<usage>
<p>The <directive>IndexIgnore</directive> directive adds to the
list of files to hide when listing a directory. <var>File</var> is a
- shell-style wildcard expression or full
- filename. Multiple IndexIgnore directives add
- to the list, rather than replacing the list of ignored
- files. By default, the list contains <code>.</code> (the current
- directory).</p>
+ filename or pattern which may contain the <code>?</code> and
+ <code>*</code> wildcard characters. <code>?</code> matches any
+ single character, and <code>*</code> matches any sequence of
+ characters, including an empty sequence. Multiple IndexIgnore
+ directives add to the list, rather than replacing the list of
+ ignored files. By default, the list contains <code>.</code> (the
+ current directory).</p>
<highlight language="config">
IndexIgnore .??* *~ *# HEADER* README* RCS CVS *,v *,t
</highlight>
+ <p>If <var>File</var> contains a <code>/</code> character, only
+ the portion after the last <code>/</code> is used for matching
+ against filenames. For example, <code>*/.??*</code> is equivalent
+ to <code>.??*</code>.</p>
+
+ <note><title>Differences from shell wildcard patterns</title>
+ <p>The wildcard matching used here is <em>not</em> the same as
+ shell-style (glob) filename pattern matching. In particular:</p>
+ <ul>
+ <li>Bracket expressions such as <code>[abc]</code> or
+ <code>[!.]</code> are not supported; they are treated as
+ literal characters.</li>
+ <li>A leading <code>.</code> in a filename is not treated
+ specially. For example, <code>*~</code> will match
+ <code>.example~</code>, whereas a shell glob would not.</li>
+ </ul>
+ </note>
+
<note><title>Regular Expressions</title>
<p>This directive does not currently work in configuration sections
that have regular expression arguments, such as <directive