Hi, I’m adding a new nf-core module that uses an R package template: https://github.com/saraamenasantamaria/modules/tree/add-updhmm-vcfcheck-module/modules/nf-core/updhmm/vcfcheck A few days ago, I asked about how to add topics. Since this module was included in a pipeline for testing, I ran lint locally from there and it passed without errors. However, running lint now from modules/nf-core produces an error related to the topics. I’ve checked the documentation and implemented it as I understood, but I’m not sure how to handle this correctly so that it passes all lint checks. Thanks for your help!

essentiall what you need instead of:

path "versions.yml", emit: versions, topic: versions

is:

tuple val("${task.process}), val(bioconductor-updhmm), eval("Rscript -e 'library(UPDhmm); cat(as.character(packageVersion('UPDhmm')))") topic:versions emit: versions_updhmm

The problem is that if I use

eval

, I get an error because it cannot be used with an R script.

Interesting, @Louis Le Nézet did that here https://github.com/nf-core/modules/blob/64ab14a6905e5c9d649f61e2757a1e600dbdb8e0/modules/nf-core/quilt/quilt/main.nf#L19 (https://github.com/nf-core/modules/pull/9917)

But it uses a Bash script, not an R template. It runs R commands from Bash. However, my code is longer, and for that the documentation recommends using templates.