<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN"
"http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd">
<book>
<title>README: Web-based Help from DocBook XML</title>
<bookinfo>
<legalnotice>
<para>Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation files
(the <quote>Software</quote>), to deal in the Software without
restriction, including without limitation the rights to use, copy,
modify, merge, publish, distribute, sublicense, and/or sell copies of
the Software, and to permit persons to whom the Software is furnished to
do so, subject to the following conditions: <itemizedlist>
<listitem>
<para>The above copyright notice and this permission notice shall
be included in all copies or substantial portions of the
Software.</para>
</listitem>
<listitem>
<para>Except as contained in this notice, the names of individuals
credited with contribution to this software shall not be used in
advertising or otherwise to promote the sale, use or other
dealings in this Software without prior written authorization from
the individuals in question.</para>
</listitem>
<listitem>
<para>Any stylesheet derived from this Software that is publicly
distributed will be identified with a different name and the
version strings in any derived Software will be changed so that no
possibility of confusion between the derived package and this
Software will exist.</para>
</listitem>
</itemizedlist></para>
<formalpara>
<title>Warranty:</title>
<para>THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
IN NO EVENT SHALL DAVID CRAMER, KASUN GAJASINGHE, OR ANY OTHER CONTRIBUTOR BE LIABLE FOR
ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.</para>
</formalpara>
<para>This package is maintained by Kasun Gajasinghe, <email>kasunbg AT
gmail DOT com</email> and David Cramer, <email>david AT thingbag DOT
net</email>.</para>
<para>This package also includes the following software written and
copyrighted by others:<itemizedlist>
<listitem>
<para>Files in <filename
class="directory">template/common/jquery</filename> are
copyrighted by <ulink url="http://jquery.com/">JQuery</ulink>
under the MIT License. The file
<filename>jquery.cookie.js</filename> Copyright (c) 2006 Klaus
Hartl under the MIT license.</para>
<indexterm>
<primary>jquery</primary>
</indexterm>
</listitem>
<listitem>
<para>Some files in the <filename
class="directory">template/content/search</filename> and <filename
class="directory">indexer</filename> directories were originally
part of N. Quaine's htmlsearch DITA plugin. The htmlsearch DITA
plugin is available from the <ulink
url="http://tech.groups.yahoo.com/group/dita-users/files/Demos/">files
page</ulink> of the DITA-users yahoogroup. The htmlsearch plugin
was released under a BSD-style license. See
<filename>indexer/license.txt</filename> for details. <indexterm>
<primary>htmlsearch</primary>
</indexterm> <indexterm>
<primary>DITA</primary>
<secondary>htmlsearch plugin</secondary>
</indexterm></para>
</listitem>
<listitem>
<para>Stemmers from the <ulink
url="http://snowball.tartarus.org/texts/stemmersoverview.html">Snowball
project</ulink> released under a BSD license.</para>
</listitem>
<listitem>
<para>Code from the <ulink url="http://lucene.apache.org/">Apache
Lucene</ulink> search engine provides support for tokenizing
Chinese, Japanese, and Korean content released under the Apache
2.0 license. </para>
</listitem>
</itemizedlist>
Webhelp for DocBook was developed as a <ulink url="http://socghop.appspot.com">Google Summer of Code</ulink> project.
</para>
</legalnotice>
<copyright>
<year>2008-2010</year>
<holder>Kasun Gajasinghe</holder>
<holder>David Cramer</holder>
</copyright>
<author>
<firstname>David</firstname>
<surname>Cramer</surname>
<email>dcramer AT motive DOT com</email>
<email>david AT thingbag DOT net</email>
</author>
<author>
<firstname>Kasun</firstname>
<surname>Gajasinghe</surname>
<email>kasunbg AT gmail DOT com</email>
</author>
<pubdate>August 2010</pubdate>
</bookinfo>
<chapter>
<chapterinfo>
<abstract>
<!-- This becomes the brief description that appears in search results UNLESS there's a para or phrase with role="summary". If there is, then the role="summary" text wins. -->
<para>Overview of the package.</para>
</abstract>
</chapterinfo>
<title>Introduction</title>
<para>A common requirement for technical publications groups is to produce a Web-based help
format that includes a table of contents pane, a search feature, and an index similar to what
you get from the Microsoft HTML Help (.chm) format or Eclipse help. If the content is help for
a Web application that is not exposed to the Internet or requires that the user be logged in,
then it is impossible to use services like Google to add search. <indexterm class="singular">
<primary>features</primary>
</indexterm>
<itemizedlist>
<title>Features</title>
<listitem>
<para>Full text search.<indexterm class="singular">
<primary>search</primary>
<secondary>features</secondary>
</indexterm></para>
<itemizedlist>
<listitem>
<para>Stemming support for English, French, and German. Stemming support can be added
for other languages by implementing a stemmer.<indexterm class="singular">
<primary>search</primary>
<secondary>stemming</secondary>
</indexterm></para>
</listitem>
<listitem>
<para>Support for Chinese, Japanese, and Korean using code from the Lucene search
engine. </para>
</listitem>
<listitem>
<para>Search highlighting shows where the searched for term appears in the results.
Use the <guibutton>H</guibutton> button to toggle the highlighting off and on.
<indexterm class="singular">
<primary>search</primary>
<secondary>highlighting</secondary>
</indexterm></para>
</listitem>
<listitem>
<para>Search results can include brief descriptions of the target.<indexterm
class="singular">
<primary>search</primary>
<secondary>descriptions</secondary>
</indexterm></para>
</listitem>
</itemizedlist>
</listitem>
<listitem>
<para>Table of contents pane with collapsible toc tree.</para>
</listitem>
<listitem>
<para>Auto-synchronization of content pane and TOC.</para>
</listitem>
<listitem>
<para>TOC and search pane implemented without the use of a frameset.</para>
</listitem>
<listitem>
<para>An Ant <filename>build.xml</filename> file to generate output. You can use this
build file by importing it into your own or use it as a model for integrating this
output format into your own build system.</para>
</listitem>
</itemizedlist>
<itemizedlist>
<title>Possible future enhancements</title>
<listitem>
<para>Move webhelp-specific parameters and gentext strings into base DocBook stylesheets.
</para>
</listitem>
<listitem>
<para>Use <sgmltag class="attribute">tabindex</sgmltag> attributes to control the tab
order in the output. The Contents and Search tabs should be first and second, then the
search box and button, then the table of contents items, and so on.</para>
</listitem>
<listitem>
<para>Add "Expand all" and "Collapse all" buttons to the table of contents.</para>
</listitem>
<listitem>
<para>Add other search options:</para>
<itemizedlist>
<listitem>
<para>Add an option to use Lucene for server-side searches with table of contents
state persisted on the server.</para>
</listitem>
<listitem>
<para>Add a simple form that uses a Google site:my.domain.com based search.</para>
</listitem>
</itemizedlist>
</listitem>
<listitem>
<para>Sort search results based on relevance</para>
</listitem>
<listitem>
<para>Support wild card characters in the search query.</para>
</listitem>
<listitem>
<para>Parameterize width of the TOC pane OR make the TOC pane resizeable by the
user.</para>
</listitem>
<listitem>
<para>Automate search results summary text:</para>
<itemizedlist>
<listitem>
<para>Automatically use the first non-heading content as the summary in the search
results.</para>
</listitem>
<listitem>
<para>Automatically limit the size of the search description to something 140
characters.</para>
</listitem>
</itemizedlist>
</listitem>
<listitem>
<para>Support boolean operators in search.</para>
</listitem>
<listitem>
<para>Parameterize list of files to exclude from indexing. Currently it's hard coded that
we don't index <filename>index.html </filename>and <filename>ix01.html</filename> (the
legal notice and index topics). It should be smarter and automatically not index the
index file even if it's not named <filename>ix01.html</filename>.</para>
</listitem>
<listitem>
<para>Improve performance by moving the table of contents div out of each page and into a
separate JavaScript file which then adds it to the page.</para>
</listitem>
<listitem>
<para>Add to the indexer the ability to specify a list of files or file patterns not to
index. Currently it does not index <filename>index.html</filename> or
<filename>ix01.html</filename>, which is generally appropriate, but it should be up to
the user to decide.</para>
</listitem>
<listitem>
<para>Add an index tab populated by a separate JavaScript file. Include a param/property
that allows the content creator to disable the index.</para>
</listitem>
<listitem>
<para>Add functionality to the <filename>build.xml</filename> file so that when a property
is set, the build generates a pdf version of the document and includes a link to it from
the header.</para>
</listitem>
<listitem>
<para>Add <ulink
url="http://www.comparenetworks.com/developers/jqueryplugins/jbreadcrumb.html"
>breadcrumbs</ulink> so the user will know what topics he's been to.</para>
</listitem>
<listitem>
<para>Consider using more advanced Lucene indexers for Chinese and Japanese than the
CJKAnalyzer</para>
</listitem>
</itemizedlist></para>
</chapter>
<chapter>
<title>Using the package</title>
<para role="summary">The following sections describe how to install and
use the package on Windows.</para>
<section>
<sectioninfo>
<abstract>
<para>Installation instructions</para>
</abstract>
</sectioninfo>
<title>Generating webhelp output</title>
<procedure>
<title>To install the package on Windows</title>
<note>
<para>The examples in this procedure assume a Windows installation,
but the process is the same in other environments,
<foreignphrase>mutatis mutandis</foreignphrase>.</para>
</note>
<step>
<para>If necessary, install <ulink
url="http://www.java.com/en/download/manual.jsp">Java 1.6</ulink> or
higher.</para>
<substeps>
<step>
<para>Confirm that Java is installed and in your
<envar>PATH</envar> by typing the following at a command prompt:
<programlisting>java -version</programlisting></para>
<note>
<para>To build the indexer, you must have the JDK.</para>
</note>
</step>
</substeps>
</step>
<step>
<para>If necessary, install <ulink
url="http://ant.apache.org/bindownload.cgi">Apache Ant</ulink> 1.6.5
or higher.</para>
<substeps>
<step>
<para>Unzip the Ant binary distribution to a convenient location
on your system. For example: <filename>c:\Program
Files</filename>.</para>
</step>
<step>
<para>Set the environment variable <envar>ANT_HOME</envar> to
the top-level Ant directory. For example: <filename>c:\Program
Files\apache-ant-1.7.1</filename>. <tip>
<para>See <ulink
url="http://support.microsoft.com/kb/310519">How To Manage
Environment Variables in Windows XP</ulink> for information
on setting environment variables.</para>
</tip></para>
</step>
<step>
<para>Add the Ant <filename>bin</filename> directory to your
<envar>PATH</envar>. For example: <filename>c:\Program
Files\apache-ant-1.7.1\bin</filename></para>
</step>
<step>
<para>Confirm that Ant is installed by typing the following at a
command prompt: <programlisting>ant -version</programlisting></para>
<note>
<para>If you see a message about the file
<filename>tools.jar</filename> being missing, you can safely
ignore it.</para>
</note>
</step>
</substeps>
</step>
<step>
<para>Download <ulink url="http://prdownloads.sourceforge.net/saxon/saxon6-5-5.zip">Saxon
6.5.x</ulink> and unzip the distribution to a convenient location on your file system.
You will use the path to <filename>saxon.jar</filename> in <xref
linkend="edit-build-properties"/> below.<note>
<para>The <filename>build.xml</filename> has only been tested with Saxon 6.5, though
it could be adapted to work with other XSLT processors. However, when you generate
output, the Saxon jar must <emphasis role="bold">not</emphasis> be in your
<envar>CLASSPATH</envar>.</para>
</note></para>
</step>
<step id="edit-build-properties">
<para>In a text editor, edit the
<filename>build.properties</filename> file in the webhelp directory
and make the changes indicated by the comments:<programlisting># The path (relative to the build.xml file) to your input document.
# To use your own input document, create a build.xml file of your own
# and import this build.xml.
input-xml=docsrc/readme.xml
# The directory in which to put the output files.
# This directory is created if it does not exist.
output-dir=docs
# If you are using a customization layer that imports webhelp.xsl, use
# this property to point to it.
stylesheet-path=${ant.file.dir}/xsl/webhelp.xsl
# If your document has image directories that need to be copied
# to the output directory, you can list patterns here.
# See the Ant documentation for fileset for documentation
# on patterns.
#input-images-dirs=images/**,figures/**,graphics/**
# By default, the ant script assumes your images are stored
# in the same directory as the input-xml. If you store your
# image directories in another directory, specify it here.
# and uncomment this line.
#input-images-basedir=/path/to/image/location
# Modify this so that it points to your copy of the Saxon 6.5 jar.
xslt-processor-classpath=/usr/share/java/saxon-6.5.5.jar
# For non-ns version only, this validates the document
# against a dtd.
validate-against-dtd=true
# Set this to false if you don't need a search tab.
webhelp.include.search.tab=true
# indexer-language is used to tell the search indexer which language
# the docbook is written. This will be used to identify the correct
# stemmer, and punctuations that differs from language to language.
# see the documentation for details. en=English, fr=French, de=German,
# zh=Chinese, ja=Japanese etc.
webhelp.indexer.language=en</programlisting></para>
</step>
<step>
<para>Test the package by running the command <code>ant webhelp
-Doutput-dir=test-ouput</code> at the command line in the webhelp directory. It should
generate a copy of this documentation in the <filename class="directory">doc</filename>
directory. Type <code>start test-output\index.html</code> to open the output in a
browser. Once you have confirmed that the process worked, you can delete the <filename
class="directory">test-output</filename> directory. <important>
<para>The Saxon 6.5 jar should <emphasis>not</emphasis> be in your
<envar>CLASSPATH</envar> when you generate the webhelp output. If you have any
problems, try running ant with an empty <envar>CLASSPATH</envar>.</para>
</important></para>
</step>
<step>
<para>To process your own document, simply refer to this package
from another <filename>build.xml</filename> in arbitrary location on
your system:</para>
<substeps>
<step>
<para>Create a new <filename>build.xml</filename> file that
defines the name of your source file, the desired output
directory, and imports the <filename>build.xml</filename> from
this package. For example: <programlisting><project>
<property name="input-xml" value="<replaceable>path-to/yourfile.xml</replaceable>"/>
<property name="input-images-dirs" value="<replaceable>images/** figures/** graphics/**</replaceable>"/>
<property name="output-dir" value="<replaceable>path-to/desired-output-dir</replaceable>"/>
<import file="<replaceable>path-to/docbook-webhelp/</replaceable>build.xml"/>
</project></programlisting></para>
</step>
<step>
<para>From the directory containing your newly created
<filename>build.xml</filename> file, type <code>ant
webhelp</code> to build your document.</para>
<important>
<para>The Saxon 6.5 jar should <emphasis>not</emphasis> be in your
<envar>CLASSPATH</envar> when you generate the webhelp output. If you have any
problems, try running ant with an empty <envar>CLASSPATH</envar>.</para>
</important>
</step>
</substeps>
</step>
</procedure>
</section>
<section>
<title>Using and customizing the output</title>
<para>To deep link to a topic inside the help set, simply link directly
to the page. This help system uses no frameset, so nothing further is
necessary. <tip>
<para>See <ulink
url="http://www.sagehill.net/docbookxsl/Chunking.html">Chunking into
multiple HTML files</ulink> in Bob Stayton's <ulink
url="http://www.sagehill.net/docbookxsl/index.html">DocBook XSL: The
Complete Guide</ulink> for information on controlling output file
names and which files are chunked in DocBook.</para>
</tip></para>
<para>When you perform a search, the results can include brief
summaries. These are populated in one of two ways:<itemizedlist>
<listitem>
<para>By adding <sgmltag>role="summary"</sgmltag> to a
<sgmltag>para</sgmltag> or <sgmltag>phrase</sgmltag> in the
<sgmltag>chapter</sgmltag> or <sgmltag>section</sgmltag>.</para>
</listitem>
<listitem>
<para>By adding an <sgmltag>abstract</sgmltag> to the
<sgmltag>chapterinfo</sgmltag> or <sgmltag>sectioninfo</sgmltag>
element.</para>
</listitem>
</itemizedlist></para>
<para>To customize the look and feel of the help, study the following
css files:<itemizedlist>
<listitem>
<para><filename>docs/common/css/positioning.css</filename>: This
handles the Positioning of DIVs in appropriate positions. For
example, it causes the <code>leftnavigation</code> div to appear
on the left, the header on top, and so on. Use this if you need to
change the relative positions or need to change the width/height
etc.</para>
</listitem>
<listitem>
<para><filename>docs/common/jquery/theme-redmond/jquery-ui-1.8.2.custom.css</filename>:
This is the theming part which adds colors and stuff. This is a
default theme comes with <ulink
url="http://jqueryui.com/download">jqueryui</ulink> unchanged. You
can get any theme based your interest from this. (Themes are on
right navigation bar.) Then replace the css theme folder
(theme-redmond) with it, and change the xsl to point to the new
css.</para>
</listitem>
<listitem>
<para><filename>docs/common/jquery/treeview/jquery.treeview.css</filename>:
This styles the toc Tree. Generally, you don't have to edit this
file.</para>
</listitem>
</itemizedlist></para>
<section>
<title>Recommended Apache configurations</title>
<para>If you are serving a long document from an Apache web server, we
recommend you make the following additions or changes to your
<filename>httpd.conf</filename> or <filename>.htaccess</filename>
file. <remark>TODO: Explain what each thing
does.</remark><programlisting>AddDefaultCharSet UTF-8 # <co
id="AddDefaultCharSet" />
# 480 weeks
<FilesMatch "\.(ico|pdf|flv|jpg|jpeg|png|gif|js|css|swf)$"> # <co
id="CachingSettings" />
Header set Cache-Control "max-age=290304000, public"
</FilesMatch>
# 2 DAYS
<FilesMatch "\.(xml|txt)$">
Header set Cache-Control "max-age=172800, public, must-revalidate"
</FilesMatch>
# 2 HOURS
<FilesMatch "\.(html|htm)$">
Header set Cache-Control "max-age=7200, must-revalidate"
</FilesMatch>
# compress text, html, javascript, css, xml:
AddOutputFilterByType DEFLATE text/plain # <co id="CompressSetting" />
AddOutputFilterByType DEFLATE text/html
AddOutputFilterByType DEFLATE text/xml
AddOutputFilterByType DEFLATE text/css
AddOutputFilterByType DEFLATE application/xml
AddOutputFilterByType DEFLATE application/xhtml+xml
AddOutputFilterByType DEFLATE application/rss+xml
AddOutputFilterByType DEFLATE application/javascript
AddOutputFilterByType DEFLATE application/x-javascript
# Or, compress certain file types by extension:
<Files *.html>
SetOutputFilter DEFLATE
</Files>
</programlisting><calloutlist>
<callout arearefs="AddDefaultCharSet">
<para>See <ulink
url="http://www.sagehill.net/docbookxsl/SpecialChars.html">Odd
characters in HTML output</ulink> in Bob Stayton's book
<citetitle>DocBook XSL: The Complete Guide</citetitle> for more
information about this setting.</para>
</callout>
<callout arearefs="CachingSettings">
<para>These lines and those that follow cause the browser to
cache various resources such as bitmaps and JavaScript files.
Note that caching JavaScript files could cause your users to
have stale search indexes if you update your document since the
search index is stored in JavaScript files.</para>
</callout>
<callout arearefs="CompressSetting">
<para>These lines cause the the server to compress html, css,
and JavaScript files and the brower to uncompress them to
improve download performance.</para>
</callout>
</calloutlist></para>
</section>
</section>
<section>
<title>Building the indexer</title>
<para role="summary">To build the indexer, you must have installed the
JDK version 1.5 or higher and set the <envar>ANT_HOME</envar>
environment variable. Run <code>ant build-indexer</code> to recompile
<filename>nw-cms.jar</filename></para>
<indexterm>
<primary>ANT_HOME</primary>
</indexterm>
<indexterm>
<primary>indexer</primary>
<secondary>building</secondary>
</indexterm>
</section>
<section>
<title>Adding support for other (non-CJKV) languages</title>
<para>To support stemming for a language, the search mechanism requires
a stemmer implemented in both Java and JavaScript. The Java version is
used by the indexer and the JavaScript verison is used to stem the
user's input on the search form. Currently the search mechanism supports
stemming for English and German. In addition, Java stemmers are included
for the following languages. Therefore, to support these languages, you
only need to implement the stemmer in JavaScript and add it to the
template. If you do undertake this task, please consider contributing
the JavaScript version back to this project and to <ulink
url="http://snowball.tartarus.org/texts/stemmersoverview.html">Martin
Porter's project</ulink>.<itemizedlist>
<listitem>
<para>Danish</para>
</listitem>
<listitem>
<para>Dutch</para>
</listitem>
<listitem>
<para>Finnish</para>
</listitem>
<listitem>
<para>Hungarian</para>
</listitem>
<listitem>
<para>Italian</para>
</listitem>
<listitem>
<para>Norwegian</para>
</listitem>
<listitem>
<para>Portuguese</para>
</listitem>
<listitem>
<para>Romanian</para>
</listitem>
<listitem>
<para>Russian</para>
</listitem>
<listitem>
<para>Spanish</para>
</listitem>
<listitem>
<para>Swedish</para>
</listitem>
<listitem>
<para>Turkish</para>
</listitem>
</itemizedlist></para>
</section>
</chapter>
<chapter>
<title>Developer Docs</title>
<para role="summary">This chapter provides an overview of how webhelp is implemented.</para>
<para>The table of contents and search panes are implemented as divs and
rendered as if they were the left pane in a frameset. As a result, the
page must save the state of the table of contents and the search in
cookies when you navigate away from a page. When you load a new page, the
page reads these cookies and restores the state of the table of contents
tree and search. The result is that the help system behaves exactly as if
it were a frameset.</para>
<section>
<title>Design</title>
<para role="summary">An overview of webhelp page structure.</para>
<para>DocBook WebHelp page structure is fully built on css-based design
abandoning frameset structure. Overall page structure can be divided in to three main sections
<itemizedlist>
<listitem>
<para>Header: Header is a separate Div which include company logo,
navigation button(prev, next etc.), page title and heading of parent topic.</para>
</listitem>
<listitem>
<para>Content: This includes the content of the documentation. The processing of this part is
done by <ulink
url="http://docbook.sourceforge.net/release/xsl/current/xhtml/chunk.xsl">
DocBook XSL Chunking customization</ulink>. Few further css-styling applied from
<filename>positioning.css</filename>.
</para>
</listitem>
<listitem>
<para>Left Navigation: This includes the table of contents and search tab. This
is customized using <ulink url="http://jqueryui.com/">jquery-ui</ulink> styling.</para>
<itemizedlist>
<listitem>
<para>Tabbed Navigation: The navigation pane is organized in to two tabs.
Contents tab, and Search tab. Tabbed output is achieved using
<ulink url="http://docs.jquery.com/UI/Tabs">JQuery Tabs plugin</ulink>.
</para>
</listitem>
<listitem>
<para>Table of Contents (TOC) tree: When building the chunked html from the
docbook file, Table of Contents is generated as an Unordered List (a list
made from <code><ul> <li></code> tags). When page loads in the browser,
we apply styling to it to achieve the nice look that you see. Styling for TOC
tree is done by a JQuery UI plugin called
<ulink url="http://bassistance.de/jquery-plugins/jquery-plugin-treeview/">
TreeView</ulink>. We can generate the tree easily by following javascript code:
<programlisting>
//Generate the tree
$("#tree").treeview({
collapsed: true,
animated: "medium",
control: "#sidetreecontrol",
persist: "cookie"
});
</programlisting>
</para>
</listitem>
<listitem>
<para>Search Tab: This includes the search feature.</para>
</listitem>
</itemizedlist>
</listitem>
</itemizedlist>
</para>
</section>
<section>
<title>Search</title>
<para role="summary">Overview design of Search mechanism.</para>
<para>
The searching is a fully client-side implementation of querying texts for
content searching, and no server is involved. That means when a user enters a query,
it is processed by JavaScript inside the browser, and displays the matching results by
comparing the query with a generated 'index', which too reside in the client-side web browser.
Mainly the search mechanism has two parts.
<itemizedlist>
<listitem>
<para>Indexing: First we need to traverse the content in the docs/content folder and index
the words in it. This is done by <filename>nw-cms.jar</filename>. You can invoke it by
<code>ant index</code> command from the root of webhelp of directory. You can recompile it
again and build the jar file by <code>ant build-indexer</code>. Indexer has some extensive
support for such as stemming of words. Indexer has extensive support for English, German,
French languages. By extensive support, what I meant is that those texts are stemmed
first, to get the root word and then indexes them. For CJK (Chinese, Japanese, Korean)
languages, it uses bi-gram tokenizing to break up the words. (CJK languages does not have
spaces between words.)
</para>
<para>
When we run <code>ant index</code>, it generates five output files:
<itemizedlist>
<listitem>
<para><filename>htmlFileList.js</filename> - This contains an array named <code>fl</code> which stores details
all the files indexed by the indexer.
</para>
</listitem>
<listitem>
<para><filename>htmlFileInfoList.js</filename> - This includes some meta data about the indexed files in an array
named <code>fil</code>. It includes details about file name, file (html) title, a summary
of the content.Format would look like,
<code>fil["4"]= "ch03.html@@@Developer Docs@@@This chapter provides an overview of how webhelp is implemented.";</code>
</para>
</listitem>
<listitem>
<para><filename>index-*.js</filename> (Three index files) - These three files actually stores the index of the content.
Index is added to an array named <code>w</code>.</para>
</listitem>
</itemizedlist>
</para>
</listitem>
<listitem>
<para>
Querying: Query processing happens totally in client side. Following JavaScript files handles them.
<itemizedlist>
<listitem>
<para><filename>nwSearchFnt.js</filename> - This handles the user query and returns the search results. It does query
word tokenizing, drop unnecessary punctuations and common words, do stemming if docbook language
supports it, etc.</para>
</listitem>
<listitem>
<para><filename>{$indexer-language-code}_stemmer.js</filename> - This includes the stemming library.
<filename>nwSearchFnt.js</filename> file calls <code>stemmer</code> method in this file for stemming.
ex: <code>var stem = stemmer(foobar);</code>
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
</itemizedlist>
</para>
<section>
<title>New Stemmers</title>
<para role="summary">Adding new Stemmers is very simple.</para>
<para>Currently, only English, French, and German stemmers are integrated in to WebHelp. But the code is
extensible such that you can add new stemmers easily by few steps.</para>
<para>What you need:
<itemizedlist>
<listitem>
<para>You'll need two versions of the stemmer; One written in JavaScript, and another in Java. But fortunately,
Snowball contains Java stemmers for number of popular languages, and are already included with the package.
You can see the full list in <ulink url="ch02s04.html">Adding support for other (non-CJKV) languages</ulink>.
If your language is listed there,
Then you have to find javascript version of the stemmer. Generally, new stemmers are getting added in to
<ulink url="http://snowball.tartarus.org/otherlangs/index.html">Snowball Stemmers in other languages</ulink> location.
If javascript stemmer for your language is available, then download it. Else, you can write a new stemmer in
JavaScript using SnowBall algorithm fairly easily. Algorithms are at
<ulink url="http://snowball.tartarus.org/">Snowball</ulink>.
</para>
</listitem>
<listitem>
<para>Then, name the JS stemmer exactly like this: <filename>{$language-code}_stemmer.js</filename>. For example,
for Italian(it), name it as, <filename>it_stemmer.js</filename>. Then, copy it to the
<filename>docbook-webhelp/template/content/search/stemmers/</filename> folder. (I assumed
<filename>docbook-webhelp</filename> is the root folder for webhelp.)
<note>
<para>Make sure you changed the <code>webhelp.indexer.language</code> property in <filename>build.properties</filename>
to your language.
</para>
</note>
</para>
</listitem>
<listitem>
<para>Now two easy changes needed for the indexer.</para>
<itemizedlist>
<listitem>
<para>Open <filename>docbook-webhelp/indexer/src/com/nexwave/nquindexer/IndexerTask.java</filename> in
a text editor and add your language code to the <code>supportedLanguages</code> String Array. </para>
<example>
<title>Add new language to supportedLanguages array</title>
<para>
change the Array from,
<programlisting>
private String[] supportedLanguages= {"en", "de", "fr", "cn", "ja", "ko"};
//currently extended support available for
// English, German, French and CJK (Chinese, Japanese, Korean) languages only.
</programlisting>
To,</para>
<programlisting>
private String[] supportedLanguages= {"en", "de", "fr", "cn", "ja", "ko", <emphasis>"it"</emphasis>};
//currently extended support available for
// English, German, French, CJK (Chinese, Japanese, Korean), and Italian languages only.
</programlisting>
</example>
</listitem>
<listitem>
<para>
Now, open <filename>docbook-webhelp/indexer/src/com/nexwave/nquindexer/SaxHTMLIndex.java</filename> and
add the following line to the code where it initializes the Stemmer (Search for
<code>SnowballStemmer stemmer;</code>). Then add code to initialize the stemmer Object in your language.
It's self understandable. See the example. The class names are at:
<filename>docbook-webhelp/indexer/src/com/nexwave/stemmer/snowball/ext/</filename>.
</para>
<example>
<title>initialize correct stemmer based on the <code>webhelp.indexer.language</code> specified</title>
<programlisting>
SnowballStemmer stemmer;
if(indexerLanguage.equalsIgnoreCase("en")){
stemmer = new EnglishStemmer();
} else if (indexerLanguage.equalsIgnoreCase("de")){
stemmer= new GermanStemmer();
} else if (indexerLanguage.equalsIgnoreCase("fr")){
stemmer= new FrenchStemmer();
}
<emphasis>else if (indexerLanguage.equalsIgnoreCase("it")){ //If language code is "it" (Italian)
stemmer= new italianStemmer(); //Initialize the stemmer to <code>italianStemmer</code> object.
} </emphasis>
else {
stemmer = null;
}
</programlisting>
</example>
</listitem>
</itemizedlist>
</listitem>
</itemizedlist>
</para>
<para>That's all. Now run <code>ant build-indexer</code> to compile and build the java code.
Then, run <code>ant webhelp</code> to generate the output from your docbook file.
For any questions, contact us or email to the docbook mailing list
<email>docbook-apps@lists.oasis-open.org</email>.
</para>
</section>
</section>
</chapter>
</book>
