root/galaxy-central/tools/unix_tools/sort_tool.xml @ 3

<tool id="cshl_sort_tool" name="Sort">
  <!-- 
        note 1:
          the 'version' sort (or natual order sort)
          requires GNU Coreutils 7.1 or later

        note 2:
          for greater efficiency, sort buffer size is very large.
          If your Galaxy server doesn't have so much memory (or the
          sorts you use don't require it) - you can decrease the memory size.
          (argument is "-S 2G")
  -->
  <command>sort -S 2G $unique 
      #for $key in $sortkeys
       '-k ${key.column},${key.column}${key.order}${key.style}'
      #end for
        $input > $out_file1
  </command>

  <inputs>
        <param format="txt" name="input" type="data" label="Sort Query" />
                
        <param name="unique" type="select" label="Output only unique values?">
                <option value="">No</option>
                <option value="-u">Yes</option>
        </param>

        <repeat name="sortkeys" title="sort key">
            <param name="column" label="on column" type="data_column" data_ref="input" accept_default="true" />
            <param name="order" type="select" display="radio" label="in">
              <option value="r">Descending order</option>
              <option value="">Ascending order</option>
            </param>
            <param name="style" type="select" display="radio" label="Flavor">
              <option value="n">Fast numeric sort ([-n])</option>
              <option value="g">General numeric sort ( scientific notation [-g])</option>
              <option value="V">Natural/Version sort ([-V]) </option>
              <option value="">Alphabetical sort</option>
            </param>
        </repeat>
  </inputs>
  <tests>
          <test>
                  <!-- Sort Descending numerical order,
                       with scientific notation -->
                  <param name="input" value="unix_sort_input1.txt" />
                  <output name="output" file="unix_sort_output1.txt" />
                  <param name="unique" value="No" />
                  <param name="column" value="2" />
                  <param name="order"  value="r" />
                  <param name="style"  value="g" />
          </test>
          <test>
                  <!-- Sort Ascending numerical order,
                  with scientific notation - outputing unique values only 

                  The catch:
                        chr15 appears twice, with the same value (0.0314 and 3.14e-2).
                        In the output, it should appear only once because of the unique flag
                  -->
                  <param name="input" value="unix_sort_input1.txt" />
                  <output name="output" file="unix_sort_output2.txt" />
                  <param name="unique" value="Yes" />
                  <param name="column" value="2" />
                  <param name="order"  value="" />
                  <param name="style"  value="g" />
          </test>
          <test>
                  <!-- Sort Ascending 'natural' order -->
                  <param name="input" value="unix_sort_input1.txt" />
                  <output name="output" file="unix_sort_output3.txt" />
                  <param name="unique" value="No" />
                  <param name="column" value="1" />
                  <param name="order"  value="" />
                  <param name="style"  value="V" />
          </test>
  </tests>
  <outputs>
    <data format="input" name="out_file1" metadata_source="input"/>
  </outputs>
  <help>

**What it does**

This tool runs the unix **sort** command on the selected data file.

-----

**Sorting Styles**

* **Fast Numeric**: sort by numeric values. Handles integer values (e.g. 43, 134) and decimal-point values (e.g. 3.14). *Does not* handle scientific notation (e.g. -2.32e2).
* **General Numeric**: sort by numeric values. Handles all numeric notations (including scientific notation). Slower than *fast numeric*, so use only when necessary.
* **Natural Sort**: Sort in 'natural' order (natural to humans, not to computers). See example below.
* **Alphabetical sort**: Sort in strict alphabetical order. See example below.




**Sorting Examples**

Given the following list::

    chr4
    chr13
    chr1
    chr10
    chr20
    chr2

**Alphabetical sort** would produce the following sorted list::

    chr1
    chr10
    chr13
    chr2
    chr20
    chr4

**Natural Sort** would produce the following sorted list::

    chr1
    chr2
    chr4
    chr10
    chr13
    chr20


.. class:: infomark

If you're planning to use the file with another tool that expected sorted files (such as *join*), you should use the **Alphabetical sort**,  not the **Natural Sort**. Natural sort order is easier for humans, but is unnatural for computer programs.

  </help>
</tool>