When parsing samplesheets, group values into ordered lists

[adamrtalbot]

It's a common pattern in nf-core modules to use an ordered list instead of tuple elements. i.e. instead of:

input:
    tuple val(meta), path(input1), path(input2)

You use:

input:
    tuple val(meta), path(inputs)

We could incorporate some method of parsing the schema to use in .fromSamplesheet, e.g.:

{
    // rest of schema above
    "lists": [
        "fastq_1",
        "fastq_2"
    ]
}

Would coerce to an ordered list, with nullable values not being included.

[adamrtalbot]

See rnaseq for example of having to parse it afterwards: https://github.com/drpatelh/nf-core-rnaseq/blob/8d9ccac23cab6c3f14149a4e8659eaea3120c6e3/workflows/rnaseq.nf#L192-L200

[nvnieuwk]

We concluded adding more examples on how to use .map would be the better solution here instead (https://nfcore.slack.com/archives/C056RQB10LU/p1690214043545869)

[mirpedrol]

Will close this issue as we added the documentation with #82, thanks!