<?xml version='1.0'?> <!-- -*- xml -*- -->
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN"
 "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd">
<book>
  <bookinfo>
    <title>Patchutils</title>
  </bookinfo>

  <reference>
    <referenceinfo>
      <author>
	<firstname>Tim</firstname>
	<surname>Waugh</surname>
	<affiliation>
	  <address><email>twaugh@redhat.com</email></address>
	</affiliation>
	<contrib>Package maintainer</contrib>
      </author>
    </referenceinfo>

    <title>Man pages</title>
    <refentry id="interdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>23 June 2009</date>
	<editor>
	  <firstname>Michael K.</firstname>
	  <surname>Johnson</surname>
	  <affiliation>
	    <address><email>johnsonm@redhat.com</email></address>
	  </affiliation>
	  <contrib>Original man page contributor</contrib>
	</editor>
      </refentryinfo>

      <refmeta>
	<refentrytitle>interdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>interdiff</refname>
	<refpurpose>show differences between two diff files</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>interdiff</command>
	  <group choice="opt">
	    <arg>-p <replaceable>n</replaceable></arg>
	    <arg>--strip-match=<replaceable>n</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-U <replaceable>n</replaceable></arg>
	    <arg>--unified=<replaceable>n</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-d <replaceable>PAT</replaceable></arg>
	    <arg>--drop-context=<replaceable>PAT</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-q</arg>
	    <arg>--quiet</arg>
	  </group>
	  <group choice="opt">
	    <arg>-z</arg>
	    <arg>--decompress</arg>
	  </group>
	  <group choice="opt">
	    <arg>-b</arg>
	    <arg>--ignore-space-change</arg>
	  </group>
	  <group choice="opt">
	    <arg>-B</arg>
	    <arg>--ignore-blank-lines</arg>
	  </group>
	  <group choice="opt">
	    <arg>-i</arg>
	    <arg>--ignore-case</arg>
	  </group>
	  <group choice="opt">
	    <arg>-w</arg>
	    <arg>--ignore-all-space</arg>
	  </group>
	  <group choice="opt">
	    <arg>--interpolate</arg>
	    <arg>--combine</arg>
	    <arg>--flip</arg>
	  </group>
	  <arg choice="opt">--no-revert-omitted</arg>
	  <arg choice="plain"><replaceable>diff1</replaceable></arg>
	  <arg choice="plain"><replaceable>diff2</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>interdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para><command>interdiff</command> creates a unified format
          diff that expresses the difference between two diffs.  The
          diffs must both be relative to the same files.  For best
          results, the diffs must have at least three lines of
          context.</para>

	<para>To reverse a patch, use <filename>/dev/null</filename>
	  for <replaceable>diff2</replaceable>.</para>

	<para>To reduce the amount of context in a patch, use:

<screen><![CDATA[interdiff -U1 /dev/null patchfile]]></screen></para>

	<para>Since <command>interdiff</command> doesn't have the
	  advantage of being able to look at the files that are to be
	  modified, it has stricter requirements on the input format
	  than <citerefentry>
	    <refentrytitle>patch</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry> does.  The output of GNU diff will be okay,
	  even with extensions, but if you intend to use a hand-edited
	  patch it might be wise to clean up the offsets and counts
	  using <citerefentry>
	    <refentrytitle>recountdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry> first.</para>

	<para>Note, however, that the two patches must both be
	  relative to the versions of the same original set of
	  files.</para>

	<para>The diffs may be in context format.  The output,
	  however, will be in unified format.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-h</option></term>
	    <listitem>
	      <para>Ignored, for compatibility with older versions of
                interdiff.  This option will go away soon.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable>,
	    <option>--strip-match=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>When comparing filenames, ignore the first
                <replaceable>n</replaceable> pathname components from
                both patches. (This is similar to the
                <option>-p</option> option to GNU <citerefentry>
		  <refentrytitle>patch</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.)</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-q</option>, <option>--quiet</option></term>
	    <listitem>
	      <para>Quieter output.  Don't emit rationale lines at the
	       beginning of each patch.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-U</option> <replaceable>n</replaceable>,
	    <option>--unified=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>Attempt to display <replaceable>n</replaceable>
                lines of context (requires at least
                <replaceable>n</replaceable> lines of context in both
                input files). (This is similar to the
                <option>-U</option> option to GNU
                <citerefentry>
		  <refentrytitle>diff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.)</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-d</option> <replaceable>PATTERN</replaceable>,
	    <option>--drop-context=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Don't display any context on files that match the
                shell wildcard <replaceable>PATTERN</replaceable>.
                This option can be given multiple times.</para>

	      <para>Note that the interpretation of the shell wildcard
	        pattern does not count slash characters or periods as
	        special (in other words, no flags are given to
	        <function>fnmatch</function>).  This is so that
	        <quote>*/basename</quote>-type patterns can be given
	        without limiting the number of pathname
	        components.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-i</option>, <option>--ignore-case</option></term>
	    <listitem>
	      <para>Consider upper- and lower-case to be the same.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-w</option>,
	    <option>--ignore-all-space</option></term>
	    <listitem>
	      <para>Ignore whitespace changes in patches.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-b</option>,
	    <option>--ignore-space-change</option></term>
	    <listitem>
	      <para>Ignore changes in the amount of whitespace.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-B</option>,
	    <option>--ignore-blank-lines</option></term>
	    <listitem>
	      <para>Ignore changes whose lines are all blank.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-z</option>,
	    <option>--decompress</option></term>
	    <listitem>
	      <para>Decompress files with extensions .gz and .bz2.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--interpolate</option></term>
	    <listitem>
	      <para>Run as <quote>interdiff</quote>.  This is the
 	        default.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--combine</option></term>
	    <listitem>
	      <para>Run as <quote>combinediff</quote>.  See
	      <citerefentry>
		  <refentrytitle>combinediff</refentrytitle>
		  <manvolnum>1</manvolnum> </citerefentry> for more
		  information about how the behaviour is altered in
		  this mode.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--no-revert-omitted</option></term>
	    <listitem>
	      <para>(For interpolation mode only)  When a file is
	        changed by the first patch but not by the second,
	        don't revert that change.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of interdiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Examples</title>

	<para>Basic usage:

<screen><![CDATA[interdiff -z 3.2pre1.patch.gz 3.2pre2.patch.gz]]></screen></para>
	<para>Reversing a patch:

<screen><![CDATA[interdiff patch /dev/null]]></screen></para>

	<para>Reversing part of a patch (and ignoring the
	  rest):

<screen><![CDATA[filterdiff -i file.c patchfile | \
  interdiff /dev/stdin /dev/null]]></screen></para>

      </refsect1>

      <refsect1>
	<title>Bugs</title>

	<para>There are currently no known bugs in interdiff; but
          there are some caveats.  If you find a bug, please report it
          (along with a minimal test case) to Tim Waugh
          <email>twaugh@redhat.com</email>.</para>

	<para>There are some sets of patches in which there is just
	  not enough information to produce a proper interdiff.  In
	  this case, the strategy employed is to revert the original
	  patch and apply the new patch.  This, unfortunately, means
	  that interdiffs are not guaranteed to be reversible.</para>

      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>combinediff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="combinediff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>23 Jan 2009</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>combinediff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>combinediff</refname>
	<refpurpose>create a cumulative unified patch from two incremental
	  patches</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>combinediff</command>
	  <group choice="opt">
	    <arg>-p <replaceable>n</replaceable></arg>
	    <arg>--strip-match=<replaceable>n</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-U <replaceable>n</replaceable></arg>
	    <arg>--unified=<replaceable>n</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-d <replaceable>PAT</replaceable></arg>
	    <arg>--drop-context=<replaceable>PAT</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-q</arg>
	    <arg>--quiet</arg>
	  </group>
	  <group choice="opt">
	    <arg>-z</arg>
	    <arg>--decompress</arg>
	  </group>
	  <group choice="opt">
	    <arg>-b</arg>
	    <arg>--ignore-space-change</arg>
	  </group>
	  <group choice="opt">
	    <arg>-B</arg>
	    <arg>--ignore-blank-lines</arg>
	  </group>
	  <group choice="opt">
	    <arg>-i</arg>
	    <arg>--ignore-case</arg>
	  </group>
	  <group choice="opt">
	    <arg>-w</arg>
	    <arg>--ignore-all-space</arg>
	  </group>
	  <group choice="opt">
	    <arg>--interpolate</arg>
	    <arg>--combine</arg>
	  </group>
	  <arg choice="plain"><replaceable>diff1</replaceable></arg>
	  <arg choice="plain"><replaceable>diff2</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>combinediff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>combinediff creates a unified diff that expresses the
          sum of two diffs.  The diff files must be listed in the
          order that they are to be applied.  For best results, the
          diffs must have at least three lines of context.</para>

	<para>Since <command>combinediff</command> doesn't have the
	  advantage of being able to look at the files that are to be
	  modified, it has stricter requirements on the input format
	  than <citerefentry>
	    <refentrytitle>patch</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry> does.  The output of GNU diff will be okay,
	  even with extensions, but if you intend to use a hand-edited
	  patch it might be wise to clean up the offsets and counts
	  using <citerefentry>
	    <refentrytitle>recountdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry> first.</para>

	<para>Note, however, that the two patches must be in strict
	incremental order.  In other words, the second patch must be
	relative to the state of the original set of files after the
	first patch was applied.</para>

	<para>The diffs may be in context format.  The output,
	  however, will be in unified format.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable>,
	    <option>--strip-match=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>When comparing filenames, ignore the first
                <replaceable>n</replaceable> pathname components from
                both patches. (This is similar to the
                <option>-p</option> option to GNU <citerefentry>
		  <refentrytitle>patch</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.)</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-q</option>, <option>--quiet</option></term>
	    <listitem>
	      <para>Quieter output.  Don't emit rationale lines at the
	       beginning of each patch.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-U</option> <replaceable>n</replaceable>,
	    <option>--unified=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>Attempt to display <replaceable>n</replaceable>
                lines of context (requires at least
                <replaceable>n</replaceable> lines of context in both
                input files). (This is similar to the
                <option>-U</option> option to GNU
                <citerefentry>
		  <refentrytitle>diff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.)</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-d</option> <replaceable>pattern</replaceable>,
	    <option>--drop-context=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Don't display any context on files that match the
                shell wildcard <replaceable>pattern</replaceable>.
                This option can be given multiple times.</para>

	      <para>Note that the interpretation of the shell wildcard
	        pattern does not count slash characters or periods as
	        special (in other words, no flags are given to
	        <function>fnmatch</function>).  This is so that
	        <quote>*/basename</quote>-type patterns can be given
	        without limiting the number of pathname
	        components.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-i</option>, <option>--ignore-case</option></term>
	    <listitem>
	      <para>Consider upper- and lower-case to be the same.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-w</option>,
	    <option>--ignore-all-space</option></term>
	    <listitem>
	      <para>Ignore whitespace changes in patches.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-b</option>,
	    <option>--ignore-space-change</option></term>
	    <listitem>
	      <para>Ignore changes in the amount of whitespace.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-B</option>,
	    <option>--ignore-blank-lines</option></term>
	    <listitem>
	      <para>Ignore changes whose lines are all blank.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-z</option>,
	    <option>--decompress</option></term>
	    <listitem>
	      <para>Decompress files with extensions .gz and .bz2.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--interpolate</option></term>
	    <listitem>
	      <para>Run as <quote>interdiff</quote>.  See
	      <citerefentry>
		  <refentrytitle>interdiff</refentrytitle>
		  <manvolnum>1</manvolnum> </citerefentry> for more
		  information about how the behaviour is altered in
		  this mode.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--combine</option></term>
	    <listitem>
	      <para>Run as <quote>combinediff</quote>.  This is the
 	        default.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of combinediff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Bugs</title>

	<para>The <option>-U</option> option is a bit erratic: it can
          control the amount of context displayed for files that are
          modified in both patches, but not for files that only appear
          in one patch (which appear with the same amount of context
          in the output as in the input).</para>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>interdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="filterdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>23 Jan 2009</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>filterdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>filterdiff</refname>
	<refpurpose>extract or exclude diffs from a diff file</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>filterdiff</command>
	  <group choice="opt">
	    <arg>-i <replaceable>PATTERN</replaceable></arg>
	    <arg>--include=<replaceable>PATTERN</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-I <replaceable>FILE</replaceable></arg>
	    <arg>--include-from-file=<replaceable>FILE</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-p <replaceable>n</replaceable></arg>
	    <arg>--strip-match=<replaceable>n</replaceable></arg>
	  </group>
	  <arg choice="opt">--strip=<replaceable>n</replaceable></arg>
	  <arg choice="opt">--addprefix=<replaceable>PREFIX</replaceable></arg>
	  <arg choice="opt">--addoldprefix=<replaceable>PREFIX</replaceable></arg>
	  <arg choice="opt">--addnewprefix=<replaceable>PREFIX</replaceable></arg>
	  <group choice="opt">
	    <arg>-x <replaceable>PATTERN</replaceable></arg>
	    <arg>--exclude=<replaceable>PATTERN</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-X <replaceable>FILE</replaceable></arg>
	    <arg>--exclude-from-file=<replaceable>FILE</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-v</arg>
	    <arg>--verbose</arg>
	  </group>
	  <arg choice="opt">--clean</arg>
	  <group choice="opt">
	    <arg>-z</arg>
	    <arg>--decompress</arg>
	  </group>
	  <group choice="opt">
	    <arg>-# <replaceable>RANGE</replaceable></arg>
	    <arg>--hunks=<replaceable>RANGE</replaceable></arg>
	  </group>
	  <arg choice="opt">--lines=<replaceable>RANGE</replaceable></arg>
	  <arg choice="opt">--files=<replaceable>RANGE</replaceable></arg>
	  <arg choice="opt">--annotate</arg>
	  <arg choice="opt">--format=<replaceable>FORMAT</replaceable></arg>
	  <arg choice="opt">--as-numbered-lines=<replaceable>WHEN</replaceable></arg>
	  <arg choice="opt">--remove-timestamps</arg>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>filterdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	    <arg>--list</arg>
	    <arg>--grep ...</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>You can use filterdiff to obtain a patch that applies to
          files matching the shell wildcard
          <replaceable>PATTERN</replaceable> from a larger collection
          of patches.  For example, to see the patches in
          <filename>patch-2.4.3.gz</filename> that apply to all files
          called <filename>lp.c</filename>:

	<screen>filterdiff -z -i '*/lp.c' patch-2.4.3.gz</screen></para>

	<para>If neither <option>-i</option> nor <option>-x</option>
	options are given, <option>-i '*'</option> is assumed.  This
	way <command>filterdiff</command> can be used to clean up an
	existing diff file, removing redundant lines from the
	beginning (eg. the text from the mail body) or between the
	chunks (eg. in CVS diffs). To extract pure patch data, use a
	command like this:

	<screen>filterdiff message-with-diff-in-the-body > patch</screen></para>

	<para>Note that the interpretation of the shell wildcard
	  pattern does not count slash characters or periods as
	  special (in other words, no flags are given to
	  <function>fnmatch</function>).  This is so that
	  <quote>*/basename</quote>-type patterns can be given without
	  limiting the number of pathname components.</para>

	<para>You can use both unified and context format diffs with
	  this program.</para>

      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-i</option> <replaceable>PATTERN</replaceable>,
	    <option>--include=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Include only files matching
	        <replaceable>PATTERN</replaceable>.  All other lines
	        in the input are suppressed.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-I</option> <replaceable>FILE</replaceable>,
	    <option>--include-from-file=<replaceable>FILE</replaceable></option></term>
	    <listitem>
	      <para>Include only files matching any pattern listed in
	        <replaceable>FILE</replaceable>, one pattern per line.
		All other lines in the input are suppressed.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-x</option> <replaceable>PATTERN</replaceable>,
	    <option>--exclude=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Exclude files matching
	        <replaceable>PATTERN</replaceable>.  All other lines
	        in the input are displayed.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-X</option> <replaceable>FILE</replaceable>,
	    <option>--exclude-from-file=<replaceable>FILE</replaceable></option></term>
	    <listitem>
	      <para>Exclude files matching any pattern listed in
	        <replaceable>FILE</replaceable>, one pattern per line.
	        All other lines in the input are displayed.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable>,
	    <option>--strip-match=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>When matching, ignore the first
	        <replaceable>n</replaceable> components of the
	        pathname.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-#</option> <replaceable>RANGE</replaceable>,
	    <option>--hunks</option>=<replaceable>RANGE</replaceable></term>
	    <listitem>
	      <para>Only include hunks within the specified
	        <replaceable>RANGE</replaceable>.  Hunks are numbered
	        from 1, and the range is a comma-separated list of
	        numbers or <quote>first-last</quote> spans; either the
	        first or the last in the span may be omitted to
	        indicate no limit in that direction.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--lines</option>=<replaceable>RANGE</replaceable></term>
	    <listitem>
	      <para>Only include hunks that contain lines from the
	        original file that lie within the specified
	        <replaceable>RANGE</replaceable>.  Lines are numbered
	        from 1, and the range is a comma-separated list of
	        numbers or <quote>first-last</quote> spans; either the
	        first or the last in the span may be omitted to
	        indicate no limit in that direction.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--files</option>=<replaceable>RANGE</replaceable></term>
	    <listitem>
	      <para>Only include files indicated by the specified
	        <replaceable>RANGE</replaceable>.  Files are numbered
	        from 1 in the order they appear in the patch input,
	        and the range is a comma-separated list of numbers or
	        <quote>first-last</quote> spans; either the first or
	        the last in the span may be omitted to indicate no
	        limit in that direction.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--annotate</option></term>

	    <listitem>
	      <para>Annotate each hunk with the filename and hunk
	        number.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--format</option>=unified|context</term>

	    <listitem>
	      <para>Use specified output format.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--strip</option>=<replaceable>n</replaceable></term>
	    <listitem>
	      <para>Remove the first <replaceable>n</replaceable>
	        components of pathnames in the output.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--addprefix</option>=<replaceable>PREFIX</replaceable></term>
	    <listitem>
	      <para>Prefix pathnames in the output by
	        <replaceable>PREFIX</replaceable>.  This will override
	        any individual settings specified with
	        the <option>--addoldprefix</option>
	        or <option>--addnewprefix</option> options.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--addoldprefix</option>=<replaceable>PREFIX</replaceable></term>
	    <listitem>
	      <para>Prefix pathnames for old or original files in the
	      output by <replaceable>PREFIX</replaceable>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--addnewprefix</option>=<replaceable>PREFIX</replaceable></term>
	    <listitem>
	      <para>Prefix pathnames for updated or new files in the
	      output by <replaceable>PREFIX</replaceable>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--as-numbered-lines</option>=before|after</term>
	    <listitem>
	      <para>Instead of a patch fragment, display the lines of
	        the selected hunks with the line number of the file
	        before (or after) the patch is applied, followed by a
	        <keycode>TAB</keycode> character and a colon, at the
	        beginning of each line.  Each hunk except the first
	        will have a line consisting of <quote>...</quote>
	        before it.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--remove-timestamps</option></term>

	    <listitem>
	      <para>Do not include file timestamps in the
	        output.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-v</option>, <option>--verbose</option></term>
	    <listitem>
	      <para>Always show non-diff lines in the output.  By
	        default, non-diff lines are only shown when excluding
	        a filename pattern.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--clean</option></term>
	    <listitem>
	      <para>Always remove all non-diff lines from the
	        output. Even when excluding a filename pattern.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-z</option>, <option>--decompress</option></term>
	    <listitem>
	      <para>Decompress files with extensions .gz and .bz2.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of filterdiff.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--list</option></term>
	    <listitem>
	      <para>Behave like <citerefentry>
		  <refentrytitle>lsdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry> instead.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--grep</option></term>
	    <listitem>
	      <para>Behave like <citerefentry>
		  <refentrytitle>grepdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry> instead.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Examples</title>

	<para>To see all patch hunks that affect the first five lines
	  of a C file:

<screen><![CDATA[filterdiff -i '*.c' --lines=-5 < patch]]></screen></para>

	<para>To see the first hunk of each file patch, use:

<screen><![CDATA[filterdiff -#1 patchfile]]></screen></para>

	<para>To see patches modifying a ChangeLog file in a
	  subdirectory, use:

<screen><![CDATA[filterdiff -p1 Changelog]]></screen></para>

	<para>To see the complete patches for each patch that modifies
	  line 1 of the original file, use:

<screen><![CDATA[filterdiff --lines=1 patchfile | lsdiff | \
  xargs -rn1 filterdiff patchfile -i]]></screen></para>

	<para>To see all but the first hunk of a particular patch, you
	  might use:

<screen><![CDATA[filterdiff -p1 -i file.c -#2- foo-patch]]></screen></para>

	<para>If you have a very specific list of hunks in a patch
	  that you want to see, list them:

<screen><![CDATA[filterdiff -#1,2,5-8,10,12,27-]]></screen></para>

	<para>To see the lines of the files that would be patched as
	  they will appear after the patch is applied, use:

<screen><![CDATA[filterdiff --as-numbered-lines=after patch.file]]></screen></para>

	<para>You can see the same context before the patch is applied
	  with:

<screen><![CDATA[filterdiff --as-numbered-lines=before
patch.file]]></screen></para>

	<para>Filterdiff can also be used to convert between unified
	  and context format diffs:

<screen><![CDATA[filterdiff -v --format=unified context.diff]]></screen></para>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>lsdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>grepdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="fixcvsdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>7 Dec 2001</date>
	<authorgroup>
	  <author>
	    <firstname>Amir</firstname>
	    <surname>Karger</surname>
	    <affiliation>
	      <address><email>karger@post.harvard.edu</email></address>
	    </affiliation>
	    <contrib>Author of the original perl script and man page</contrib>
	  </author>

	  <author>
	    <firstname>John</firstname>
	    <surname>Levon</surname>
	    <affiliation>
	      <address><email>moz@compsoc.man.ac.uk</email></address>
	    </affiliation>
	    <contrib>Author of the original perl script and man page</contrib>
	  </author>
	</authorgroup>
      </refentryinfo>

      <refmeta>
	<refentrytitle>fixcvsdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>fixcvsdiff</refname>
	<refpurpose>fix problematic diff files</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>fixcvsdiff</command>
	  <arg choice="opt">-p</arg>
	  <arg choice="opt">-b</arg>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>fixcvsdiff modifies diff files created from the
          <command>cvs diff</command> command, where files have been
          added or removed. CVS tends to create diff files that
          <citerefentry>
	    <refentrytitle>patch</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry> mis-interprets. The diff file must retain the
          lines beginning with <quote>Index:</quote> in order for the
          correction to work.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-b</option></term>
	    <listitem>
	      <para>Keep a .bak backup file.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-p</option></term>
	    <listitem>
	      <para>Print out the <command>cvs add</command> and
                <command>cvs remove</command> commands that must be
                made after applying the diff, but before running
                <command>cvs commit</command>.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>diff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>patch</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>cvs</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="lsdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>23 Jan 2009</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>lsdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>lsdiff</refname>
	<refpurpose>show which files are modified by a patch</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>lsdiff</command>
	  <group choice="opt">
	    <arg>-n</arg>
	    <arg>--line-number</arg>
	  </group>
	  <group choice="opt">
	    <arg>-p <replaceable>n</replaceable></arg>
	    <arg>--strip-match=<replaceable>n</replaceable></arg>
	  </group>
	  <arg choice="opt">--strip=<replaceable>n</replaceable></arg>
	  <arg choice="opt">--addprefix=<replaceable>PREFIX</replaceable></arg>
	  <group choice="opt">
	    <arg>-s</arg>
	    <arg>--status</arg>
	  </group>
	  <group choice="opt">
	    <arg>-E</arg>
	    <arg>--empty-files-as-removed</arg>
	  </group>
	  <group choice="opt">
	    <arg>-i <replaceable>PATTERN</replaceable></arg>
	    <arg>--include=<replaceable>PATTERN</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-x <replaceable>PATTERN</replaceable></arg>
	    <arg>--exclude=<replaceable>PATTERN</replaceable></arg>
	  </group>
	  <group>
	    <arg>-z</arg>
	    <arg>--decompress</arg>
	  </group>
	  <group choice="opt">
	    <arg>-# <replaceable>RANGE</replaceable></arg>
	    <arg>--hunks=<replaceable>RANGE</replaceable></arg>
	  </group>
	  <arg choice="opt">--lines=<replaceable>RANGE</replaceable></arg>
	  <arg choice="opt">--files=<replaceable>RANGE</replaceable></arg>
	  <group choice="opt">
	    <arg>-H</arg>
	    <arg>--with-filename</arg>
	  </group>
	  <group choice="opt">
	    <arg>-h</arg>
	    <arg>--no-filename</arg>
	  </group>
	  <group choice="opt" rep="repeat">
	    <arg>-v</arg>
	    <arg>--verbose</arg>
	  </group>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>lsdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	    <arg>--filter ...</arg>
	    <arg>--grep ...</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>List the files modified by a patch.</para>

	<para>You can use both unified and context format diffs with
	  this program.</para>

      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-n</option>, <option>--line-number</option></term>
	    <listitem>
	      <para>Display the line number that each patch begins at.
	        If verbose output is requested (using
	        <option>-nv</option>), each hunk of each patch is
	        listed as well.</para>

	      <para>For each file that is modified, a line is
	        generated containing the line number of the beginning
	        of the patch, followed by a <keycode>Tab</keycode>
	        character, followed by the name of the file that is
	        modified.  If <option>-v</option> is given once,
	        following each of these lines will be one line for
	        each hunk, consisting of a <keycode>Tab</keycode>
	        character, the line number that the hunk begins at,
	        another <keycode>Tab</keycode> character, the string
	        <quote>Hunk #</quote>, and the hunk number (starting
	        at 1).</para>

	      <para>If the <option>-v</option> is given twice in
	        conjunction with <option>-n</option>
	        (i.e. <option>-nvv</option>), the format is slightly
	        different: hunk-level descriptive text
	        is shown after each hunk number, and the
	        <option>--number-files</option> option is
	        enabled.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--number-files</option></term>
	    <listitem>
	      <para>File numbers are listed, beginning at 1, before
	      each filename.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-#</option> <replaceable>RANGE</replaceable>,
	    <option>--hunks</option>=<replaceable>RANGE</replaceable></term>
	    <listitem>
	      <para>Only list hunks within the specified
	        <replaceable>RANGE</replaceable>.  Hunks are numbered
	        from 1, and the range is a comma-separated list of
	        numbers or <quote>first-last</quote> spans; either the
	        first or the last in the span may be omitted to
	        indicate no limit in that direction.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--lines</option>=<replaceable>RANGE</replaceable></term>
	    <listitem>
	      <para>Only list hunks that contain lines from the
	        original file that lie within the specified
	        <replaceable>RANGE</replaceable>.  Lines are numbered
	        from 1, and the range is a comma-separated list of
	        numbers or <quote>first-last</quote> spans; either the
	        first or the last in the span may be omitted to
	        indicate no limit in that direction.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--files</option>=<replaceable>RANGE</replaceable></term>
	    <listitem>
	      <para>Only list files indicated by the specified
	        <replaceable>RANGE</replaceable>.  Files are numbered
	        from 1 in the order they appear in the patch input,
	        and the range is a comma-separated list of numbers or
	        <quote>first-last</quote> spans; either the first or
	        the last in the span may be omitted to indicate no
	        limit in that direction.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable>,
	    <option>--strip-match=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>When matching, ignore the first
	        <replaceable>n</replaceable> components of the
	        pathname.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--strip</option>=<replaceable>n</replaceable></term>
	    <listitem>
	      <para>Remove the first <replaceable>n</replaceable>
	        components of the pathname before displaying
	        it.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--addprefix</option>=<replaceable>PREFIX</replaceable></term>
	    <listitem>
	      <para>Prefix the pathname with
	        <replaceable>PREFIX</replaceable> before displaying
	        it.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-s</option>, <option>--status</option></term>
	    <listitem>
	      <para>Show file additions, modifications and removals.
	       A file addition is indicated by a <quote>+</quote>, a
	       removal by a <quote>-</quote>, and a modification by a
	       <quote>!</quote>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-E</option>,
	    <option>--empty-files-as-removed</option></term>
	    <listitem>
	      <para>Treat empty files as absent for the purpose of
	      displaying file additions, modifications and
	      removals.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-i</option> <replaceable>PATTERN</replaceable>,
	    <option>--include=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Include only files matching
	        <replaceable>PATTERN</replaceable>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-x</option> <replaceable>PATTERN</replaceable>,
	    <option>--exclude=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Exclude files matching
	        <replaceable>PATTERN</replaceable>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-z</option>,
	    <option>--decompress</option></term>
	    <listitem>
	      <para>Decompress files with extensions .gz and .bz2.
              </para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-H</option>, <option>--with-filename</option></term>
	    <listitem>
	      <para>Print the name of the patch file containing each
	        patch.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-h</option>, <option>--no-filename</option></term>
	    <listitem>
	      <para>Suppress the name of the patch file containing each
	        patch.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-v</option>, <option>--verbose</option></term>
	    <listitem>
	      <para>Verbose output.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of lsdiff.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--filter</option></term>
	    <listitem>
	      <para>Behave like <citerefentry>
		  <refentrytitle>filterdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry> instead.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--grep</option></term>
	    <listitem>
	      <para>Behave like <citerefentry>
		  <refentrytitle>grepdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry> instead.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>filterdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>grepdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>

      <refsect1>
	<title>Examples</title>

	<para>To sort the order of touched files in a patch, you can
	  use:

<screen><![CDATA[lsdiff patch | sort -u | \
  xargs -rn1 filterdiff patch -i]]></screen></para>

	<para>To show only added files in a patch:

<screen><![CDATA[lsdiff -s patch | grep '^+' | \
  cut -c2- | xargs -rn1 filterdiff patch -i]]></screen></para>

	<para>To show the headers of all file hunks:

<screen><![CDATA[lsdiff -n patch | (while read n file
  do sed -ne "$n,$(($n+1))p" patch
done)]]></screen></para>
      </refsect1>

    </refentry>

    <refentry id="splitdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>21 Oct 2003</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>splitdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>splitdiff</refname>
	<refpurpose>separate out incremental patches</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>splitdiff</command>
	  <arg choice="opt">-a</arg>
	  <arg choice="opt">-d</arg>
	  <arg choice="opt">-p <replaceable>n</replaceable></arg>
	  <arg choice="opt"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>splitdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>If you have a patch file composed of several incremental
          patches, you can use splitdiff to separate them out.  You
          may want to do this in preparation for re-combining them
          with <citerefentry>
          <refentrytitle>combinediff</refentrytitle>
          <manvolnum>1</manvolnum></citerefentry>.</para>

	<para>The effect of running splitdiff is to separate its input
	  into a set of output files, with no output file patching the
	  same file more than once.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-a</option></term>
	    <listitem>
	      <para>Split out every single file-level patch.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-d</option></term>
	    <listitem>
	      <para>Create file names such as
	        <filename>a_b.c</filename> for a patch that modifies
	        <filename>a/b.c</filename>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable></term>
	    <listitem>
	      <para>Strip the first <replaceable>n</replaceable>
	        components of the pathname to aid comparisons.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of splitdiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>combinediff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>lsdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="rediff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>13 May 2002</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>rediff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>rediff</refname>
	<refname>editdiff</refname>
	<refpurpose>fix offsets and counts of a hand-edited diff</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>rediff</command>
	  <arg choice="plain"><replaceable>ORIGINAL</replaceable></arg>
	  <arg choice="plain"><replaceable>EDITED</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>rediff</command>
	  <arg choice="plain"><replaceable>EDITED</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>rediff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>editdiff</command>
	  <arg choice="plain"><replaceable>FILE</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>editdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>You can use rediff to correct a hand-edited unified
          diff.  Take a copy of the diff you want to edit, and edit it
          without changing any offsets or counts (the lines that begin
          <quote>@@</quote>).  Then run rediff, telling it the name of
          the original diff file and the name of the one you have
          edited, and it will output the edited diff file but with
          corrected offsets and counts.</para>

	<para>A small script, editdiff, is provided for editing a diff
          file in-place.</para>

	<para>The types of changes that are currently handled
	  are:

	<itemizedlist>
	    <listitem>
	      <para>Modifying the text of any file content line (of
	        course).</para>
	    </listitem>

	    <listitem>
	      <para>Adding new line insertions or deletions.</para>
	    </listitem>

	    <listitem>
	      <para>Adding, changing or removing context lines.  Lines
	        at the context horizon are dealt with by adjusting the
	        offset and/or count.</para>
	    </listitem>

	    <listitem>
	      <para>Adding a single hunk (@@-prefixed section).</para>
	    </listitem>

	    <listitem>
	      <para>Removing multiple hunk (@@-prefixed
	        sections).</para>
	    </listitem>
	  </itemizedlist></para>

	<para>Alternatively, if only one argument is provided, it is
	  taken to be the edited file and the counts and offsets are
	  adjusted as appropriate.  Some assumptions are made when
	  used in this mode.  See <citerefentry>
	    <refentrytitle>recountdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry> for more information.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of rediff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>interdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>recountdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="grepdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>23 Jan 2009</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>grepdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>grepdiff</refname>
	<refpurpose>show files modified by a diff containing a
          regex</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>grepdiff</command>
	  <group choice="opt">
	    <arg>-n</arg>
	    <arg>--line-number</arg>
	  </group>
	  <arg choice="opt">--number-files</arg>
	  <group choice="opt">
	    <arg>-p <replaceable>n</replaceable></arg>
	    <arg>--strip-match=<replaceable>n</replaceable></arg>
	  </group>
	  <arg choice="opt">--strip=<replaceable>n</replaceable></arg>
	  <arg choice="opt">--addprefix=<replaceable>PREFIX</replaceable></arg>
	  <group choice="opt">
	    <arg>-s</arg>
	    <arg>--status</arg>
	  </group>
	  <group choice="opt">
	    <arg>-i <replaceable>PATTERN</replaceable></arg>
	    <arg>--include=<replaceable>PATTERN</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-x <replaceable>PATTERN</replaceable></arg>
	    <arg>--exclude=<replaceable>PATTERN</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-v</arg>
	    <arg>--verbose</arg>
	  </group>
	  <group choice="opt">
	    <arg>-E</arg>
	    <arg>--extended-regexp</arg>
	  </group>
	  <group choice="opt">
	    <arg>-H</arg>
	    <arg>--with-filename</arg>
	  </group>
	  <group choice="opt">
	    <arg>-h</arg>
	    <arg>--no-filename</arg>
	  </group>
	  <arg choice="opt">--output-matching=<replaceable>WHAT</replaceable></arg>
	  <group choice="req">
	    <arg><replaceable>REGEX</replaceable></arg>
	    <arg>-f <replaceable>FILE</replaceable></arg>
	  </group>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>grepdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	    <arg>--list</arg>
	    <arg>--filter ...</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para>For each file modified by a patch, if the patch hunk
	  contains the <replaceable>REGEX</replaceable> then the
	  file's name is printed.</para>

	<para>The regular expression is treated as POSIX Basic Regular
	  Expression syntax, unless the <option>-E</option> option is
	  given in which case POSIX Extended Regular Expression syntax
	  is used.</para>

	<para>For example, to see the patches in
	  <filename>my.patch</filename> which contain the regular
	  expression <quote>pf_gfp_mask</quote>, use:

<screen><![CDATA[grepdiff pf_gfp_mask my.patch | \
  xargs -rn1 filterdiff my.patch -i]]></screen></para>

	<para>You can use both unified and context format diffs with
	  this program.</para>

      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-n</option>, <option>--line-number</option></term>
	    <listitem>
	      <para>Display the line number that each patch begins at.
	        If verbose output is requested, each matching hunk is
	        listed as well.</para>

	      <para>For a description of the output format see
	        <citerefentry>
		  <refentrytitle>lsdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--number-files</option></term>
	    <listitem>
	      <para>File numbers are listed, beginning at 1, before
	      each filename.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable>,
	    <option>--strip-match=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>When matching, ignore the first
	        <replaceable>n</replaceable> components of the
	        pathname.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--strip</option>=<replaceable>n</replaceable></term>
	    <listitem>
	      <para>Remove the first <replaceable>n</replaceable>
	        components of the pathname before displaying
	        it.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--addprefix</option>=<replaceable>PREFIX</replaceable></term>
	    <listitem>
	      <para>Prefix the pathname with
	        <replaceable>PREFIX</replaceable> before displaying
	        it.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-s</option></term>
	    <listitem>
	      <para>Show file additions, modifications and removals.
	       A file addition is indicated by a <quote>+</quote>, a
	       removal by a <quote>-</quote>, and a modification by a
	       <quote>!</quote>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-i</option> <replaceable>PATTERN</replaceable>,
	    <option>--include=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Include only files matching
	        <replaceable>PATTERN</replaceable>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-x</option> <replaceable>PATTERN</replaceable>
	    <option>--exclude=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Exclude files matching
	        <replaceable>PATTERN</replaceable>.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-E</option>,
	    <option>--extended-regexp</option></term>
	    <listitem>
	      <para>Use POSIX Extended Regular Expression
	        syntax.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-H</option>, <option>--with-filename</option></term>
	    <listitem>
	      <para>Print the name of the patch file containing each
	        match.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-h</option>, <option>--no-filename</option></term>
	    <listitem>
	      <para>Suppress the name of the patch file containing each
	        match.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-f</option> <replaceable>FILE</replaceable>,
	    <option>--file=<replaceable>FILE</replaceable></option></term>
	    <listitem>
	      <para>Read regular expressions from
	        <replaceable>FILE</replaceable>, one per line.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--output-matching</option>=hunk|file</term>
	    <listitem>
	      <para>Display the matching hunk-level or file-level
	       diffs.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of grepdiff.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--filter</option></term>
	    <listitem>
	      <para>Behave like <citerefentry>
		  <refentrytitle>filterdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry> instead.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--list</option></term>
	    <listitem>
	      <para>Behave like <citerefentry>
		  <refentrytitle>lsdiff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry> instead.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>filterdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, <citerefentry>
	    <refentrytitle>lsdiff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="recountdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>5 July 2002</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>recountdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>recountdiff</refname>
	<refpurpose>recompute patch counts and offsets</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>recountdiff</command>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>recountdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para><command>recountdiff</command> recomputes the counts and
	  offsets in unified context diffs.  It does this in two
	  passes: first the counts are corrected by inspection of
	  the actual patch lines, and then the offsets are adjusted
	  according to the computed counts.</para>

	<para>The corrected diff is sent to standard output.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of recountdiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Limitations</title>

	<para>There are some things that cannot be dealt with.  If you
	  have removed the first context line in a hunk, for example,
	  there is no way for <command>recountdiff</command> to know
	  that.  For more precise fixing up, use <citerefentry>
	    <refentrytitle>rediff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry>, which uses the original and modified
	  diffs.</para>
      </refsect1>

      <refsect1>
	<title>See also</title>

	<para><citerefentry>
	    <refentrytitle>rediff</refentrytitle>
	    <manvolnum>1</manvolnum>
	  </citerefentry></para>
      </refsect1>
    </refentry>

    <refentry id="unwrapdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>17 January 2003</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>unwrapdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>unwrapdiff</refname>
	<refpurpose>demangle word-wrapped patches</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>unwrapdiff</command>
	  <arg choice="opt">-v</arg>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>unwrapdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para><command>unwrapdiff</command> demangles patches that
	  have been word-wrapped, in an attempt to make them
	  useful.</para>

	<para>The corrected diff is sent to standard output.  Note
	  that you will probably need to use the <option>-l</option>
	  option when applying the patch, in order to ignore any
	  whitespace differences there may be.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-v</option></term>
	    <listitem>
	      <para>Verbose operation.  A list of lines that are
	        modified in a way that might be wrong is sent to
	        stderr.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of unwrapdiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Limitations</title>

	<para>Some heuristics are used to decide whether use a space
	  to recombine a wrapped line, or just join them
	  together.  Currently this is done by comparing with last
	  three characters of a line with the first two characters of
	  its continuation, and using a space if any of them are
	  different.</para>

	<para>The patch needs to have been valid before being
	  word-wrapped.</para>

	<para>The last line of a hunk is nearly always ambiguous.  If
 	  the next line begins <quote>@@</quote>,
 	  <quote>Index:&nbsp;</quote>, <quote>diff&nbsp;</quote> or
 	  <quote>---&nbsp;</quote> then it is taken to be complete;
 	  otherwise it is unwrapped using the next line.</para>
      </refsect1>
    </refentry>

    <refentry id="dehtmldiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>17 January 2003</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>dehtmldiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>dehtmldiff</refname>
	<refpurpose>get usable diff from an HTML page</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>dehtmldiff</command>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>dehtmldiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para><command>dehtmldiff</command> creates usable diff files
	  from HTML pages.  This is useful in the scenario where a
	  patch has been posted to a mailing list viewable via a web
	  page.</para>

	<para>The corrected diff is sent to standard output.  Note
	  that you will probably need to use the <option>-l</option>
	  option when applying the patch, in order to ignore any
	  whitespace differences there may be.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of dehtmldiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Limitations</title>

	<para>The way this command works is fairly primitive; it boils
	  down to some simple text replacements, followed by an
	  <command>unwrapdiff</command> pass.  As a result, all of the
	  limitations that apply to <command>unwrapdiff</command> also
	  apply to <command>dehtmldiff</command>.</para>
      </refsect1>
    </refentry>

    <refentry id="flipdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>23 January 2009</date>
      </refentryinfo>

      <refmeta>
	<refentrytitle>flipdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>flipdiff</refname>
	<refpurpose>exchange the order of two incremental patches</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>flipdiff</command>
	  <group choice="opt">
	    <arg>-p <replaceable>n</replaceable></arg>
	    <arg>--strip-match=<replaceable>n</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-U <replaceable>n</replaceable></arg>
	    <arg>--unified=<replaceable>n</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-d <replaceable>PAT</replaceable></arg>
	    <arg>--drop-context=<replaceable>PAT</replaceable></arg>
	  </group>
	  <group choice="opt">
	    <arg>-q</arg>
	    <arg>--quiet</arg>
	  </group>
	  <group choice="opt">
	    <arg>-z</arg>
	    <arg>--decompress</arg>
	  </group>
	  <group choice="opt">
	    <arg>-b</arg>
	    <arg>--ignore-space-change</arg>
	  </group>
	  <group choice="opt">
	    <arg>-B</arg>
	    <arg>--ignore-blank-lines</arg>
	  </group>
	  <group choice="opt">
	    <arg>-i</arg>
	    <arg>--ignore-case</arg>
	  </group>
	  <group choice="opt">
	    <arg>-w</arg>
	    <arg>--ignore-all-space</arg>
	  </group>
	  <arg choice="opt">--in-place</arg>
	  <arg choice="plain"><replaceable>diff1</replaceable></arg>
	  <arg choice="plain"><replaceable>diff2</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>flipdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para><command>flipdiff</command> exchanges the order of two
	  patch files that apply one after the other.  The patches
	  must be <quote>clean</quote>: the context lines must match
	  and there should be no mis-matched offsets.</para>

	<para>The swapped patches are sent to standard output, with a
	  marker line
	  (<quote><literal>===&nbsp;8&lt;&nbsp;===&nbsp;cut&nbsp;here&nbsp;===&nbsp;8&lt;&nbsp;===</literal></quote>)
	  between them, unless the <option>--in-place</option> option
	  is passed.  In that case, the output is written back to the
	  original input files.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>-p</option> <replaceable>n</replaceable>,
	    <option>--strip-match=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>When comparing filenames, ignore the first
                <replaceable>n</replaceable> pathname components from
                both patches. (This is similar to the
                <option>-p</option> option to GNU <citerefentry>
		  <refentrytitle>patch</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.)</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-q</option>, <option>--quiet</option></term>
	    <listitem>
	      <para>Quieter output.  Don't emit rationale lines at the
	       beginning of each patch.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-U</option> <replaceable>n</replaceable>,
	    <option>--unified=<replaceable>n</replaceable></option></term>
	    <listitem>
	      <para>Attempt to display <replaceable>n</replaceable>
                lines of context (requires at least
                <replaceable>n</replaceable> lines of context in both
                input files). (This is similar to the
                <option>-U</option> option to GNU
                <citerefentry>
		  <refentrytitle>diff</refentrytitle>
		  <manvolnum>1</manvolnum>
		</citerefentry>.)</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-d</option> <replaceable>pattern</replaceable>,
	    <option>--drop-context=<replaceable>PATTERN</replaceable></option></term>
	    <listitem>
	      <para>Don't display any context on files that match the
                shell wildcard <replaceable>pattern</replaceable>.
                This option can be given multiple times.</para>

	      <para>Note that the interpretation of the shell wildcard
	        pattern does not count slash characters or periods as
	        special (in other words, no flags are given to
	        <function>fnmatch</function>).  This is so that
	        <quote>*/basename</quote>-type patterns can be given
	        without limiting the number of pathname
	        components.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-i</option>, <option>--ignore-case</option></term>
	    <listitem>
	      <para>Consider upper- and lower-case to be the same.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-w</option>,
	    <option>--ignore-all-space</option></term>
	    <listitem>
	      <para>Ignore whitespace changes in patches.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-b</option>,
	    <option>--ignore-space-change</option></term>
	    <listitem>
	      <para>Ignore changes in the amount of whitespace.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-B</option>,
	    <option>--ignore-blank-lines</option></term>
	    <listitem>
	      <para>Ignore changes whose lines are all blank.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>-z</option>,
	    <option>--decompress</option></term>
	    <listitem>
	      <para>Decompress files with extensions .gz and .bz2.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--in-place</option></term>
	    <listitem>
	      <para>Write output to the original input files.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of flipdiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Limitations</title>

	<para>This is only been very lightly tested, and may not even
	  work.  Using <option>--in-place</option> is not recommended
	  at the moment.</para>

	<para>There are some cases in which it is not possible to
	 meaningfully flip patches without understanding the semantics
	 of the content.  This program only uses complete lines that
	 appear at some stage during the application of the two
	 patches, and never composes a line from parts.</para>

	<para>Because of this, it is generally a good idea to read
	 through the output to check that it makes sense.</para>
      </refsect1>
    </refentry>

    <refentry id="espdiff">
      <refentryinfo>
	<productname>patchutils</productname>
	<date>5 March 2003</date>
	<authorgroup>
	  <author>
	    <firstname>Mike</firstname>
	    <surname>Harris</surname>
	    <affiliation>
	      <address><email>mharris@redhat.com</email></address>
	    </affiliation>
	    <contrib>Conceptual designer</contrib>
	  </author>
	</authorgroup>
      </refentryinfo>

      <refmeta>
	<refentrytitle>espdiff</refentrytitle>
	<manvolnum>1</manvolnum>
      </refmeta>

      <refnamediv>
	<refname>espdiff</refname>
	<refpurpose>apply the appropriate transformation to a set of
	 patches</refpurpose>
      </refnamediv>

      <refsynopsisdiv>
	<cmdsynopsis>
	  <command>espdiff</command>
	  <arg choice="opt">--deep-brainwave-mode</arg>
	  <arg choice="opt">--recurse</arg>
	  <arg choice="opt">--compare</arg>
	  <arg choice="opt" rep="repeat"><replaceable>file</replaceable></arg>
	</cmdsynopsis>

	<cmdsynopsis>
	  <command>espdiff</command>
	  <group choice="req">
	    <arg>--help</arg>
	    <arg>--version</arg>
	  </group>
	</cmdsynopsis>
      </refsynopsisdiv>

      <refsect1>
	<title>Description</title>

	<para><command>espdiff</command> applies the appropriate
	  transformation to a patch or set of patches, depending on
	  what you intend to accomplish.</para>

	<para>The input patches must be <quote>clean</quote>: in other
	 words they must apply without fuzz or offsets in whichever
	 order they are meant to be applied.  The exception is if you
	 intend <command>espdiff</command> to clean them for
	 you (good luck).</para>

	<para>You may find it useful to cross your fingers while the
	 program performs its task, or to screw your eyes tight shut
	 while imagining it doing the right thing.</para>
      </refsect1>

      <refsect1>
	<title>Options</title>

	<variablelist>
	  <varlistentry>
	    <term><option>--deep-brainwave-mode</option></term>
	    <listitem>
	      <para>Probes your brain deeply in a manner that takes
	       longer, but produces better extra sensory
	       results.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--recurse</option></term>
	    <listitem>
	      <para>Recurses neural pathways throughout all parts of
               the brain, in some cases determining code changes you
               might make far off in the future.  You may feel a
               gentle tickling sensation when using this
               option.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--compare</option></term>
	    <listitem>
	      <para>Allows the program to scan the current directory
               examining existing patches to determine areas of code
               you are likely to change again, and concentrating on
               these areas more closely.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--help</option></term>
	    <listitem>
	      <para>Display a short usage message.</para>
	    </listitem>
	  </varlistentry>

	  <varlistentry>
	    <term><option>--version</option></term>
	    <listitem>
	      <para>Display the version number of espdiff.</para>
	    </listitem>
	  </varlistentry>
	</variablelist>
      </refsect1>

      <refsect1>
	<title>Limitations</title>

	<para>Do not use this program while sleep-walking, or before
	 your first cup of coffee.</para>

	<para>There are some cases in which it is not possible to
	 determine what the intention of the user is.  In these cases,
	 you should construct the output you desire using an editor,
	 and send the input files together with the desired output
	 file to Tim Waugh <email>twaugh@redhat.com</email> asking for
	 a new diff tool.</para>
      </refsect1>
    </refentry>
  </reference>
</book>