<tool id="cshl_sed_tool" name="sed">
  <description></description>
  <!-- NOTE
  	  'sandbox' is a patched SED program,
	  which blocks executing shell commands and file reading/writing.

	  Hopefully, it is safe enough to allow users to execute their own SED commands
	  -->
  <command interpreter="sh">sed_wrapper.sh $silent $input $output '$url_paste'</command>
  <inputs>
    <param format="txt" name="input" type="data" label="File to process" />

    <!-- Note: the parameter ane MUST BE 'url_paste' -
         This is a hack in the galaxy library (see ./lib/galaxy/util/__init__.py line 142)
	 If the name is 'url_paste' the string won't be sanitized, and all the non-alphanumeric characters 
	 will be passed to the shell script -->
    <param name="url_paste" type="text" area="true" size="5x35" label="SED Program" help=""> 
    	<validator type="expression" message="Invalid Program!">value.find('\'')==-1</validator>
    </param>

    <param name="silent" type="select"  label="operation mode" help="(Same as 'sed -n', leave at 'normal' unless you know what you're doing)" > 
      <option value="">normal</option>
      <option value="-n">silent</option>
    </param>

  </inputs>
  <outputs>
    <data format="input" name="output" metadata_source="input" />
  </outputs>
<help>

**What it does**

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

.. class:: infomark

**TIP:** This tool uses the **extended regular** expression syntax (same as running 'sed -r').



**Further reading**

- Short sed tutorial (http://www.linuxhowtos.org/System/sed_tutorial.htm)
- Long sed tutorial (http://www.grymoire.com/Unix/Sed.html)
- sed faq with good examples (http://sed.sourceforge.net/sedfaq.html)
- sed cheat-sheet (http://www.catonmat.net/download/sed.stream.editor.cheat.sheet.pdf)
- Collection of useful sed one-liners (http://student.northpark.edu/pemente/sed/sed1line.txt)

-----

**Sed commands**

The most useful sed command is **s** (substitute).

**Examples**

- **s/hsa//**  will remove the first instance of 'hsa' in every line.
- **s/hsa//g**  will remove all instances (beacuse of the **g**) of 'hsa' in every line.
- **s/A{4,}/--&amp;--/g**  will find sequences of 4 or more consecutive A's, and once found, will surround them with two dashes from each side. The **&amp;** marker is a place holder for 'whatever matched the regular expression'.
- **s/hsa-mir-([^ ]+)/short name: \\1 full name: &amp;/**  will find strings such as 'hsa-mir-43a' (the regular expression is 'hsa-mir-' followed by non-space characters) and will replace it will string such as 'short name: 43a full name: hsa-mir-43a'.  The **\\1** marker is a place holder for 'whatever matched the first parenthesis' (similar to perl's **$1**) .


**sed's Regular Expression Syntax**

The select tool searches the data for lines containing or not containing a match to the given pattern. A Regular Expression is a pattern descibing a certain amount of text. 

- **( ) { } [ ] . * ? + \ ^ $** are all special characters. **\\** can be used to "escape" a special character, allowing that special character to be searched for.
- **^** matches the beginning of a string(but not an internal line).
- **(** .. **)** groups a particular pattern.
- **{** n or n, or n,m **}** specifies an expected number of repetitions of the preceding pattern.

  - **{n}** The preceding item is matched exactly n times.
  - **{n,}** The preceding item ismatched n or more times. 
  - **{n,m}** The preceding item is matched at least n times but not more than m times. 

- **[** ... **]** creates a character class. Within the brackets, single characters can be placed. A dash (-) may be used to indicate a range such as **a-z**.
- **.** Matches any single character except a newline.
- ***** The preceding item will be matched zero or more times.
- **?** The preceding item is optional and matched at most once.
- **+** The preceding item will be matched one or more times.
- **^** has two meaning:
  - matches the beginning of a line or string. 
  - indicates negation in a character class. For example, [^...] matches every character except the ones inside brackets.
- **$** matches the end of a line or string.
- **\|** Separates alternate possibilities. 


**Note**: SED uses extended regular expression syntax, not Perl syntax. **\\d**, **\\w**, **\\s** etc. are **not** supported.

</help>
</tool>