Better handling of minute init --input parameter

[cnluzon]

I mistakenly left out validation of --input on minute init. So the command would run, and generate invalid groups.tsv file. So now:

• An error is printed if --barcodes is provided but --input is not.
• An error is printed if the value of --input does not match the prefix of any of the .fastq.gz files in the --reads directory.

I think the make_groups_from_barcodes_and_reads should be the function raising an exception if input is not valid, but I also think it is not so nice if the user gets an output with an entire Traceback and all, so I put it on a try/catch block, but still printed the message coming from it in a CommandLineError, which I am unsure if it is the best way to do.

[marcelm]

I think the make_groups_from_barcodes_and_reads should be the function raising an exception if input is not valid, but I also think it is not so nice if the user gets an output with an entire Traceback and all, so I put it on a try/catch block, but still printed the message coming from it in a CommandLineError, which I am unsure if it is the best way to do.

This is IMO the right way to do it: The function itself should juts raise an appropriate exception, and then the actual command-line code (main or run_init in this case) catches it and converts it to a CommandLineError. This separates responsibilities nicely: The function doesn’t need to even know that it is being run from the command line. And main() can take the exception and its message and make it nicer to read.