Skip to content

Inputs

This page summarizes the different input types available for Nextflow processes. Use these declarations to specify how data is provided to your process scripts.

val

Declare a variable input. The received value can be any type and is made available to the process body as a variable with the given identifier.

  • Syntax: val(identifier)
  • Input: Any value.
  • Usage: 
    process example {
    input:
    val(x)
    script:
    """
    echo $x
    """
}

See also: Outputs.md val, the output-side counterpart for emitting a value from the process body.

path

Declare a file input. The value should be a file or collection of files and is staged into the task directory.

  • Syntax: path(identifier | stageName)
  • Input: File or collection of files.
  • Options:
    • arity: Number or range of expected files (e.g., 1, 1..*). Task fails if invalid.
    • name / stageAs: Name or pattern for the file(s) in the task directory.
  • Usage: 
    process example {
    input:
    path('input.txt', stageAs: 'renamed.txt')
    script:
    """
    cat renamed.txt
    """
}

See also: Outputs.md path, the reverse declaration for emitting files instead of receiving them — shares the same arity option. Directives.md stageInMode controls how these files are staged into the task directory. Operators.md splitCsv/splitFasta/splitFastq/splitText/splitJson are the usual way to turn a path input's contents into a per-record channel.

env

Declare an environment variable input. The value should be a string and is exported to the task environment as the given variable name.

  • Syntax: env(name)
  • Input: String.
  • Usage: 
    process example {
    input:
    env(MY_VAR)
    script:
    """
    echo \$MY_VAR
    """
}

See also: Outputs.md env, the output-side counterpart for capturing an environment variable from the task.

stdin

Declare a stdin input. The value should be a string and is provided as standard input to the task script. Only one stdin input can be declared per process.

  • Syntax: stdin
  • Input: String.
  • Usage: 
    process example {
    input:
    stdin
    script:
    """
    cat
    """
}

See also: Outputs.md stdout, the output-side counterpart for capturing a task's standard output.

tuple

Declare a tuple input. Each argument should be an input declaration (val, path, env, or stdin). The received value should be a tuple matching the declaration.

  • Syntax: tuple(arg1, arg2, ...)
  • Input: Tuple with elements matching the declarations.
  • Usage: 
    process example {
    input:
    tuple(val(x), path(y))
    script:
    """
    echo $x
    cat $y
    """
}

See also: Outputs.md tuple, the output-side counterpart. Operators.md groupTuple/join/combine/cross are the main operators for building and manipulating tuple-shaped channels feeding into this declaration.