Current UNTAR scheme inefficent and can cause overwriting for database sheet input

[jfy133]

Description of the bug

Currently we 'blindly' untar every row in the database samplesheet. If someone has submitted the same database but with different parameters, we will untar that database repeatedly. If we go to publish this, we we will repeatedly overwrite the same files - the latter causes an error with strict

We should improve the efficiency of untarring to only do it once per file - so likely we have to group by file name, untar, then spread again with the metas to separate the differnet parameters.

Picked up by setting nextflow.enabled.strict = true in test config and running the test_malt profile

Command used and terminal output

No response

Relevant files

No response

System information

No response

[jfy133]

#461 done