Having trouble using collate() with formatter()

[Tom Harrop]

Hello,

I am struggling to get my head around the collate syntax using formatters.

I've made an example where I try to collate ['data/aa_data', 'output/aa/analysis1/results.txt'] to ['output/aa/analysis2/results.txt'].

I have tried a few variations, this is what I have at the moment:

from ruffus import *


def orig(output_files):
    print("readfile = ", output_files)
    with open(output_files, "w") as oo:
        pass


def next_task(input_files, output_parameters):
    print("input_parameters = ", input_files)
    print("output_parameters = ", output_parameters)
    for f in output_parameters:
        with open(f, "w") as oo:
            pass


def main():
    main_pipeline = Pipeline.pipelines["main"]

    data = main_pipeline.originate(name="data",
                                    task_func=orig,
                                    output="data/aa_data")

    result = main_pipeline.originate(
        name="result",
        task_func=orig,
        output="output/aa/analysis1/results.txt")

    main_pipeline.collate(
        name="next_task",
        task_func=next_task,
        input=[result, data],
        filter=formatter(),
        output=["output/{subdir[0][1]}/analysis2/results.txt",
                "output/{basename[1][0:2]}/analysis2/results.txt"])

    pipeline_printout_graph("flowchart.pdf", "pdf")

    main_pipeline.run(verbose=5)

if __name__ == "__main__":
    main()

According to my understanding of the Ruffus manual I expect both {subdir[0][1]} and {basename[1][0:2]} to resolve to "aa", so this should collate the output. However I am getting the following message:

Task = 'next_task' 
       !!! No jobs for this task.
       Are you sure there is not a error in your code / regular expression?
       Job Warning: Input substitution failed:
         Missing key = {} in 'output/{basename[1][0:2]}/analysis2/results.txt'.
               input =  '',
              filter = formatter(['output/aa/analysis1/results.txt'])..
       Job Warning: Input substitution failed:
         Missing key = {} in 'output/{basename[1][0:2]}/analysis2/results.txt'.
               input =  '',
              filter = formatter(['data/aa_data'])..

Can anyone tell me what I'm missing?

Thanks for reading,

Tom