ch05_07.htm

编程珍珠,里面很多好用的代码,大家可以参考学习呵呵,

print "postmatch: $'\n";          # , and enter.
print "lastmatch: $+\n";          # &lt;/EM&gt;</pre>
</blockquote>

For more explanation of these magical Elvish variables (and for a way
to write them in English), see
<a href="ch28_01.htm">Chapter 28, "Special Names"</a>.</p>

<p>
<a name="INDEX-1654"></a><a name="INDEX-1655"></a>
The <tt class="literal">@-</tt> (<tt class="literal">@LAST_MATCH_START</tt>) array holds the offsets of the
beginnings of any submatches, and <tt class="literal">@+</tt> (<tt class="literal">@LAST_MATCH_END</tt>) holds the
offsets of the ends:
<blockquote>
<pre class="programlisting">#!/usr/bin/perl
$alphabet = "abcdefghijklmnopqrstuvwxyz";
$alphabet =~ /(hi).*(stu)/;

print "The entire match began at $-[0] and ended at $+[0]\n";
print "The first  match began at $-[1] and ended at $+[1]\n";
print "The second match began at $-[2] and ended at $+[2]\n";</pre>
</blockquote>
</p>

<p>If you really want to match a literal parenthesis character instead of having
it interpreted as a metacharacter, backslash it:
<blockquote>
<pre class="programlisting">/\(e.g., .*?\)/</pre>
</blockquote>

This matches a parenthesized example (e.g., this statement).  But since dot
is a wildcard, this also matches any parenthetical statement with the
first letter <tt class="literal">e</tt> and third letter <tt class="literal">g</tt> (ergo, this statement too).</p>

<a name="INDEX-1656"></a><a name="INDEX-1657"></a>






<h3 class="sect2">5.7.2. Clustering</h3>

<a name="INDEX-1658"></a><a name="INDEX-1659"></a>
<p>
<a name="INDEX-1660"></a><a name="INDEX-1661"></a>
Bare parentheses both cluster and capture.  But sometimes you don't
want that.  Sometimes you just want to group portions of the pattern
without creating a backreference.  You can use an extended form of
parentheses to suppress capturing: the <tt class="literal">(?:</tt><em class="replaceable">PATTERN</em><tt class="literal">)</tt> notation
will <em class="emphasis">cluster</em> without capturing.</p>

<p>
<a name="INDEX-1662"></a>
There are at least three reasons you might want to cluster without
capturing:</p>

<ol>
<li>
<p>To quantify something.</p>
</li>
<li>
<p>
<a name="INDEX-1663"></a>
To limit the scope of interior alternation; for example,
<tt class="literal">/^cat|cow|dog$/</tt> needs to be <tt class="literal">/^(?:cat|cow|dog)$/</tt> so that the
cat doesn't run away with the <tt class="literal">^</tt>.</p>
</li>
<li>
<p>To limit the scope of an embedded pattern modifier to a particular
subpattern, such as in <tt class="literal">/foo(?-i:Case_Matters)bar/i</tt>.
(See the next section, <a href="ch05_07.htm#ch05-sect-cloist">Section 5.7.3, "Cloistered Pattern Modifiers"</a>)</p>
</li>
</ol>
<p>In addition, it's more efficient to suppress the capture of something
you're not going to use.  On the minus side, the notation is a little
noisier, visually speaking.</p>

<p>
<a name="INDEX-1664"></a>
In a pattern,
a left parenthesis immediately followed by a question mark denotes a
regex <em class="emphasis">extension</em>.  The current regular expression
bestiary is relatively fixed--we don't dare create a new
metacharacter, for fear of breaking old Perl programs.  Instead, the
extension syntax is used to add new features to the bestiary.</p>

<p>In the remainder of the chapter, we'll see many
more regex extensions, all of which cluster without capturing, as well
as doing something else.  The
<tt class="literal">(?:</tt><em class="replaceable">PATTERN</em><tt class="literal">)</tt>
extension is just special in that it does nothing else.  So if you
say:
<blockquote>
<pre class="programlisting">@fields = split(/\b(?:a|b|c)\b/)</pre>
</blockquote>
it's like:
<blockquote>
<pre class="programlisting">@fields = split(/\b(a|b|c)\b/)</pre>
</blockquote>
<a name="INDEX-1665"></a>
but doesn't spit out extra fields.  (The <tt class="literal">split</tt> operator is a bit
like <tt class="literal">m//g</tt> in that it will emit extra fields for all the captured
substrings within the pattern.  Ordinarily, <tt class="literal">split</tt> only returns what it
<em class="emphasis">didn't</em> match.  For more on <tt class="literal">split</tt> see <a href="ch29_01.htm">Chapter 29, "Functions"</a>.)</p>





<a name="ch05-sect-cloist"></a>
<h3 class="sect2">5.7.3. Cloistered Pattern Modifiers</h3>

<p>
<a name="INDEX-1666"></a><a name="INDEX-1667"></a><a name="INDEX-1668"></a>
You may <em class="emphasis">cloister</em> the <tt class="literal">/i</tt>, <tt class="literal">/m</tt>, <tt class="literal">/s</tt>, and <tt class="literal">/x</tt> modifiers
within a portion of your pattern by inserting them (without the
slash) between the <tt class="literal">?</tt> and <tt class="literal">:</tt> of the clustering notation.  If you say:
<blockquote>
<pre class="programlisting">/Harry (?i:s) Truman/</pre>
</blockquote>

it matches both "<tt class="literal">Harry S Truman</tt>" and
"<tt class="literal">Harry s Truman</tt>", whereas:
<blockquote>
<pre class="programlisting">/Harry (?x: [A-Z] \.? \s )?Truman/</pre>
</blockquote>

matches both "<tt class="literal">Harry S Truman</tt>" and "<tt class="literal">Harry
S. Truman</tt>", as well as "<tt class="literal">Harry Truman</tt>", and:
<blockquote>
<pre class="programlisting">/Harry (?ix: [A-Z] \.? \s )?Truman/</pre>
</blockquote>

matches all five, by combining the <tt class="literal">/i</tt> and
<tt class="literal">/x</tt> modifiers within the cloister.</p>

<p>
<a name="INDEX-1669"></a>
You can also subtract modifiers from a cloister with a minus sign:
<blockquote>
<pre class="programlisting">/Harry (?x-i: [A-Z] \.? \s )?Truman/i</pre>
</blockquote>

This matches any capitalization of the name--but if the middle initial
is provided, it must be capitalized, since the <tt class="literal">/i</tt>
applied to the overall pattern is suspended inside the cloister.</p>

<p>By omitting the colon and <em class="replaceable">PATTERN</em>, you can export modifier settings
to an outer cluster, turning it into a cloister.  That is, you can
selectively turn modifiers on and off for the cluster one level outside
the modifiers' parentheses, like so:
<blockquote>
<pre class="programlisting">/(?i)foo/            # Equivalent to /foo/i
/foo((?-i)bar)/i     # "bar" must be lower case
/foo((?x-i) bar)/    # Enables /x and disables /i for "bar"</pre>
</blockquote>

Note that the second and third examples create backreferences.
If that wasn't what you wanted, then you should have been using
<tt class="literal">(?-i:bar)</tt> and <tt class="literal">(?x-i: bar)</tt>, respectively.</p>

<p>Setting modifiers on a portion of your pattern is particularly useful
when you want "<tt class="literal">.</tt>" to match newlines in part of your pattern but
not in the rest of it.  Setting <tt class="literal">/s</tt> on the whole pattern doesn't help
you there.</p>



<!-- BOTTOM NAV BAR -->

<hr width="515" align="left">
<div class="navbar">
<table width="515" border="0">
<tr>
<td align="left" valign="top" width="172"><a href="ch05_06.htm"><img src="../gifs/txtpreva.gif" alt="Previous" border="0"></a></td><td align="center" valign="top" width="171"><a href="index.htm"><img src="../gifs/txthome.gif" alt="Home" border="0"></a></td><td align="right" valign="top" width="172"><a href="ch05_08.htm"><img src="../gifs/txtnexta.gif" alt="Next" border="0"></a></td>
</tr>
<tr>
<td align="left" valign="top" width="172">5.6. Positions</td><td align="center" valign="top" width="171"><a href="index/index.htm"><img src="../gifs/index.gif" alt="Book Index" border="0"></a></td><td align="right" valign="top" width="172">5.8. Alternation</td>
</tr>
</table>
</div>
<hr width="515" align="left">

<!-- LIBRARY NAV BAR -->

<img src="../gifs/smnavbar.gif" usemap="#library-map" border="0" alt="Library Navigation Links"><p>
<font size="-1"><a href="copyrght.htm">Copyright &copy; 2001</a> O'Reilly &amp; Associates. All rights reserved.</font>
</p>
<map name="library-map"> <area shape="rect" coords="2,-1,79,99" href="../index.htm"><area shape="rect" coords="84,1,157,108" href="../perlnut/index.htm"><area shape="rect" coords="162,2,248,125" href="../prog/index.htm"><area shape="rect" coords="253,2,326,130" href="../advprog/index.htm"><area shape="rect" coords="332,1,407,112" href="../cookbook/index.htm"><area shape="rect" coords="414,2,523,103" href="../sysadmin/index.htm">
</map>

<!-- END OF BODY -->

</body>
</html>