<!-- jEdit buffer-local properties: -->
<!-- :indentSize=1:noTabs=true: -->
<!-- :xml.root=users-guide.xml: -->

<chapter id="updating-modes"><title>Updating Edit Modes for jEdit 4.1/4.2</title>
 <para>
  In jEdit 4.1, the mode file grammar has been cleaned up somewhat. As a result, some
  edit modes written for jEdit 4.0 and earlier need to be updated:
 </para>
 <itemizedlist>
  <listitem>
   <para>
    Defining <literal>&lt;WHITESPACE&gt;</literal> rules is no longer necessary
    and doing so will print warnings to the activity logs.
   </para>
  </listitem>
  <listitem>
   <para>
    The <literal>&lt;KEYWORDS&gt;</literal> tag no longer accepts an
    <literal>IGNORE_CASE</literal> attribute.
    Set the <literal>IGNORE_CASE</literal> attribute of the
    <literal>&lt;RULES&gt;</literal> tag instead.
   </para>
  </listitem>
  <listitem>
   <para>
    The <literal>&lt;END&gt;</literal> tag of the <literal>&lt;SPAN&gt;</literal> rule
    used to be optional, in which case
    any occurrence of the start string would cause the remainder of the buffer
    to be highlighted with the span. In jEdit 4.1, the <literal>&lt;END&gt;</literal>
    tag can no longer be omitted, however
    a <literal>&lt;SEQ&gt;</literal> tag with a <literal>DELEGATE</literal> attribute
    can be used to achieve the same effect as endless span.
   </para>
  </listitem>
  <listitem>
   <para>
    Defining <literal>&lt;SEQ TYPE="NULL"&gt;</literal> rules for word separators
    is no longer necessary. Now, any non-alphanumeric character not appearing in
    a keyword definition or the ruleset's <literal>NO_WORD_SEP</literal> attribute
    is considered a word separator.
   </para>
  </listitem>
 </itemizedlist>
</chapter>