Does the

Channel.fromSamplesheet()

factory support an

index

? So can I somehow straightforwardly add the line number of the sample sheet entry to the

meta

map? I am asking because we would like to ensure that even if a sample ID is specified twice, we have a unique technical ID and combining it with the line number seemed like a straightforward approach. At first, I just wanted to

merge()

the channel with a numeric

Channel.of(0…length of sample sheet)

, but apart from my inability to get that to work (because I fail to convert the value channel

Channel.fromPath(file("samplesheet.csv")).countLines()

to an Integer), this will not be deterministic and thus cause issues when resuming the pipeline.

I don’t believe

fromSamplesheet

can inject an index itself (feature request maybe!?) but I think the output from it might be deterministic and then something like the following may do the job?

ch = Channel.fromSamplesheet()

ch  | toList()
    | flatMap { it.withIndex().collect{ entry, idx -> [ entry[0] + [index: idx], entry[1] ] } }
    | view

[[id:1, index:0], a]
[[id:2, index:1], b]
[[id:3, index:2], c]

Thanks a lot, I will experiment in that direction!

Alternatively you could also migrate to nf-schema which has a

samplesheetToList()

function that converts the samplesheet to a list. You can then add a unique ID based on the list index and convert that list to a channel using

Channel.fromList()