<?xml version="1.0"?>
<tool id="windowmasker_mkcounts" name="WindowMasker_mkcounts" version="1.0">
    <description>Construct WindowMasker unit counts table</description>

    <macros>
        <import>windowmasker_macros.xml</import>
    </macros>

    <expand macro="requirements" />

    <command detect_errors="exit_code">
<![CDATA[
    @OPTIONAL_PARAM_FUNC@

    windowmasker -mk_counts -infmt fasta

        $optional_param("-genome_size", $adv.genome_size)
        $optional_param("-t_low",       $adv.t_low)
        $optional_param("-t_high",      $adv.t_high)
        $optional_param("-unit",        $adv.unit)

        -sformat ${adv.sformat} -checkdup ${adv.checkdup}
        -in "${fasta_input}" -out "${mkcount_output}"
]]>
    </command>
    <inputs>
        <param name="fasta_input" type="data" format="fasta"
                label="FASTA sequence file" />

        <section name="adv" title="Advanced options" expanded="false" >
            <param name="checkdup" type="boolean" checked="false"
                    truevalue="true" falsevalue="false"
                    label="Check input file for potentially duplicated sequences"
                    help="-checkdup" />

            <param name="genome_size" type="integer" label="Genome size"
                    min="0" optional="true"
                    help="Use this genome size to compute the unit length" />

            <param name="sformat" type="select" label="Unit counts format"
                help="The format the unit counts data should be generated" >

                <option value="ascii">Plain text (ascii)</option>
                <option value="oascii" selected="true">Optimized text (oascii)</option>
                <!-- Omit support for non-portable binary formats (binary, obinary) -->
            </param>

            <param name="t_low" type="integer" label="T_low"
                    min="0" optional="true"
                    help="Save only units that appears at least this many times" />

            <param name="t_high" type="integer" label="T_high"
                    min="0" optional="true"
                    help="Set the maximum count value for units" />

            <param name="unit" type="integer" label="Unit length"
                    min="1" max="16" optional="true"
                    help="Unit length used to mask the genome" />
        </section>
    </inputs>
    <outputs>
        <data name="mkcount_output" format="txt" />
    </outputs>
    <tests>
        <test>
            <!-- Test mkcounts with default settings -->
            <param name="fasta_input" value="chr4_part.fa" ftype="fasta" />
            <output name="mkcount_output" file="chr4_part.oascii.counts" />
        </test>
        <test>
            <!-- Test mkcounts with custom genome_size -->
            <param name="fasta_input" value="chr4_part.fa" ftype="fasta" />
            <param name="genome_size" value="500000" />
            <output name="mkcount_output" file="chr4_part.500k_size.counts" />
        </test>
        <test>
            <!-- Test mkcounts with ascii output format -->
            <param name="fasta_input" value="chr4_part.fa" ftype="fasta" />
            <param name="sformat" value="ascii" />
            <output name="mkcount_output" file="chr4_part.ascii.counts" />
        </test>
        <test>
            <!-- Test mkcounts with custom unit settings -->
            <param name="fasta_input" value="chr4_part.fa" ftype="fasta" />
            <param name="t_low" value="3" />
            <param name="t_high" value="20" />
            <param name="unit" value="9" />
            <output name="mkcount_output" file="chr4_part.custom_unit.counts" />
        </test>
    </tests>
    <help>
<![CDATA[
**What it does**

This tool runs `stage 1 <https://www.ncbi.nlm.nih.gov/IEB/ToolBox/CPP_DOC/lxr/source/src/app/winmasker/>`_
of the WindowMasker analysis to produce a unit counts file for a genome assembly.

    ]]></help>

    <expand macro="citations" />
</tool>