Add match-all.action to the files section.

[privoxy.git] / doc / source / privoxy-man-page.sgml

<!--
 File        :  $Source: /cvsroot/ijbswa/current/doc/source/privoxy-man-page.sgml,v $

 Purpose     :  Manual Page
                This file belongs into
                ijbswa.sourceforge.net:/home/groups/i/ij/ijbswa/htdocs/
                
 $Id: privoxy-man-page.sgml,v 2.24 2009/02/12 16:08:26 fabiankeil Exp $

 Copyright (C) 2001-2009 Privoxy Developers http://www.privoxy.org/
 See LICENSE.

 ========================================================================
 NOTE: Please read developer-manual/documentation.html before touching 
 anything in this, or other Privoxy documentation. 
 ========================================================================

 Doc NOTES: This is some tricky markup! There are some quirks
 to how this markup is handled. It is not always so co-operative.
 Please don't change the markup unless you can verify the changes 
 will improve finished output! 
 
 literallayout tags are particularly sensitive to where they are placed.
 The 'replaceable' and 'command' tags are used here somewhat unconventionally,
 since it seems to generate the proper formatting (at least for me :).

 Create man page: 'make man'

 Requires docbook2man (short perl script), see CVS
 http://sources.redhat.com/docbook-tools/. Also requires openjade and SGMLSpm
 perl module. 
 
 For man page references, see:
 http://www.linuxdoc.org/HOWTO/mini/DocBook-Install/using.html
 http://docbook.org/tdg/en/html/ch02.html#making-refentry

-->
<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook V4.1//EN"[
<!entity % dummy "IGNORE"> 
<!entity p-intro SYSTEM "privoxy.sgml">
<!entity seealso SYSTEM "seealso.sgml">
<!entity copyright SYSTEM "copyright.sgml">
<!entity license SYSTEM "license.sgml">
<!entity authors SYSTEM "p-authors.sgml">
<!entity p-version "3.0.11">
<!entity p-status "stable">
<!entity % p-not-stable "IGNORE">
<!entity % p-stable "INCLUDE">
<!entity % p-text "IGNORE">           <!-- define we are not a text only doc -->
<!entity % p-authors-formal "IGNORE"> <!-- exclude additional formating      -->
<!entity my-copy "(C)">               <!-- db2man barfs on copyright symbol  -->
]>

<refentry id="privoxy">
<refentryinfo>
 <date>2009-02-15</date>
</refentryinfo>
<refmeta>
 <refentrytitle>privoxy</refentrytitle> 
 <manvolnum>1</manvolnum>
 <refmiscinfo>
  Privoxy &p-version;<![%p-not-stable;[ &p-status;]]>
 </refmiscinfo>
</refmeta>

<refnamediv>
 <refname><application>privoxy</application></refname>
 <refpurpose>Privacy Enhancing Proxy</refpurpose>
</refnamediv>

<refsynopsisdiv>
 <cmdsynopsis> 
  <command>privoxy</command>
  <arg><option>--help</option></arg>
  <arg><option>--version</option></arg>
  <arg><option>--no-daemon</option></arg>
  <arg><option>--pidfile </option><replaceable class="parameter">pidfile</replaceable></arg>  
  <arg><option>--user </option><replaceable class="parameter">user[.group]</replaceable></arg> 
  <arg><option>--chroot</option></arg>
  <arg><option>--pre-chroot-nslookup </option><replaceable class="parameter">hostname</replaceable></arg> 
  <arg><replaceable class="parameter">configfile</replaceable></arg>        
 </cmdsynopsis>

</refsynopsisdiv>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Options</title>
 <para>
  <command>Privoxy</command> may be invoked with the following command line
  options:
 </para>

 <variablelist>                                          
  <varlistentry>
    <term>--help</term>
      <listitem>
       <para>
         Print brief usage info and exit.
        </para>
      </listitem>
  </varlistentry>

  <varlistentry>
    <term>--version</term>
      <listitem>
       <para>
         Print version info and exit.
        </para>
      </listitem>
  </varlistentry>

  <varlistentry>
    <term>--no-daemon</term>
     <listitem>
      <para>
        Don't  become  a daemon, i.e. don't fork and become process group
        leader, don't detach from controlling tty, and do all logging there.
      </para>
    </listitem>
  </varlistentry>

  <varlistentry>
    <term>--pidfile <replaceable class="parameter">pidfile</replaceable></term>
     <listitem>
      <para>
        On startup, write the process ID to <replaceable class="parameter">pidfile</replaceable>.
        Delete the <replaceable class="parameter">pidfile</replaceable> on exit.
        Failure to create or delete the <replaceable class="parameter">pidfile</replaceable>
        is non-fatal. If no <command>--pidfile</command> option is given, no PID file will be used.
      </para>
    </listitem>
  </varlistentry>

  <varlistentry>
    <term>--user <replaceable class="parameter">user[.group]</replaceable></term>
     <listitem>
      <para>
       <!-- Note: replaceable is maybe the wrong tag, but generates  -->
       <!-- correct looking man output.                              -->
       After (optionally) writing the PID file, assume the user ID of
       <replaceable class="parameter">user</replaceable> and the GID of
       <replaceable class="parameter">group</replaceable>, or, if the optional
       <replaceable class="parameter">group</replaceable> was not given, the default group of
       <replaceable class="parameter">user</replaceable>. Exit if the privileges are not
       sufficient to do so.
     </para>
    </listitem>
  </varlistentry>
  <varlistentry>
    <term>--chroot</term>
     <listitem>
      <para>
       Before changing to the user ID given in the --user option, chroot to
       that user's home directory, i.e. make the kernel pretend to the
       <command>Privoxy</command> process that the directory tree starts
       there. If set up carefully, this can limit the impact of possible
       vulnerabilities in <command>Privoxy</command> to the files contained in
       that hierarchy. 
      </para>
    </listitem>
  </varlistentry>
  <varlistentry>
    <term>--pre-chroot-nslookup <replaceable class="parameter">hostname</replaceable></term>
     <listitem>
      <para>
        Initialize the resolver library using <replaceable class="parameter">hostname</replaceable>
        before chroot'ing. On some systems this reduces the number of files
        that must be copied into the chroot tree.
     </para>
    </listitem>
  </varlistentry>
 </variablelist>

 <para>
  If the <filename>configfile</filename> is not specified on  the  command  line,
  <command>Privoxy</command>  will  look for a file named
  <filename>config</filename> in the current directory . If no
  <filename>configfile</filename> is found, <command>Privoxy</command> will 
  fail to start.
 </para>

</refsect1>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Description</title>
<!-- Include privoxy.sgml boilerplate: -->
 &p-intro;
<!-- end boilerplate -->
</refsect1>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Installation and Usage</title>
<para>
 Browsers can either be individually configured to use
 <command>Privoxy</command> as a HTTP proxy (recommended),
 or <command>Privoxy</command> can be combined with a packet
 filter to build an intercepting proxy
 (see <filename>config</filename>).  The default setting is  for
 localhost,  on port  8118 (configurable in the main config file).  To set the
 HTTP proxy in Netscape and Mozilla, go through:  <command>Edit</command>;
 <command>Preferences</command>;  <command>Advanced</command>;
 <command>Proxies</command>;  <command>Manual Proxy Configuration</command>;
 <command>View</command>. 
</para>
<para>
 For Firefox, go through: <command>Tools</command>; 
 <command>Options</command>; <command>General</command>;
 <command>Connection Settings</command>;
 <command>Manual Proxy Configuration</command>. 
</para>
<para>
 For Internet Explorer, go through: <command>Tools</command>; 
 <command>Internet Properties</command>; <command>Connections</command>;
 <command>LAN Settings</command>. 
</para>
<para>
 The Secure (SSL) Proxy should also be set to the same values, otherwise
 https: URLs will not be proxied. Note: <command>Privoxy</command> can only
 proxy HTTP and HTTPS traffic. Do not try it with FTP or other protocols.
 HTTPS presents some limitations, and not all features will work with HTTPS 
 connections.
</para>

<para>
 For other browsers, check the documentation.
</para>
</refsect1>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Configuration</title>
<para>
 <command>Privoxy</command> can be configured with the various configuration
 files. The default configuration files are: <filename>config</filename>,
 <filename>default.filter</filename>, and
 <filename>default.action</filename>. <filename>user.action</filename> should 
 be used for locally defined exceptions to the default rules of
 <filename>default.action</filename>, and <filename>user.filter</filename> for 
 locally defined filters. These are well commented.  On Unix
 and Unix-like systems, these are located in
 <filename>/etc/privoxy/</filename> by default. 
</para>
<para> 
 <command>Privoxy</command> uses the concept of <command>actions</command> 
 in order to manipulate the data stream between the browser and remote sites.
 There are various actions available with specific functions for such things 
 as blocking web sites, managing cookies, etc. These actions can be invoked
 individually or combined, and used against individual URLs, or groups of URLs 
 that can be defined using wildcards and regular expressions. The result is
 that the user has greatly enhanced control and freedom.
</para>
<para>
 The actions list (ad blocks, etc) can also be configured with your
 web browser at <ulink url="http://config.privoxy.org/">http://config.privoxy.org/</ulink> 
 (assuming the configuration allows it).
 <command>Privoxy's</command> configuration parameters  can also  be viewed at
 the same page. In addition, <command>Privoxy</command> can be toggled on/off.
 This is an internal page, and does not require Internet access.
</para>
<para>
 See the <ulink
 url="http://www.privoxy.org/user-manual/"><citetitle>User Manual</citetitle></ulink> for a detailed
 explanation of installation, general usage, all configuration options, new
 features and notes on upgrading.
</para>
</refsect1>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Sample Configuration</title>
<para>
 A brief example of what a simple <filename>default.action</filename>
 configuration might look like:
</para>

<literallayout>
 # Define a few useful custom aliases for later use
 {{alias}}

 # Useful aliases that combine more than one action
 +crunch-cookies = +crunch-incoming-cookies +crunch-outgoing-cookies
 -crunch-cookies = -crunch-incoming-cookies -crunch-outgoing-cookies
 +block-as-image = +block +handle-as-image

 # Fragile sites should have the minimum changes
 fragile     = -block -deanimate-gifs -fast-redirects -filter \
               -hide-referer -prevent-cookies

 ## Turn some actions on ################################
 ## NOTE: Actions are off by default, unless explictily turned on 
 ## otherwise with the '+' operator.

{ \
+change-x-forwarded-for{block} \
+deanimate-gifs{last} \
+filter{refresh-tags} \
+filter{img-reorder} \
+filter{banners-by-size} \
+filter{webbugs} \
+filter{jumping-windows} \
+filter{ie-exploits} \
+hide-from-header{block} \
+hide-referrer{conditional-block} \
+session-cookies-only \
+set-image-blocker{pattern} \
}
/ # '/' Match *all* URL patterns

 
 # Block all URLs that match these patterns
 { +block }
  ad.
  ad[sv].
  .*ads.
  banner?.
  /.*count(er)?\.(pl|cgi|exe|dll|asp|php[34]?)
  .hitbox.com 
  media./.*(ads|banner)

 # Block, and treat these URL patterns as if they were 'images'.
 # We would expect these to be ads.
 { +block-as-image }
  .ad.doubleclick.net
  .a[0-9].yimg.com/(?:(?!/i/).)*$
  ad.*.doubleclick.net

 # Make exceptions for these harmless ones that would be 
 # caught by our +block patterns just above.
 { -block }
  adsl.
  adobe.
  advice.
  .*downloads.
  # uploads or downloads
  /.*loads
</literallayout>

<para>
 Then for a <filename>user.action</filename>, we would put local,
 narrowly defined exceptions:
</para>

<literallayout>
 # Re-define aliases as needed here
 {{alias}}

 # Useful aliases
 -crunch-cookies = -crunch-incoming-cookies -crunch-outgoing-cookies
 
 # Set personal exceptions to the policies in default.action #######

 # Sites where we want persistent cookies, so allow *all* cookies
 { -crunch-cookies -session-cookies-only }
  .redhat.com
  .sun.com
  .msdn.microsoft.com
 
 # These sites break easily. Use our "fragile" alias here.
 { fragile }
  .forbes.com
  mybank.example.com

 # Replace example.com's style sheet with one of my choosing
 { +redirect{http://localhost/css-replacements/example.com.css} }
  .example.com/stylesheet.css

</literallayout>

<para>
 See the comments in the configuration files themselves, or the 
 <citetitle>User Manual</citetitle>
 for full explanations of the above syntax, and other <command>Privoxy</command>
 configuration options.
</para>

</refsect1>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Files</title>
<!-- this is a cheesy way to do this, but WTF. -->
<literallayout> 
 <filename>/usr/sbin/privoxy</filename>
 <filename>/etc/privoxy/config</filename>
 <filename>/etc/privoxy/match-all.action</filename>
 <filename>/etc/privoxy/default.action</filename>
 <filename>/etc/privoxy/user.action</filename>
 <filename>/etc/privoxy/default.filter</filename>
 <filename>/etc/privoxy/user.filter</filename>
 <filename>/etc/privoxy/trust</filename>
 <filename>/etc/privoxy/templates/*</filename>
 <filename>/var/log/privoxy/logfile</filename>
</literallayout>

<para>
 Various other files should be included, but may vary depending on platform
 and build configuration. Additional documentation should be included in the local
 documentation directory.
</para>

</refsect1>


<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Signals</title>
<para>
 <!-- command tag is used here to get proper looking format  -->
 <command>Privoxy</command> terminates on the <command>SIGINT</command>,
 <command>SIGTERM</command> and <command>SIGABRT</command> signals. Log
 rotation scripts may cause a re-opening of the logfile by sending a 
 <command>SIGHUP</command> to <command>Privoxy</command>. Note that unlike
 other daemons,  <command>Privoxy</command> does not need to be made aware of
 config file changes by <command>SIGHUP</command> -- it will detect them
 automatically. 
</para>

</refsect1>

<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Notes</title>
<![%p-not-stable;[
<para>
 This is a &p-status; version of <command>Privoxy</command>. Not 
 all features are well tested.
</para>]]>
<para>
 Please see the <citetitle>User Manual</citetitle> on how to contact the
 developers, for feature requests, reporting problems, and other questions.
</para>

</refsect1>

<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>See Also</title>
<!-- Include seealso.sgml boilerplate: -->
 &seealso;
<!-- end boilerplate -->
</refsect1>

<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Development Team</title>
<!-- Include p-authors.sgml boilerplate: -->
 &authors;
<!-- end boilerplate -->
</refsect1>

<!--   ~~~~~       New section      ~~~~~     -->
<refsect1><title>Copyright and License</title>

<refsect2><title>Copyright</title>
<!-- Include copyright.sgml boilerplate: -->
 &copyright;
<!-- end boilerplate -->
</refsect2>

<refsect2><title>License</title>
<!-- Include license.sgml boilerplate: -->
 &license;
<!-- end boilerplate -->
</refsect2>
</refsect1>

</refentry>