Using Bioclipse in Nextflow.

demo.nf

#!/usr/bin/env nextflow
@Grab(group='net.bioclipse.bacting', module='managers-cdk', version='0.0.3')
import net.bioclipse.managers.CDKManager

params.str = "./data.tsv"
params.size = 5000

process splitWikidata {
    input:
    file str from Channel.fromPath(params.str)
    output:
    file 'chunk_*' into smiChunks

    """
    cat $str | tail -n +2 | split -l ${params.size} - chunk_
    """
}

process checkSMILES {
  input: 
  val x from smiChunks

  output:
  stdout results

  exec:
    def cdk = new CDKManager(".");
    new File(x).readLines().each {
      def fields = it.split()
      try {
        // println "x: " + fields[1]
        cdk.fromSMILES(fields[1])
      } catch (Exception exc) {
        println fields[0] + ": " + exc.message
      }
    }
}

results.subscribe {
  println it
}

This code fails for me with this error message:

ERROR ~ Error executing process > 'checkSMILES (1)'

Caused by:
  Could not find matching constructor for: java.io.File(ArrayList)

Source block:
  def cdk = new CDKManager(".");
  new File(x).readLines().each {
    def fields = it.split()
    try {
      // println "x: " + fields[1]
      cdk.fromSMILES(fields[1])
    } catch (Exception exc) {
      println fields[0] + ": " + exc.message
    }
  }

Work dir:
  /home/egonw/tmp/nf/work/29/1cd5958914f3c7f5d5c4a88add0a0e

Tip: view the complete command output by changing to the process work dir and entering the command `cat .command.out`

 -- Check '.nextflow.log' file for details

The problem seems to be that x is an Array, so when I iterate over that first it works:

    x.each { chunk ->
      // println "x: " + chunk
      chunk.readLines().each { line ->
        def fields = line.split()
        try {
          // println "x: " + fields[1]
          cdk.fromSMILES(fields[1])
        } catch (Exception exc) {
          println fields[0] + ": " + exc.message
        }
      }
    }

But this change may explain why it only uses one cpu core?

My fault. The input should be declared as:

  input: 
  val x from smiChunks.flatten()