ch24_02.htm

编程珍珠,里面很多好用的代码,大家可以参考学习呵呵,

<em class="emphasis">group</em> files) that are apt to be reused. It's
particularly important to cache entries from the network.  For
example, to cache the return value from
<tt class="literal">gethostbyaddr</tt> when you are converting numeric
addresses (like <tt class="literal">204.148.40.9</tt>) to names (like
"<tt class="literal">www.oreilly.com</tt>"), you can use something like:
<blockquote>
<pre class="programlisting">sub numtoname {
    local ($_) = @_;
    unless (defined $numtoname{$_}) {
        my (@a) = gethostbyaddr(pack('C4', split(/\./)),2);
        $numtoname{$_} = @a &gt; 0 ? $a[0] : $_;
    }
    return $numtoname{$_};
}</pre>
</blockquote>
</p>
</li>
<li>
<p>
<a name="INDEX-4231"></a>
Avoid unnecessary syscalls. Operating system calls tend to be rather
expensive.  So for example, don't call the <tt class="literal">time</tt> operator when a
cached value of <tt class="literal">$now</tt> would do.  Use the special <tt class="literal">_</tt> filehandle to
avoid unnecessary <em class="emphasis">stat</em>(2) calls.  On some systems, even a minimal
syscall may execute a thousand instructions.</p>
</li>
<li>
<p>
<a name="INDEX-4232"></a>
Avoid unnecessary <tt class="literal">system</tt> calls. The <tt class="literal">system</tt> function has to fork a
subprocess in order to execute the program you specify--or worse, execute a
shell to execute the program.  This can easily execute a
million instructions.</p>
</li>
<li>
<p>
<a name="INDEX-4233"></a>
Worry about starting subprocesses, but only if they're frequent. Starting a single <em class="emphasis">pwd</em>, <em class="emphasis">hostname</em>, or <em class="emphasis">find</em> process isn't going to
hurt you much--after all, a shell starts subprocesses all day long.  We
do occasionally encourage the toolbox approach, believe it or not.</p>
</li>
<li>
<p>Keep track of your working directory yourself rather than calling <em class="emphasis">pwd</em>
repeatedly. (A standard module is provided for this. See
 <tt class="literal">Cwd</tt> in <a href="ch30_01.htm">Chapter 30, "The Standard Perl Library"</a>.)</p>
</li>
<li>
<p>
<a name="INDEX-4234"></a>
Avoid shell metacharacters in commands--pass lists to <tt class="literal">system</tt> and
<tt class="literal">exec</tt> where appropriate.</p>
</li>
<li>
<p>
<a name="INDEX-4235"></a><a name="INDEX-4236"></a>
Set the sticky bit on the Perl interpreter on machines without demand
paging:
<blockquote>
<pre class="programlisting">chmod +t /usr/bin/perl</pre>
</blockquote>
</p>
</li>
<li>
<p>Allowing built-in functions' arguments to default to <tt class="literal">$_</tt> doesn't make your
program faster.</p>
</li>
</ul>






<h3 class="sect2">24.2.2. Space Efficiency</h3>

<ul>
<li>
<p>
<a name="INDEX-4237"></a>
You can use <tt class="literal">vec</tt> for compact integer array storage
if the integers are of fixed width.  (Integers of variable width can
be stored in a UTF-8 string.)</p>
</li>
<li>
<p>Prefer numeric values over equivalent string values--they require less

memory.
</p>
</li>
<li>
<p>Use <tt class="literal">substr</tt> to store constant-length strings in a longer string.</p>
</li>
<li>
<p>Use the <tt class="literal">Tie::SubstrHash</tt> module for very compact storage of a hash array,
if the key and value lengths are fixed.</p>
</li>
<li>
<p>Use <tt class="literal">__END__</tt> and the <tt class="literal">DATA</tt> filehandle to avoid storing program data
as both a string and an array.</p>
</li>
<li>
<p>Prefer <tt class="literal">each</tt> to <tt class="literal">keys</tt> where order doesn't matter.</p>
</li>
<li>
<p>Delete or <tt class="literal">undef</tt> globals that are no longer in use.</p>
</li>
<li>
<p>Use some kind of DBM to store hashes.</p>
</li>
<li>
<p>Use temp files to store arrays.</p>
</li>
<li>
<p>Use pipes to offload processing to other tools.</p>
</li>
<li>
<p>Avoid list operations and entire file slurps.</p>
</li>
<li>
<p>Avoid using <tt class="literal">tr///</tt>.  Each <tt class="literal">tr///</tt> expression must store a
sizable translation table.</p>
</li>
<li>
<p>Don't unroll your loops or inline your subroutines.</p>
</li>
</ul>






<h3 class="sect2">24.2.3. Programmer Efficiency</h3>

<ul>
<li>
<p>
<a name="INDEX-4238"></a>
Use defaults.</p>
</li>
<li>
<p>Use funky shortcut command-line switches like <span class="option">-a</span>, <span class="option">-n</span>, <span class="option">-p</span>,
<span class="option">-s</span>, and <span class="option">-i</span>.</p>
</li>
<li>
<p>Use <tt class="literal">for</tt> to mean <tt class="literal">foreach</tt>.</p>
</li>
<li>
<p>Run system commands with backticks.</p>
</li>
<li>
<p>Use <tt class="literal">&lt;*&gt;</tt> and such.</p>
</li>
<li>
<p>Use patterns created at run time.</p>
</li>
<li>
<p>Use <tt class="literal">*</tt>, <tt class="literal">+</tt>, and <tt class="literal">{}</tt> liberally in your patterns.</p>
</li>
<li>
<p>Process whole arrays and slurp entire files.</p>
</li>
<li>
<p>Use <tt class="literal">getc</tt>.</p>
</li>
<li>
<p>Use <tt class="literal">$`</tt>, <tt class="literal">$&amp;</tt>, and <tt class="literal">$'</tt>.</p>
</li>
<li>
<p>Don't check error values on <tt class="literal">open</tt>, since <tt class="literal">&lt;</tt><em class="replaceable">HANDLE</em><tt class="literal">&gt;</tt> and
<tt class="literal">print</tt><em class="replaceable">HANDLE</em> will simply behave as no-ops when given an invalid handle.</p>
</li>
<li>
<p>Don't <tt class="literal">close</tt> your files--they'll be closed on the next <tt class="literal">open</tt>.</p>
</li>
<li>
<p>Don't pass subroutine arguments. Use globals.</p>
</li>
<li>
<p>Don't name your subroutine parameters. You can access them directly as
<tt class="literal">$_[</tt><em class="replaceable">EXPR</em><tt class="literal">]</tt>.</p>
</li>
<li>
<p>Use whatever you think of first.</p>
</li>
</ul>






<h3 class="sect2">24.2.4. Maintainer Efficiency</h3>

<ul>
<li>
<p>Don't use defaults.</p>
</li>
<li>
<p>Use <tt class="literal">foreach</tt> to mean <tt class="literal">foreach</tt>.</p>
</li>
<li>
<p>Use meaningful loop labels with <tt class="literal">next</tt> and <tt class="literal">last</tt>.</p>
</li>
<li>
<p>Use meaningful variable names.</p>
</li>
<li>
<p>Use meaningful subroutine names.</p>
</li>
<li>
<p>Put the important thing first on the line using <tt class="literal">and</tt>, <tt class="literal">or</tt>, and
statement modifiers (like <tt class="literal">exit if $done</tt>).</p>
</li>
<li>
<p>Close your files as soon as you're done with them.</p>
</li>
<li>
<p>Use packages, modules, and classes to hide your implementation details.</p>
</li>
<li>
<p>Pass arguments as subroutine parameters.</p>
</li>
<li>
<p>Name your subroutine parameters using <tt class="literal">my</tt>.</p>
</li>
<li>
<p>Parenthesize for clarity.</p>
</li>
<li>
<p>Put in lots of (useful) comments.</p>
</li>
<li>
<p>Include embedded pod documentation.</p>
</li>
<li>
<p>
<tt class="literal">use warnings</tt>.</p>
</li>
<li>
<p>
<tt class="literal">use strict</tt>.</p>
</li>
</ul>






<h3 class="sect2">24.2.5. Porter Efficiency</h3>

<ul>
<li>
<p>
<a name="INDEX-4239"></a>
Wave a handsome tip under his nose.</p>
</li>
<li>
<p>Avoid functions that aren't implemented everywhere. You can use <tt class="literal">eval</tt>
tests to see what's available.</p>
</li>
<li>
<p>Use the <tt class="literal">Config</tt> module or the <tt class="literal">$^O</tt> variable to find out what kind of
machine you're running on.</p>
</li>
<li>
<p>Don't expect native float and double to <tt class="literal">pack</tt> and <tt class="literal">unpack</tt> on foreign
machines.</p>
</li>
<li>
<p>Use network byte order (the "<tt class="literal">n</tt>" and "<tt class="literal">N</tt>" formats for <tt class="literal">pack</tt>) when
sending binary data over the network.</p>
</li>
<li>
<p>Don't send binary data over the network.  Send ASCII.  Better, send UTF-8.
Better yet, send money.</p>
</li>
<li>
<p>Check <tt class="literal">$]</tt> or <tt class="literal">$^V</tt> to see if the current version supports all the
features you use.</p>
</li>
<li>
<p>Don't use <tt class="literal">$]</tt> or <tt class="literal">$^V</tt>.  Use <tt class="literal">require</tt> or <tt class="literal">use</tt> with a version number.</p>
</li>
<li>
<p>Put in the <tt class="literal">eval exec</tt> hack even if you don't use it, so your program
will run on those few systems that have Unix-like shells but don't
recognize the <tt class="literal">#!</tt> notation.</p>
</li>
<li>
<p>Put the <tt class="literal">#!/usr/bin/perl</tt> line in even if you don't use it.</p>
</li>
<li>
<p>Test for variants of Unix commands. Some <em class="emphasis">find</em> programs can't handle the <em class="emphasis">-xdev</em> switch,
for example.</p>
</li>
<li>
<p>Avoid variant Unix commands if you can do it internally. Unix commands
don't work too well on MS-DOS or VMS.</p>
</li>
<li>
<p>Put all your scripts and manpages into a single network filesystem that's
mounted on all your machines.</p>
</li>
<li>
<p>Publish your module on CPAN.  You'll get lots of feedback if it's not
portable.</p>
</li>
</ul>






<h3 class="sect2">24.2.6. User Efficiency</h3>

<ul>
<li>
<p>
<a name="INDEX-4240"></a>
Instead of making users enter data line by line, pop users into
their favorite editor.</p>
</li>
<li>
<p>
<a name="INDEX-4241"></a><a name="INDEX-4242"></a><a name="INDEX-4243"></a>
Better yet, use a GUI like the Perl/Tk extension, where users can
control the order of events.  (Perl/Tk is available on CPAN.)</p>
</li>
<li>
<p>Put up something for users to read while you continue doing work.</p>
</li>
<li>
<p>Use autoloading so that the program <em class="emphasis">appears</em> to run faster.</p>
</li>
<li>
<p>Give the option of helpful messages at every prompt.</p>
</li>
<li>
<p>Give a helpful usage message if users don't give correct input.</p>
</li>
<li>
<p>Display the default action at every prompt, and maybe a few
alternatives.</p>
</li>
<li>
<p>Choose defaults for beginners.  Allow experts to change the defaults.</p>
</li>
<li>
<p>Use single character input where it makes sense.</p>
</li>
<li>
<p>Pattern the interaction after other things the user is familiar with.</p>
</li>
<li>
<p>
<a name="INDEX-4244"></a>
Make error messages clear about what needs fixing.  Include all
pertinent information such as filename and error code, like this:
<blockquote>
<pre class="programlisting">open(FILE, $file) or die "$0: Can't open $file for reading: $!\n";</pre>
</blockquote>
</p>
</li>
<li>
<p>Use <tt class="literal">fork &amp;&amp; exit</tt> to detach from the terminal when the rest of the script is just batch
processing.</p>
</li>
<li>
<p>Allow arguments to come from either the command line or standard
input.</p>
</li>
<li>
<p>Don't put arbitrary limitations into your program.</p>
</li>
<li>
<p>Prefer variable-length fields over fixed-length fields.</p>
</li>
<li>
<p>Use text-oriented network protocols.</p>
</li>
<li>
<p>Tell everyone else to use text-oriented network protocols!</p>
</li>
<li>
<p>Tell everyone else to tell everyone else to use text-oriented network protocols!!!</p>
</li>
<li>
<p>Be vicariously lazy.</p>
</li>
<li>
<p>Be nice.</p>
</li>
</ul>
<a name="INDEX-4245"></a><a name="INDEX-4246"></a><a name="INDEX-4247"></a>



<!-- 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="ch24_01.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="ch24_03.htm"><img src="../gifs/txtnexta.gif" alt="Next" border="0"></a></td>
</tr>
<tr>
<td align="left" valign="top" width="172">24.1. Common Goofs for Novices</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">24.3. Programming with Style</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>