Hi, I have a hopefully quick question. If I have a large bash script inside my process, is there a way I can have it in a separate

.sh

file and call it from the script, or directly import the script instead of the

script

block?

Well you can use

template

(docs) but the template function is generally advised against as noted in the docs. Alternatively you could put it in your pipeline

bin

and then run

my script.sh

in the

script

block. If you are using

wave

you could also put it in a module binary

If you put the script in the bin folder, make it executable and add a shebang, your process should automatically find it and be able to run it

Thanks! The

bin

folder in the module? Is that added to the

PATH

variable?

I currently only have one bin folder at the project root and that's automatically added to the PATH and I can imagine it would be the same for a bin folder in the module directory. I am still figuring out nf-core best practice

The pipeline

bin

is top level (alongside the pipeline

main.nf

) and "just works" (automatically added to path and exposed to processes/modules). A

bin

at the module level is covered by the module binaries link in my first response and (unfortunately imo) only works if you are using

wave

currently

Thanks! Unfortunately, we are not using wave yet. But thanks for the suggestions 🙂

Module binaries work without wave when not using a cloud executor.