-httr
r-cran-igraph
r-cran-isoweek
r-cran-jsonlite
r-cran-lava
r-cran-lavaan
r-cran-lazyeval
r-cran-lexrankr
r-cran-lubridate
r-cran-magrittr
r-cran-maldiquant
r-cran-maldiquantforeign
r-cran-matchit
r-cran-memoise
r-cran-mlmrev
r-cran-modelmetrics
r-cran-munsell
r-cran-natserv
r-cran-openssl
r-cran-optparse
r-cran-phangorn
r-cran-phylobase
r-cran-pillar
r-cran-pkgmaker
r-cran-plotly
r-cran-plyr
r-cran-polycub
r-cran-praise
r-cran-prettyunits
r-cran-progress
r-cran-proto
r-cran-purrr
r-cran-qtl
r-cran-r6
r-cran-rcpproll
r-cran-recipes
r-cran-rentrez
r-cran-repr
r-cran-reshape2
r-cran-ritis
r-cran-rlang
r-cran-rlist
r-cran-rncl
r-cran-rnetcdf
r-cran-rnexml
r-cran-rotl
r-cran-rprojroot
r-cran-rredlist
r-cran-rsdmx
r-cran-rsqlite
r-cran-scales
r-cran-seroincidence
r-cran-shiny
r-cran-shinyjs
r-cran-solrium
r-cran-sourcetools
r-cran-stringr
r-cran-surveillance
r-cran-taxize
r-cran-tibble
r-cran-tidyr
r-cran-tidyselect
r-cran-tikzdevice
r-cran-tm
r-cran-treescape
r-cran-treespace
r-cran-triebeard
r-cran-urltools
r-cran-utf8
r-cran-v8
r-cran-webmockr
r-cran-webutils
r-cran-wikidatar
r-cran-wikipedir
r-cran-wikitaxa
r-cran-withr
r-cran-worrms
r-cran-xml2
r-cran-yaml
r-cran-zelig

-- 
http://fam-tille.de


Reply to: