path: root/htdocs/docs/install_win.html

<!--#set var="revision" value="\$Id: install_win.html,v 1.7 2006-02-10 14:52:17 bjoern Exp $" -->
<!--#set var="date" value="\$Date: 2006-02-10 14:52:17 $" -->
<!--#set var="title" value="Installation Documentation for The W3C Markup Validation Service (Windows, Apache)" -->
<!--#set var="relroot" value="../" -->
<!--#include virtual="../header.html" -->
<div class="doc">
<h2>Installing the W3C-Validator on a Apache under Windows</h2>

<h3 id="TableOfContents">Table of Contents</h3>
<div id="toc">
  <ul>
    <li><a href="#about">About this guide</a></li>
    <li><a href="#needed-programs">Needed programs and other prerequisites</a></li>
    <li><a href="#directory-structure">Directory Structure</a></li>
    <li><a href="#install-programs">Installation of the Programs</a></li>
    <li><a href="#modules">Integration of the Perl-Modules</a>
      <ul>
        <li><a href="#modules-local">Integration from a local repository</a></li>
        <li><a href="#modules-remote">Integration from a remote repository</a></li>
        <li><a href="#modules-opensp">Adding the opensp152.dll</a></li>
      </ul></li>
    <li><a href="#config-apache">Configuration of the Apache Web server</a></li>
    <li><a href="#adaption-of-hosts">Adaptation of the hosts-File</a></li>
    <li><a href="#config-validator">Configuration of the Validator</a></li>
    <li><a href="#adaption-of-check">Adaptation of the check-Script</a></li>
    <li><a href="#hints">Hints</a></li>
  </ul>
</div>

<div id="install" class="stb">
<a name="skip" id="skip"></a>  
<h3>Important Note</h3>
<p>This version of the How-To refers to the current Validator-Version out of the
  CVS (HEAD). Some things may change up to the final release.<br />
  If you're using the current stable version (0.7), please use the guide at
  <a href="http://validator.w3.org/docs/install_win.html">http://validator.w3.org/docs/install_win.html</a>.</p>

<h3 id="about">About this guide</h3>

<p>This guide for the installation of the W3C Markup Validator was
contributed by David Tibbe. It is also available 
<a href="http://aktuell.de.selfhtml.org/artikel/server/validator/">in german</a>.</p>

<p>See the <a href="install.html">generic installation guide</a> for instructions on how
to install the Markup Validator on other platforms and <a href="install.html#otherguides">links
to other platform-specific guides</a>.</p>

<h3 id="needed-programs">Needed programs and other prerequisites</h3>
<p>The first prerequisite to run the validator on your system is a Web server. 
  This guide will assume that the server is already installed, and its configuration
  will not be discussed here. The server used in this guide is 
  <a href="http://httpd.apache.org/">Apache</a> 2, and the instructions should be applicable
  to most versions of Apache.</p>
  
<p>The Markup Validator itself is basically a script written in Perl, so you will
  need that, too. <a href="http://www.activestate.com/Products/Download/Download.plex?id=ActivePerl">ActivePerl</a> (version 5.8),
  is one of the options, and thanks to its installer should not be difficult to set up.</p>
  
<p>Of course, you will also need the Validator itself. It is available for 
 <a href="http://validator.w3.org/source/">download</a> as two tar-archives:
 the <a href="http://validator.w3.org/validator.tar.gz">validator itself</a>
  (~300kB), and a <a href="http://validator.w3.org/sgml-lib.tar.gz">collection of 
  <acronym title="Document Type Definition">DTDs</acronym></a>
  (~400kB).</p> 
  

<p>The validator relies on a number of Perl libraries, or "modules". 
At <a href="http://ppm.activestate.com/">Active State</a>,
  has a list of all available modules for Active Perl. In this list is also mentioned
  if a module is &quot;Core&quot; (which means it is build in) or if it
  is downloadable.<br />
  For the validator, the following modules are required:</p>
<ul>
  <li>Class-Accessor</li>
  <li>Config-General</li>
  <li>HTML-Encoding</li>
  <li>HTML-Template</li>
  <li>Net-IP</li>
  <li>SGML-Parser-OpenSP</li>
</ul>
<p>You can get each of them in a single package at ActiveState. They are also 
  packed  ready-to-install <a href="http://www.myhpi.de/~dtibbe/download/ppm.zip">in a single zip-file</a>,
  (courtesy of the guide's author).<br />
  As long as version 0.7 is current, you will find the modules for the CVS-version at
  <a href="http://www.myhpi.de/~dtibbe/download/ppm-08.zip">http://www.myhpi.de/~dtibbe/download/ppm-08.zip</a>.</p>

<p>Finally you will need some calm and patience. A whole installation of
  the validator (including the Apache and Perl) will take about an hour 
  if you do not have much experience.</p>
  
<h3 id="directory-structure">Directory Structure</h3>
<p>It might be a good idea to think about the directories the program should be
  installed to. Just clicking &quot;Next&quot; in all installation routines is not a very good idea.</p>

<p>One suggested method is to  have a directory <code>C:\www</code> in which all programs concerning the
  Web server are located (in subfolders, obviously). For instance, Apache is installed in
  <code>C:\www\Apache2</code>, Perl in <code>C:\www\perl</code>, the perl modules unzipped in
  <code>C:\www\pmm</code>, the validator itself in <code>C:\www\validator</code> and finally
  the DTD-Collection in <code>C:\www\validator\htdocs\sgml-lib</code></p>

<p>The rest of this guide will assume that these paths are being used. If you want to use another 
directory structure, you will have to adapt the instructions and sample configuration to your
own setup.</p>

<h3 id="install-programs">Installation of the Programs</h3>
<p>Next, you will be installing all the programs and libraries one after the other.</p>

<p>The Apache Web server
  should first be installed and successfully started. As mentioned, this will not be
  explained in here, but you will find a lot of How-Tos all over the net, e.g.
  in the <a href="http://httpd.apache.org/docs/2.0/platform/windows.html">Documentation on the Apache site</a>
  The only thing that is important to know is that the installation
  routine will create a subfolder <code>Apache2</code> by itself. Therefore, you will want to choose
  <code>C:\www</code> as installation directory, and Apache will be installed to
  <code>C:\www\Apache2</code>.</p>
  
<p>Perl is the next program to be installed. Because of the installer, it is
  also easy. Just make sure that you install ppm3, too. This
  program is needed for adding the Perl-modules.</p>
  
<h3 id="modules">Adding the Perl-Modules</h3>
<p>Now comes a slightly more difficult step: installing the Perl Modules.</p>
<p>ppm3 (the &quot;Programmer's
  Package Manager&quot;), which you downloaded and installed with ActivePerl, should be used.</p>
<p> 
  You can start ppm3 by clicking the
  <kbd>Start</kbd>-Button, selecting <kbd>Run...</kbd>, typing <kbd>ppm3</kbd> and pressing return.
  The program will look like this:</p>
<p style="text-align: center"><img src="<!--#echo var="relroot" -->images/ppm_1.png" alt="PPM after the startup" /></p>
<p>When typing <kbd>rep</kbd>, you will get a list of all (in)active repositories. These are
  locations for modules out of which you can download and install the needed
  ones.</p>
<p>You can add the modules in two ways: In the first case, the modules can be
  found local (e.g. extracted from the zip-archive) or in the second case they will
  be downloaded from a server (e.g. the ActiveState-Server). In the following,
  the first way will be discussed, but is is quite similar to the second one.
  So it is recommended in any case to read this section.</p>
  
<h4 id="modules-local">Integration from a local repository</h4>
<p>When ppm3 is started, it has to be told that such a repository can be found
  on the local hard disk, precise at <code>C:\www\ppm</code>. Therefore,
  <kbd>rep add local C:\www\ppm</kbd> is typed. Afterwards, a repository
  called local exists.<br />
  Since you want ppm to search packages only in local, all the others have
  to be disabled. When having a look at the output of the <kbd>rep</kbd>-command, you
  will find numbers in front of each repository. Typing <kbd>rep off 1</kbd> will
  disable the repository identified by the number 1. Do this for
  all repositories except the local one. When having finished, you might
  see a screen like this:</p>
<p style="text-align: center"><img src="<!--#echo var="relroot" -->images/ppm_2.png" alt="PPM displaying the available repositories" /></p>
<p>Now, ppm3 is ready for the installation of the modules. When typing <kbd>s *</kbd>,
  local is searched and all five modules will be found. Typing <kbd>install 1</kbd> will
  install the modules identified by the number 1. You have to repeat this for
  all modules (by typing their numbers instead, of course).<br />
  Notice: You have to install the module Class-Accessor before adding SGML-Parser-OpenSP.<br />
  The result should be similar to this:</p>
<p style="text-align: center"><img src="<!--#echo var="relroot" -->images/ppm_3.png" alt="PPM after having installed the modules" /></p>
<p>If you do not get a &quot;Successfully installed...&quot;-message as the
  last output of such an installation, something has failed and you
  should try again.<br />
  After having done that, you can leave ppm3 by typing <kbd>quit</kbd>. The
  modules are ready to use now. All programs are installed right now,
  but they do not work together yet. You have to do some configuration
  changes.</p>

<h4 id="modules-remote">Integration from a remote repository</h4>
<p>The installation from a remote repository is quite the same as from a local one.
  First, you have to check which repositories are active, but you might not need
  to change anything. After that, the procedure is the same: By typing <kbd>s *</kbd>
  you will get a list of all available modules again. But it is much longer,
  so restrict your search. Instead of typing <kbd>s *</kbd>, type e.g 
  <kbd>s Conf</kbd> for retrieving a list of all modules beginning 
  with Conf. Download and install all needed modules in this way
  (again by typing <kbd>install 5</kbd> for the 5th
  result of your search). When getting different versions of a module,
  the newest might be the best choice.</p>
  
<h4 id="modules-opensp">Adding the opensp152.dll</h4>
<p>The Perl module SGML-Parser-OpenSP requires the <acronym title="Dynamic Link Library">DLL</acronym>
  opensp152.dll. It is included in my ppm-collection. Copy it into
  your system32-dirctory (eg. C:\windows\system32).</p>
  
<h3 id="config-apache">Configuration of the Apache Web server</h3>
<p>The first file to edit is the <code>httpd.conf</code> located at 
  <code>C:\www\Apache2\conf</code>, the central configuration file 
  of your Apache. It is recommended to make a backup before editing
  it.</p>
<p>The validator pages are composed by using 
  <acronym title="Server Side Include">SSI</acronym>. Therefore, the 
  Apache has to load the required module. In
  &quot;Section&nbsp;1&nbsp;Global&nbsp;Environment&quot;, the different
  modules are loaded. SSI needs <code>mod_include</code> for working.
  The line</p>
<pre>
LoadModule include_module modules/mod_include.so
</pre>
<p>has to be unquoted (by deleting the # at the beginning of the line) or added
  completely if the module is not loaded yet.</p>
<p>The next step is to create a virtual host. It is needed, because the validator
  should be run from a different directory and logically separated from the
  default host. At the end of the <code>httpd.conf</code>, in 
  &quot;Section&nbsp;3:&nbsp;Virtual&nbsp;Hosts&quot;, the following lines have to be added:</p>
<pre>
NameVirtualHost 127.0.0.2:80

&lt;VirtualHost 127.0.0.2:80&gt;
    ServerName validator.example.org
    DocumentRoot &quot;C:/www/validator/htdocs&quot;

    ErrorLog logs/error_validator.log
    CustomLog logs/access_validator.log common

    ScriptAlias /cgi-bin &quot;C:/www/validator/httpd/cgi-bin&quot;
    ScriptAlias /check &quot;C:/www/validator/httpd/cgi-bin/check&quot;

    AddType text/html .html
    AddOutputFilter INCLUDES .html

    &lt;Directory &quot;C:/www/validator/htdocs&quot;&gt;
        Options ExecCGI Includes Indexes MultiViews
        AllowOverride None
        Order deny,allow
        Allow from localhost
    &lt;/Directory&gt;
    
    &lt;Directory &quot;C:/www/validator/httpd/cgi-bin&quot;&gt;
        Options ExecCGI Includes Indexes MultiViews
        AllowOverride None
        Order deny,allow
        Allow from localhost
    &lt;/Directory&gt;
    
&lt;/VirtualHost&gt;
</pre>
<p>The meaning of all these lines will not be discussed here. Who is
  interested in their sense can have a look to the manual or in one
  of the many How-Tos found by Google. Just some short notes about:
  The first line specifies the IP-address, the validator should run
  at. It is a loop back address, so the validator will be accessible
  only from your machine. The following lines specify the name of
  the host, the locations of the log files and some &quot;short
  cuts&quot; for the cgi-bin directory and check-script. The next 
  three lines make the Apache parsing HTML-files for SSI-directives.
  The last two sections set some permission for the used
  directories.<br />
  The files <code>error_validator.log</code> and
  <code>access_validator.log</code> can be found in the 
  <code>C:\www\Apache2\logs</code>-directory. They log every 
  request and error occurring on this host and give you some 
  helpful advices in these cases.</p>
<p>Finally, the Apache has to be restarted so that the changes
  take effect. You can do that by choosing the shortcut in the 
  Apache-program group (<kbd>Start</kbd>, <kbd>Programs</kbd>,
  <kbd>Apache HTTP Server</kbd>, <kbd>Control Apache Server</kbd>
  <kbd>Restart</kbd>). A DOS-box will appear shortly;
  when it disappears, the Apache is restarted.</p>
<p>When you call http://127.0.0.2/ in your browser, you should 
  see the well-known site from 
  <a href="http://validator.w3.org/">http://validator.w3.org/</a>.
  In the configuration file of the Apache, a name for the Virtual
  Host was defined (validator.example.org), but it is not 
  resolved yet. That is changed in the next step.</p>

<h3 id="adaption-of-hosts">Adaptation of the hosts-File</h3>
<p>The <code>hosts</code>-file can be seen as a local
  DNS-configuration. It is located in at
  <code>C:\windows\hosts</code> on Win9x and at
  <code>C:\Windows\system32\drivers\etc\hosts</code> on
  WinXP. It might be possible, that the file is missing,
  but a file <code>hosts.sam</code> can be found instead.
  In that case you have to rename it by deleting the file
  extension and its leading dot.</p>
<p>When opening it in an editor, you will find a leading 
  comment in there. A line</p>
<pre>
127.0.0.1 localhost
</pre>
<p>follows.</p>
<p>That line means that a request to localhost is redirected
  to 127.0.0.1 (i.e. calling http://localhost/ in your browser
  effects a request to http://127.0.01/).</p>
<p>Edit the file to the following:</p>
<pre>
127.0.0.1 localhost
127.0.0.1 www.example.org
127.0.0.2 validator.example.org
</pre>
<p>After these changes, the server is available at 
  http://localhost/ but can be accessed also at 
  http://www.example.org/, too. Requests for 
  http://validator.example.org/ are redirected to 
  http://127.0.0.2/.</p>
<p>The server configuration is finished right now. But if you 
  try to validate a page, you will get an &quot;Internal Server Error&quot;,
  because the <code>check</code>-script has not been configured yet.</p>

<h3 id="config-validator">Configuration of the Validator</h3>
<p>In the directory <code>c:\www\validator\htdocs\config</code>, a file 
  called <code>validator.conf</code> can be found. Open it with the editor.
  Lines beginning with # are comments.</p>
<p>Because of using a Windows-system, we have to define a
  base path to the validator's directory. That is done by
  uncommenting the line</p>
<pre>
#Base = /usr/local/validator
</pre>
<p>and changing it to</p>
<pre>
Base = C:/www/validator
</pre>
<p>For the SGML <code>Library</code>, <code>c:/www/sgml-lib</code>
  is declared. Notice that slashes are used here instead of 
  windows-like back slashes.</p>

<p>The option <code>Home Page</code> has to be changed to the URL of 
  the validator, http://validator.example.org/ in our case.</p>
<p>The last option that has to be modified is Allow Private 
  <code>IPs = { no | yes }</code>. It must be set to
  &quot;yes&quot;. Otherwise, you cannot validate files 
  from the local PC and you will get just an access fault
  because of security reasons.</p>
<p>After that, the configuration file will look like this:</p>
<pre>
#
# Main Configuration File for the W3C Markup Validation Service.
#
# $Id: install_win.html,v 1.7 2006-02-10 14:52:17 bjoern Exp $
#
# See 'perldoc Config::General' for the syntax, and be aware that the
# 'SplitPolicy' is 'equalsign', ie. keys and values are separated by '\s*=\s*',
# and that 'InterPolateVars' is in effect.
#

#
# Base Path for Markup Validator files.
#
# You MUST set these unless you use the default locations for the files.
# e.g. the config files in &quot;/etc/w3c/&quot; and everything else in
# &quot;/usr/local/validator/&quot;.
#
# Make sure all file paths below do NOT end with a slash

&lt;Paths&gt;
  #
  # Base path.  Defaults to the value of the W3C_VALIDATOR_HOME environment
  # variable or /usr/local/validator if the variable does not exist.
  Base = C:/www/validator

  #
  # Location of template files
  Templates = $Base/share/templates

  &lt;SGML&gt;
    #
    # The SGML Library Path.
    Library = $Base/htdocs/sgml-lib
  &lt;/SGML&gt;
&lt;/Paths&gt;

#
# This controls whether the debugging options are allowed to be enabled. 
Allow Debug = yes

#
# This lets you permanently enable the debugging options. Can be overridden
# with CGI options (unlike &quot;Allow Debug&quot; above).
Enable Debug = no

#
# Whether private RFC1918 addresses are allowed.
Allow Private IPs = no

#
# Whether the (highly experimental!) SOAP support should be enabled.
Enable SOAP = no

#
# Whether the validator will check its own output.
# 0 means it will refuse to check its own output, 1 means it will but it will
# refuse to check the results of it checking itself. Etc.
Max Recursion = 0

#
# Protocols the validator is allowed to use for retrieving documents.
# The default is to allow http and https.
&lt;Protocols&gt;
  Allow = data,http,https
&lt;/Protocols&gt;

#
# Email address of the maintainer of this service.
Maintainer = validator-admin@example.org

#
# The "Home Page" for the service.  Make sure this ends with a slash.
Home Page = http://validator.example.org

#
# Base URI for the Element Reference.
Element Ref URI = http://www.htmlhelp.com/reference/html40/



#
# Mapping tables etc...
#

#
# Maps element names to URLs (cf. &quot;Element Ref URI&quot; above).
&lt;Elements&gt;
  Include eref.cfg
&lt;/Elements&gt;

#
# Main document Type Registry; contains all information on the types
# of documents we support and how they are processed.
&lt;Types&gt;
  Include types.conf
&lt;/Types&gt;

#
# Mapping of charset names to their IANA names and how iconv(3) knows them.
&lt;Charsets&gt;
  Include charset.cfg
&lt;/Charsets&gt;

#
# Map MIME Media Type to Parse Mode mapping.
&lt;MIME&gt;
  text/xml              = XML
  image/svg             = XML
  image/svg+xml         = XML
  application/smil      = XML
  application/xml       = XML
  text/html             = TBD
  text/vnd.wap.wml      = XML
  application/xhtml+xml = XML
  application/mathml+xml = XML
&lt;/MIME&gt;

#
# Source for the "Tip of The Day" blurbs.
&lt;Tips&gt;
  Include tips.cfg
&lt;/Tips&gt;
</pre>
<p>Now, the validator has been successfully configured. But it
  is not ready for use yet, some lines in the validator script
  itself have to be changed before.</p>

<h3 id="adaption-of-check">Adaptation of the check-Script</h3>
<p>The following changes are necessary because the script is 
  written for a Unixserver where some things are different to
  a Windows-system.</p>
<p>The <code>check</code>-script located in the directory 
  <code>C:\www\validator\httpd\cgi-bin</code> can be opened
  with the editor. I will not mention any line numbers in 
  the following steps, because they might differ in
  later versions. Some advices are always placed in the lines above
  in the script so that you can orientate yourself to those lines.</p>
<p>The very first line of the script has to be changed to</p>
<pre>
#!c:/www/perl/bin/perl.exe
</pre>
<p>This is the path to the Perl interpreter, up to now in
  Unix-style. So it has to be changed to Windows-style. The
  parameter <code>-T</code> is replaced by that, too.</p>
<p>In a next step, the script is told where to find the 
  configuration file. That is done after the comment in these lines</p>
<pre>
#
# Read Config Files.
eval {
  my %config_opts = (
     -ConfigFile =&gt; ($ENV{W3C_VALIDATOR_CFG} || '/etc/w3c/validator.conf'),
</pre>
<p>We do not define an environment variable but specify the full path
  to the file. So the lines have to be changed to</p>
<pre>
#
# Read Config Files.
eval {
  my %config_opts = (
     -ConfigFile =&gt; ('C:/www/validator/htdocs/config/validator.conf'),
</pre>     
<p>After saving the script, you can use it at http://validator.example.org/
  as you know it from http://validator.w3.org/.</p>
<p>That is it, your own validator is working.</p>

<h3 id="hints">Hints</h3>
<p>In further versions of the validator, some other Perl modules
  will be needed perhaps. They can be downloaded by ppm3. You 
  can discover that case very easily: When trying to run the script,
  you will get an output like</p>
<pre>
Can't locate Config/General.pm in @INC (@INC contains: C:/www/perl/lib C:/www/perl/site/lib .)
    at C:/www/validator/httpd/cgi-bin/check line 46.
BEGIN failed--compilation aborted at C:/www/validator/httpd/cgi-bin/check line 46.
</pre>  
<p>It is easy to see that the missing module is &quot;Config General&quot;,
  that has to be installed.</p>
<p>Windows XP with Service Pack 2 installed might have problems with the
  loop back address 127.0.0.2. The Problem and its solution are 
  described at <a href="http://support.microsoft.com/default.aspx?kbid=884020">http://support.microsoft.com/default.aspx?kbid=884020</a>.</p>
<p>If you do have any further questions or suggestions to improve this guide,
 you can use our <a href="../feedback.html">feedback</a> channels.</p>
</div>
</div>
<!--#include virtual="../footer.html" -->
  </body>
</html>