<?xml version="1.0"?>
<tool id="sequence_content_trimmer" version="0.2.3" name="Sequence Content Trimmer">
  <description>trim reads based on certain bases</description>
  <command detect_errors="exit_code"><![CDATA[
  #if $paired.is_paired and (('fasta' in $input1.extension and 'fastq' in $input2.extension) or \
      ('fastq' in $input1.extension and 'fasta' in $input2.extension))
    echo 'Both input files must be either fastq or fasta (no mixing the two).' >&2
  #else
    python '$__tool_directory__/trimmer.py' '$input1'
    #if $paired.is_paired:
      '$input2' '$output1' '$output2'
    #end if
    #if $input1.extension in ('fastq', 'fastqsanger', 'fastqillumina', 'fastqsolexa')
      -f fastq
    #elif $input1.extension == 'fasta'
      -f fasta
    #else
      -f '$input1.extension'
    #end if
    -b '$bases' -t '$thres' -w '$win_len' $invert
    #if $min_len.has_min_len:
      -m '$min_len.value'
    #end if
    #if not $paired.is_paired:
      > '$output1'
    #end if
  #end if
  ]]>
  </command>
  <inputs>
    <conditional name="paired">
      <param name="is_paired" type="select" label="Paired reads?">
        <option value="" selected="True">Unpaired</option>
        <option value="true">Paired</option>
      </param>
      <when value="true">
        <param name="input1" type="data" format="fasta,fastq" label="Input reads (mate 1)"/>
        <param name="input2" type="data" format="fasta,fastq" label="Input reads (mate 2)"/>
      </when>
      <when value="">
        <param name="input1" type="data" format="fasta,fastq" label="Input reads"/>
      </when>
    </conditional>
    <param name="bases" type="text" value="N" label="Bases to filter on"/>
    <param name="thres" type="float" value="0.5" min="0" max="1" label="Frequency threshold" help="Trim when the frequency of filter bases (or non-filter bases, if inverting) exceeds this value."/>
    <param name="win_len" type="integer" value="10" min="1" label="Size of the window"/>
    <param name="invert" type="boolean" truevalue="--invert" falsevalue="" checked="False" label="Invert filter bases" help="Trim when the frequency of bases NOT in the &quot;filter bases&quot; list exceeds the threshold."/>
    <conditional name="min_len">
      <param name="has_min_len" type="boolean" truevalue="true" falsevalue="" checked="False" label="Set a minimum read length"/>
      <when value="true"> 
        <param name="value" type="integer" value="10" min="0" label="Minimum read length" help="Reads trimmed to less than this length will be omitted from the output. Pairs will be preserved: both must exceed this threshold to be kept."/>
      </when>
    </conditional>
  </inputs>
  <outputs>
    <data name="output1" format_source="input1" label="$tool.name on $on_string"/>
    <data name="output2" format_source="input2" label="$tool.name on $on_string (mate 2)">
      <filter>paired['is_paired']</filter>
    </data>
  </outputs>

  <help>

.. class:: infomark

**What it does**

This tool trims the 3' ends of reads based on the presence of the given bases. For instance, trim when N's are encountered or when the GC content exceeds a certain frequency.


.. class:: infomark

**How it works**

This will slide along the read with a window, and trim once the frequency of filter bases exceeds the frequency threshold (unless "Invert filter bases" is enabled, in which case it will trim once non-filter bases exceed the threshold).

The trim point will be just before the first (leftmost) filter base in the final window (the one where the frequency exceeded the threshold).


.. class:: infomark

**Input**

The inputs can be in the following formats: fasta, fastq, fastqsanger, fastqillumina, and fastqsolexa. Both must be either a fasta or fastq type (no mixing fastq and fasta).

  </help>

</tool>