Bubble up some more errors

Last week I ran into a situation where it would be nice to have some extra error handling.

Essentially I was running nf-test on a number of subworkflows and modules for our internal modules repo, at some point the nf-core modules cache (the .git database) got corrupted, potentially me spamming too many runs over too short a period of time.

After this, the runner would stop at Installing nf-core modules.... at one point I left this overnight (prior to realising the cache was corrupted) and it was still at the same point. It wouldn't properly die until I exit the program.

Running later on with nf-core to install a module resulted in a an error that better led to the issue and I understand now, that many of these commands are actually wrappers for an underlying nf-core tools command. Which has taken me to where I think the error comes from...

Ignore the inlines pointing to specific lines, I was trying to link to the whole function.

https://github.com/nf-core/nft-utils/blob/9a9092107021fed6262cad539c079079ce1a399c/src/main/java/nf_core/nf/test/utils/NfCoreUtils.java#L47

https://github.com/nf-core/nft-utils/blob/9a9092107021fed6262cad539c079079ce1a399c/src/main/java/nf_core/nf/test/utils/NfCoreUtils.java#L111-L121

The error of nf-core crashing due to a corrupted nf-core-modules repo seems to fall outside of the current error catching.

I'll try and re-create the situation for more details.

Start with src/main/java/nf_core/nf/test/utils/NfCoreUtils.java, especially the linked sections around lines 47 and 111-121, and trace how nf-core module installation errors are caught. Reproduce or inspect the corrupted nf-core-modules cache scenario if possible. Done means a corrupted cache produces a surfaced error instead of leaving the runner stuck at module installation.