default is Latin before Greek.)
        </para>
       </listitem>
      </varlistentry>

      <varlistentry id="collation-managing-create-icu-en-u-kf-upper">
       <term><literal>CREATE COLLATION upperfirst (provider = icu, locale = 'en-u-kf-upper');</literal></term>
       <listitem>
        <para>
         Sort upper-case letters before lower-case letters.  (The default is
         lower-case letters first.)
        </para>
       </listitem>
      </varlistentry>

      <varlistentry id="collation-managing-create-icu-en-u-kf-upper-kr-grek-latn">
       <term><literal>CREATE COLLATION special (provider = icu, locale = 'en-u-kf-upper-kr-grek-latn');</literal></term>
       <listitem>
        <para>
         Combines both of the above options.
        </para>
       </listitem>
      </varlistentry>
     </variablelist>
   </sect3>

   <sect3 id="icu-tailoring-rules">
    <title>ICU Tailoring Rules</title>

    <para>
     If the options provided by the collation settings shown above are not
     sufficient, the order of collation elements can be changed with tailoring
     rules, whose syntax is detailed at <ulink
     url="https://unicode-org.github.io/icu/userguide/collation/customization/"></ulink>.
    </para>

    <para>
     This small example creates a collation based on the root locale with a
     tailoring rule:
<programlisting>
<![CDATA[CREATE COLLATION custom (provider = icu, locale = 'und', rules = '&V << w <<< W');]]>
</programlisting>
     With this rule, the letter <quote>W</quote> is sorted after
     <quote>V</quote>, but is treated as a secondary difference similar to an
     accent.  Rules like this are contained in the locale definitions of some
     languages.  (Of course, if a locale definition already contains the
     desired rules, then they don't need to be specified again explicitly.)
    </para>

    <para>
     Here is a more complex example.  The following statement sets up a
     collation named <literal>ebcdic</literal> with rules to sort US-ASCII
     characters in the order of the EBCDIC encoding.

<programlisting>
<![CDATA[CREATE COLLATION ebcdic (provider = icu, locale = 'und',
rules = $$
& ' ' < '.' < '<' < '(' < '+' < \|
< '&' < '!' < '$' < '*' < ')' < ';'
< '-' < '/' < ',' < '%' < '_' < '>' < '?'
< '`' < ':' < '#' < '@' < \' < '=' < '"'
<*a-r < '~' <*s-z < '^' < '[' < ']'
< '{' <*A-I < '}' <*J-R < '\' <*S-Z <*0-9
$$);]]>

SELECT c
FROM (VALUES ('a'), ('b'), ('A'), ('B'), ('1'), ('2'), ('!'), ('^')) AS x(c)
ORDER BY c COLLATE ebcdic;
 c
---
 !
 a
 b
 ^
 A
 B
 1
 2
</programlisting>
    </para>
   </sect3>

   <sect3 id="icu-external-references">
    <title>External References for ICU</title>

    <para>
     This section (<xref linkend="icu-custom-collations"/>) is only a brief
     overview of ICU behavior and language tags. Refer to the following
     documents for technical details, additional options, and new behavior:
    </para>

    <itemizedlist>
     <listitem>
      <para>
       <ulink url="https://www.unicode.org/reports/tr35/tr35-collation.html">Unicode Technical Standard #35</ulink>
      </para>
     </listitem>
     <listitem>
      <para>
       <ulink url="https://www.rfc-editor.org/info/bcp47">BCP 47</ulink>
      </para>
     </listitem>
     <listitem>
      <para>
       <ulink url="https://github.com/unicode-org/cldr/blob/master/common/bcp47/collation.xml">CLDR repository</ulink>
      </para>
     </listitem>
     <listitem>
      <para>
       <ulink url="https://unicode-org.github.io/icu/userguide/locale/"></ulink>
      </para>
     </listitem>
     <listitem>
      <para>
       <ulink url="https://unicode-org.github.io/icu/userguide/collation/"></ulink>
      </para>
     </listitem>
    </itemizedlist>
   </sect3>
  </sect2>
 </sect1>

 <sect1 id="multibyte">
  <title>Character Set Support</title>

  <indexterm zone="multibyte"><primary>character set</primary></indexterm>

  <para>
   The character set support in <productname>PostgreSQL</productname>
   allows