<tool id="sqlite" name="SQL" version="0.4">
  <description>Runner (SQLite)</description>
  <command interpreter="python">
    sqlite.py $output_file $explain_plan $header "$sql"
    #for $i in $files
      ${i.file} ${i.header} '${i.index}'
    #end for
  </command>
  <inputs>
    <repeat name="files" title="Table">
      <param name="file" type="data" format="tabular" label="File (tabular)"/>
      <param name="header" type="boolean" truevalue="1" falsevalue="0" checked="False" label="Use the first row as column names (If unchecked, use c1 c2 c3 ...)" />
      <param name="index" type="text" size='20' value="" label="Index" />
    </repeat>
    <param name="header" type="boolean" truevalue="1" falsevalue="0" checked="False" label="Print column names in the first row" />
    <param name="sql" type="text" area="true" size="10x50" label="SQL">
      <sanitizer sanitize="False"/>
    </param>
    <param name="explain_plan" type="boolean" truevalue="1" falsevalue="0" checked="False" label="Explain SQL execution plan. (Ignore if you are not a developer)" />
  </inputs>
  <outputs>
    <data name="output_file" format="tabular" label="${tool.name} on ${on_string}"/>
  </outputs>
  <help>

**Tables**

Select all table to be joined.

----

**Use the first line as column names**

When the input file has column header::

    id    name
    101   John
    102   Mary

Check this box and use column names ('id' and 'name') to write queries:

    SELECT * FROM t1 WHERE c2 = 'John';

When the input file does not have column header::

    101   John
    102   Mary

Uncheck this box and use c1, c2, c3, ... to write queries:

    SELECT * FROM t1 WHERE name = 'John';

----

**Index**

Comma-delimited list of column numbers.

Example:

    2,3

This creates NOT a composite index BUT two different indices on column 2 and on column 3.

----

**Print column names in the first row**

Example::

    col1 col2 col3
    aaaa bbbb cccc
    aaaa dddd eeee

----

**SQL**

  </help>
</tool>