Hello, I have a process whose emits a file (e.g., let's say

sample/subdir/<INDEX>/counts.h5ad

) for each sample and each index. Then, I have a process which needs to access all these counts files for each sample. I'm waiting for all the indices before putting them all as an input of the other process, i.e. it will have an input

path(counts_files)

for a given sample. The problem is, all counts files end up in the current workdir (with the same name) instead of keeping the original hierarchy in

sample/subdir/<INDEX>

. How to keep this hierarchy? I tried to use

stageAs

, but couldn't find the right way to make it work

can you rename the files that the process emits? Maybe with a prefix? The prefix could be sample-specific, e.g. meta.id?

Yes this is definitely possible! Is it the only way? If so, I can do that, but if there is a direct solution it would be super nice 🙂

If the naming of the files is of no relevance in the work dir, you could also use:

input:
tuple val(meta), path('input/counts_*.h5ad')

where * will be substituted by the file count, i.e 1,2,3,4,...

No unfortunately I need to keep the original file name with the right index

Not sure. Maybe ask in the nextflow slack?

Actually I'll stick with the file renaming - it adds quite some lines of code, but it's working well, thanks 🙂