<tool id="sed_stream_editor" name="Manipulation" version="0.0.1">
  <description>of text lines with regular expressions (sed)</description>
  <command>sed -r '$pattern' $input > $outfile</command>
  <inputs>
    <param format="txt, tabular" name="input" type="data" label="Replace lines from"/>
    <param name="pattern" size="40" type="text" value="" label="the pattern" help="here you can enter your sed expression (No syntax check or sanitising!)" />
  </inputs>
  <outputs>
    <data format="input" name="outfile" metadata_source="input"/>
  </outputs>
  <options sanitize="False"/>
 <requirements>
    <requirement type="binary">sed</requirement>
 </requirements>
  <tests>

  </tests>
  <help>

.. class:: warningmark

Use with caution! Its a plain wrapper around **sed** and the input is not sanitized.


-----

**What it does**

Changes every line of a text file according to a given regular expression.

-----

**Syntax**

Use the **sed**-syntax -&gt; **s/find-pattern/replace-pattern/**

**Example**

- **s/x/-/** Replace all **x** with **-**.

- **s/_.*//** Splits a string after **_** and replaces the rest with nothing.
- **s/[^_]*_*//** Splits a string after **_** and replaces the first part with nothing.

- **s/\\s.*//** Splits a string after whitespaces and replaces the rest with nothing.
- **s/\\S*\\s*//** Splits a string after whitespaces and replaces the first part with nothing.


</help>
</tool>