<refentry xmlns="http://docbook.org/ns/docbook"
          xmlns:xlink="http://www.w3.org/1999/xlink"
          xmlns:xi="http://www.w3.org/2001/XInclude"
          xmlns:src="http://nwalsh.com/xmlns/litprog/fragment"
          xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
          version="5.0" xml:id="index.method">
<refmeta>
<refentrytitle>index.method</refentrytitle>
<refmiscinfo class="other" otherclass="datatype">list</refmiscinfo>
<refmiscinfo class="other" otherclass="value">basic</refmiscinfo>
<refmiscinfo class="other" otherclass="value">kosek</refmiscinfo>
<refmiscinfo class="other" otherclass="value">kimber</refmiscinfo>
</refmeta>
<refnamediv>
<refname>index.method</refname>
<refpurpose>Select method used to group index entries in an index</refpurpose>
</refnamediv>

<refsynopsisdiv>
<src:fragment xml:id="index.method.frag">
<xsl:param name="index.method">basic</xsl:param>
</src:fragment>
</refsynopsisdiv>

<refsection><info><title>Description</title></info>

<para>This parameter lets you select which method to use for sorting and grouping
 index entries in an index.
Indexes in Latin-based languages that have accented characters typically
sort together accented words and unaccented words.
Thus <quote>Á</quote> (U+00C1 LATIN CAPITAL LETTER A WITH ACUTE) would sort together
with <quote>A</quote> (U+0041 LATIN CAPITAL LETTER A), so both would appear in the <quote>A</quote>
section of the index.
Languages using other alphabets (such as Russian,  which is written in the Cyrillic alphabet)
and languages using ideographic chararacters (such as Japanese)
require grouping specific to the languages and alphabets.
</para>

<para>The default indexing method is limited.
It can group accented characters in Latin-based languages only.
It cannot handle non-Latin alphabets or ideographic languages.
The other indexing methods require extensions of one type or
another, and do not work with
all XSLT processors, which is why they are not used by default.</para>

<para>The three choices for indexing method are:</para>
<variablelist>
<varlistentry>
<term><literal>basic</literal></term>
<listitem>
<para>
(default)  Sort and groups words based only on the Latin alphabet.
Words with accented Latin letters will group and sort with
their respective primary letter, but
words in non-Latin alphabets will be
put in the <quote>Symbols</quote> section of the index.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><literal>kosek</literal></term>
<listitem>
<para>
This method sorts and groups words based on letter groups configured in
the DocBook locale file for the given language.
See, for example, the French locale file <filename>common/fr.xml</filename>.
This method requires that the XSLT processor
supports the EXSLT extensions (most do).
It also requires support for using 
user-defined functions in xsl:key (xsltproc does not).
</para>
<para>This method is suitable for any language for which you can
list all the individual characters that should appear
in each letter group in an index.
It is probably not practical to use it for ideographic languages
such as Chinese that have hundreds or thousands of characters.
</para>

<para>To use the kosek method, you must:</para>

<orderedlist>
<listitem>
<para>Use a processor that supports its extensions, such as
Saxon 6 or Xalan (xsltproc and Saxon 8 do not).
</para>
</listitem>
<listitem>
<para>Set the index.method parameter's value to <quote>kosek</quote>.
</para>
</listitem>
<listitem>
<para>Import the appropriate index extensions stylesheet module
<filename>fo/autoidx-kosek.xsl</filename> or 
<filename>html/autoidx-kosek.xsl</filename> into your
customization.
</para>
</listitem>
</orderedlist>

</listitem>
</varlistentry>
<varlistentry>
<term><literal>kimber</literal></term>
<listitem>
<para>
This method uses extensions to the Saxon processor to implement
sophisticated indexing processes. It uses its own 
configuration file, which can include information for any number of
languages. Each language's configuration can group
words using one of two processes. In the
enumerated process similar to that used in the kosek method,
you indicate the groupings character-by-character.
In the between-key process, you specify the
break-points in the sort order that should start a new group.
The latter configuration is useful for ideographic languages
such as Chinese, Japanese, and Korean.
You can also define your own collation algorithms and how you
want mixed Latin-alphabet words sorted.</para>
<itemizedlist>
<listitem>
<para>For a whitepaper describing the extensions, see:
<link xlink:href="http://www.innodata-isogen.com/knowledge_center/white_papers/back_of_book_for_xsl_fo.pdf">http://www.innodata-isogen.com/knowledge_center/white_papers/back_of_book_for_xsl_fo.pdf</link>.
</para>
</listitem>
<listitem>
<para>To download the extension library, see
<link xlink:href="http://www.innodata-isogen.com/knowledge_center/tools_downloads/i18nsupport">http://www.innodata-isogen.com/knowledge_center/tools_downloads/i18nsupport</link>.
</para>
</listitem>
</itemizedlist>

<para>To use the kimber method, you must:</para>

<orderedlist>
<listitem>
<para>Use Saxon (version 6 or 8) as your XSLT processor.
</para>
</listitem>
<listitem>
<para>Install and configure the Innodata Isogen library, using
the documentation that comes with it.
</para>
</listitem>
<listitem>
<para>Set the index.method parameter's value to <quote>kimber</quote>.
</para>
</listitem>
<listitem>
<para>Import the appropriate index extensions stylesheet module
<filename>fo/autoidx-kimber.xsl</filename> or 
<filename>html/autoidx-kimber.xsl</filename> into your
customization.
</para>
</listitem>
</orderedlist>

</listitem>
</varlistentry>
</variablelist>

</refsection>
</refentry>