ch19_01.htm

编程珍珠,里面很多好用的代码,大家可以参考学习呵呵,

You can even put backup copies of the original files into another
directory (provided that the directory already exists):
<blockquote>
<pre class="programlisting">
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -pi'old/*.orig' -e 's/foo/bar/' xyx</b></tt> # backup to 'old/xyx.orig'</pre>
</blockquote>
These pairs of one-liners are equivalent:
<blockquote>
<pre class="programlisting">
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -pi -e 's/foo/bar/' xyx</b></tt>             # overwrite current file
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -pi'*' -e 's/foo/bar/' xyx</b></tt>          # overwrite current file

<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -pi'.orig' -e 's/foo/bar/' xyx</b></tt>      # backup to 'xyx.orig'
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -pi'*.orig' -e 's/foo/bar/' xyx</b></tt>     # backup to 'xyx.orig'</pre>
</blockquote>

From the shell, saying:
<blockquote>
<pre class="programlisting">
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -p -i.orig -e "s/foo/bar/;"</b></tt>
</pre>
</blockquote>

is the same as using the program:
<blockquote>
<pre class="programlisting">#!/usr/bin/perl -pi.orig
s/foo/bar/;</pre>
</blockquote>

which is convenient shorthand for the remarkably longer:
<blockquote>
<pre class="programlisting">#!/usr/bin/perl
$extension = '.orig';
LINE: while (&lt;&gt;) {
    if ($ARGV ne $oldargv) {
        if ($extension !~ /\*/) {
            $backup = $ARGV . $extension;
        }
        else {
            ($backup = $extension) =~ s/\*/$ARGV/g;
        }
        unless (rename($ARGV, $backup)) {
            warn "cannot rename $ARGV to $backup: $!\n";
            close ARGV;
            next;
        } 
        open(ARGVOUT, "&gt;$ARGV");
        select(ARGVOUT);
        $oldargv = $ARGV;
    }
    s/foo/bar/;
}
continue {
    print;  # this prints to original filename
}
select(STDOUT);</pre>
</blockquote>
<a name="INDEX-3411"></a><a name="INDEX-3412"></a>
This long code is virtually identical to the simple one-liner with the
<span class="option">-i</span> switch, except that the <span class="option">-i</span> form
doesn't need to compare <tt class="literal">$ARGV</tt> to
<tt class="literal">$oldargv</tt> to know when the filename has changed.  It
does, however, use <tt class="literal">ARGVOUT</tt> for the selected
filehandle and restore the old <tt class="literal">STDOUT</tt> as the
default output filehandle after the loop.  Like the code above, Perl
creates the backup file irrespective of whether any output has truly
changed.  See the description of the <tt class="literal">eof</tt> function
for examples of

how to use use <tt class="literal">eof</tt> without
parentheses to locate the end of each input file, in case you want to
append to each file or to reset line numbering.</p>

<p>If, for a given file, Perl is unable to create the backup file as
specified in the <em class="replaceable">EXTENSION</em>, it will issue a
warning to that effect and continue processing any other remaining
files listed.</p>

<p>You cannot use <span class="option">-i</span> to create directories or to strip
extensions from files.  Nor can you use it with a <tt class="literal">~</tt>
to indicate a home directory--which is just as well, since some folks
like to use that character for their backup files:
<blockquote>
<pre class="programlisting">
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -pi~ -e 's/foo/bar/' file1 file2 file3...</b></tt>
</pre>
</blockquote>

Finally, the <span class="option">-i</span> switch does not stop Perl from
running if no filenames are given on the command line.  When this
happens, no backup is made since the original file cannot be
determined, and processing proceeds from <tt class="literal">STDIN</tt> to
<tt class="literal">STDOUT</tt> as might be expected.</p>
</dd>


<dt>
<b><span class="option">-I</span><em class="replaceable">DIRECTORY</em></b>
</dt>
<dd>
<p>
<a name="INDEX-3413"></a><a name="INDEX-3414"></a><a name="INDEX-3415"></a><a name="INDEX-3416"></a><a name="INDEX-3417"></a>
Directories specified by <span class="option">-I</span> are prepended to <tt class="literal">@INC</tt>, which holds
the search path for modules.  <span class="option">-I</span> also tells the C preprocessor
where to search for include files.  The C preprocessor is invoked
with <span class="option">-P</span>; by default it searches <em class="emphasis">/usr/include</em> and <em class="emphasis">/usr/lib/perl</em>.
Unless you're going to be using the C preprocessor (and almost no
one does any more), you're better off using the <tt class="literal">use lib</tt> directive
within your script.  Like <tt class="literal">use lib</tt>, however, the <span class="option">-I</span> switch
implicitly adds platform-specific directories.  See <tt class="literal">use lib</tt> in
<a href="ch31_01.htm">Chapter 31, "Pragmatic Modules"</a>, for details.</p>
</dd>


<dt>
<b><span class="option">-l</span><em class="replaceable">OCTNUM</em></b>
</dt>
<dt>
<b><span class="option">-l</span></b>
</dt>
<dd>
<p>
<a name="INDEX-3418"></a><a name="INDEX-3419"></a><a name="INDEX-3420"></a><a name="INDEX-3421"></a><a name="INDEX-3422"></a>
Enables automatic line-end processing.  It has two effects:  first, it
automatically <tt class="literal">chomp</tt>s the line terminator when used with <span class="option">-n</span> or
<span class="option">-p</span>, and second, it sets <tt class="literal">$\</tt> to the value of <em class="replaceable">OCTNUM</em> so that any print
statements will have a line terminator of ASCII value <em class="replaceable">OCTNUM</em> added
back on.  If <em class="replaceable">OCTNUM</em> is omitted, <span class="option">-l</span>
sets <tt class="literal">$\</tt> to the current value of
<tt class="literal">$/</tt>, typically newline.  So, to trim lines to 80 columns, say this:
<blockquote>
<pre class="programlisting">
<tt class="computeroutput">%</tt> <tt class="userinput"><b>perl -lpe 'substr($_, 80) = ""'</b></tt>
</pre>
</blockquote>

Note that the assignment <tt class="literal">$\ = $/</tt> is done when the switch is
processed, so the input record separator can be different from the
output record separator if the <span class="option">-l</span> switch is followed by a <span class="option">-0</span>
switch:
<blockquote>
<pre class="programlisting">
<tt class="computeroutput">%</tt> <tt class="userinput"><b>gnufind / -print0 | perl -ln0e 'print "found $_" if -p'</b></tt>
</pre>
</blockquote>

This sets <tt class="literal">$\</tt> to newline and later sets <tt class="literal">$/</tt> to the null character.
(Note that <tt class="literal">0</tt> would have been interpreted as part of the <span class="option">-l</span> switch
had it followed the <span class="option">-l</span> directly.  That's why we bundled the <span class="option">-n</span>
switch between them.)</p>
</dd>


<dt>
<b><span class="option">-m</span> and <span class="option">-M</span></b>
</dt>
<dd>
<p>These switches load a <em class="replaceable">MODULE</em> as if you'd executed a <tt class="literal">use</tt>, unless
you specify <span class="option">-</span><em class="replaceable">MODULE</em> instead of <em class="replaceable">MODULE</em>, in which case they
invoke <tt class="literal">no</tt>.  For example, <span class="option">-Mstrict</span> is like <tt class="literal">use strict</tt>, while
<span class="option">-M-strict</span> is like <tt class="literal">no strict</tt>.</p>

<dl>
<dt>
<b><span class="option">-m</span><em class="replaceable">MODULE</em></b>
</dt>
<dd>
<p>
<a name="INDEX-3423"></a><a name="INDEX-3424"></a><a name="INDEX-3425"></a><a name="INDEX-3426"></a><a name="INDEX-3427"></a>
Executes <tt class="literal">use</tt><em class="replaceable">MODULE</em><tt class="literal">()</tt> before executing your script.</p>
</dd>


<dt>
<b><span class="option">-M</span><em class="replaceable">MODULE</em></b>
</dt>
<dt>
<b><span class="option">-M'</span><em class="replaceable">MODULE</em> <span class="option">...'</span></b>
</dt>
<dd>
<p>Executes <tt class="literal">use</tt><em class="replaceable">MODULE</em>
before executing your script.  The command is formed by mere
interpolation of the rest of the argument after the
<span class="option">-M</span>, so you can use quotes to add extra code after the
module name, for example, <span class="option">-M'MODULE qw(foo bar)'</span>.</p>
</dd>


<dt>
<b><span class="option">-M</span><em class="replaceable">MODULE</em><span class="option">=</span><em class="replaceable">arg1</em><span class="option">,</span><em class="replaceable">arg2</em><span class="option">...</span></b>
</dt>
<dd>
<p>A little built-in syntactic sugar means you can also say
<tt class="userinput"><b>-Mmodule=foo,bar</b></tt> as a shortcut for
<tt class="userinput"><b>-M'module qw(foo bar)'</b></tt>.  This avoids the need
to use quotes when importing symbols.  The actual code generated by
<tt class="userinput"><b>-Mmodule=foo,bar</b></tt> is:
<blockquote>
<pre class="programlisting">use module split(/,/, q{foo,bar})</pre>
</blockquote>

Note that the <tt class="literal">=</tt> form removes the distinction
between <span class="option">-m</span> and <span class="option">-M</span>, but it's better
to use the uppercase form to avoid confusion.</p>
</dd>

</dl>

<p>You may only use the <span class="option">-M</span> and <span class="option">-m</span>
switches from a real command-line invocation of Perl, not as options
picked up on the <tt class="literal">#!</tt> line.  (Hey, if you're gonna
put it in the file, why not just write the equivalent
<tt class="literal">use</tt> or <tt class="literal">no</tt> instead?)</p>
</dd>


<dt>
<b><span class="option">-n</span></b>
</dt>
<dd>
<p>
<a name="INDEX-3428"></a><a name="INDEX-3429"></a>
Causes Perl to assume the following loop around your script, which makes
it iterate over filename arguments much as <em class="emphasis">sed -n</em> or <em class="emphasis">awk</em> do:
<blockquote>
<pre class="programlisting">LINE:
while (&lt;&gt;) {
    ...                # your script goes here
}</pre>
</blockquote>

You may use <tt class="literal">LINE</tt> as a loop label from within you
script, even though you can't see the actual label in your file.</p>

<p>Note that the lines are not printed by default.  See
<span class="option">-p</span> to have lines printed.  Here is an efficient way
to delete all files older than a week:
<blockquote>
<pre class="programlisting">find . -mtime +7 -print | perl -nle unlink</pre>
</blockquote>

This is faster than using the <em class="emphasis">-exec</em> switch of
<em class="emphasis">find</em>(1) because you don't have to start a
process on every filename found. By an amazing coincidence,
<tt class="literal">BEGIN</tt> and <tt class="literal">END</tt> blocks may be used
to capture control before or after the implicit loop, just as in
<em class="emphasis">awk</em>.</p>
</dd>


<dt>
<b><span class="option">-p</span></b>
</dt>
<dd>
<p>
<a name="INDEX-3430"></a><a name="INDEX-3431"></a>
Causes Perl to assume the following loop around your script, which makes
it iterate over filename arguments much as <em class="emphasis">sed</em> does:
<blockquote>
<pre class="programlisting">LINE:
while (&lt;&gt;) {
    ...         # your script goes here
}
continue {
    print or die "-p destination: $!\n";
}</pre>
</blockquote>

You may use <tt class="literal">LINE</tt> as a loop label from within you
script, even though you can't see the actual label in your file.</p>

<p>If a file named by an argument cannot be opened for some reason, Perl
warns you about it, and moves on to the next file.  Note that the
lines are printed automatically.  An error occurring during printing
is treated as fatal.  By yet another amazing coincidence,
<tt class="literal">BEGIN</tt> and <tt class="literal">END</tt> blocks may be used
to capture control before or after the implicit loop, just as in
<em class="emphasis">awk</em>.</p>
</dd>


<dt>
<b><span class="option">-P</span></b>
</dt>
<dd>
<p>
<a name="INDEX-3432"></a><a name="INDEX-3433"></a><a name="INDEX-3434"></a>
Causes your script to be run through the C preprocessor before
compilation by Perl.  (Since both comments and
<em class="emphasis">cpp</em>(1) directives begin with the
<tt class="literal">#</tt> character, you should avoid starting comments
with any words recognized by the C preprocessor such as
"<tt class="literal">if</tt>", "<tt class="literal">else</tt>" or
"<tt class="literal">define</tt>".)  Whether you use <span class="option">-P</span>
switch or not, Perl still pays attention to <tt class="literal">#line</tt>
directives to control the line number and filename, so any
preprocessor can apprise Perl of these things. See the section
<a href="ch24_05.htm#ch24-sect-gen">Section 19.5.2, "Generating Perl in Other Languages"</a> in <a href="ch24_01.htm">Chapter 24, "Common Practices"</a>.</p>
</dd>


<dt>
<b><span class="option">-s</span></b>
</dt>
<dd>
<p>
<a name="INDEX-3435"></a><a name="INDEX-3436"></a>
Enables rudimentary switch-parsing for switches on the command line
after the script name but before any filename arguments or a
"<span class="option">--</span>" switch-processing terminator.  Any
switch found is removed from <tt class="literal">@ARGV</tt>, and a variable
by the same name as the switch is set in Perl.  Switch bundling is not
allowed because multicharacter switches are permitted.</p>

<p>The following script prints "<tt class="literal">true</tt>" only when the
script is invoked with a <tt class="literal">-foo</tt> switch.
<blockquote>
<pre class="programlisting">#!/usr/bin/perl -s
if ($foo) { print "true\n" }</pre>
</blockquote>