1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">
<html>
<head>
<title>perlperf - perldoc.perl.org</title>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta http-equiv="Content-Language" content="en-gb">
<link rel="search" type="application/opensearchdescription+xml" title="Search perldoc.perl.org" href="/static/search.xml"/>
<link href="static/css-20100830.css" rel="stylesheet" rev="stylesheet" type="text/css" media="screen">
<link href="static/exploreperl.css" rel="stylesheet" rev="stylesheet" type="text/css">
</head>
<body onLoad="perldoc.startup();" onPageShow="if (event.persisted) perldoc.startup();">
<div id="page">
<div id="header">
<div id="homepage_link">
<a href="index.html"></a>
</div>
<div id="strapline">
Perl Programming Documentation
</div>
<div id="download_link" class="download">
<a href="http://www.perl.org/get.html">Download Perl</a>
</div>
<div id="explore_link" class="download">
<a id="explore_anchor" href="#">Explore</a>
</div>
</div>
<div id="body">
<div id="left_column">
<div class="side_group">
<div class="side_panel doc_panel">
<p>Manual</p>
<ul>
<li><a href="index-overview.html">Overview</a>
<li><a href="index-tutorials.html">Tutorials</a>
<li><a href="index-faq.html">FAQs</a>
<li><a href="index-history.html">History / Changes</a>
<li><a href="index-licence.html">License</a>
</ul>
</div>
<div class="side_panel doc_panel">
<p>Reference</p>
<ul>
<li><a href="index-language.html">Language</a>
<li><a href="index-functions.html">Functions</a>
<li><a href="perlop.html">Operators</a>
<li><a href="perlvar.html">Special Variables</a>
<li><a href="index-pragmas.html">Pragmas</a>
<li><a href="index-utilities.html">Utilities</a>
<li><a href="index-internals.html">Internals</a>
<li><a href="index-platforms.html">Platform Specific</a>
</ul>
</div>
<div class="side_panel doc_panel">
<p>Modules</p>
<ul>
<li>
<a href="index-modules-A.html">A</a>
•
<a href="index-modules-B.html">B</a>
•
<a href="index-modules-C.html">C</a>
•
<a href="index-modules-D.html">D</a>
•
<a href="index-modules-E.html">E</a>
<li>
<a href="index-modules-F.html">F</a>
•
<a href="index-modules-G.html">G</a>
•
<a href="index-modules-H.html">H</a>
•
<a href="index-modules-I.html">I</a>
•
<a href="index-modules-L.html">L</a>
<li>
<a href="index-modules-M.html">M</a>
•
<a href="index-modules-N.html">N</a>
•
<a href="index-modules-O.html">O</a>
•
<a href="index-modules-P.html">P</a>
•
<a href="index-modules-S.html">S</a>
<li>
<a href="index-modules-T.html">T</a>
•
<a href="index-modules-U.html">U</a>
•
<a href="index-modules-X.html">X</a>
</ul>
</div>
<div class="side_panel doc_panel">
<p>Tools</p>
<ul>
<li><a href="preferences.html">Preferences</a>
</ul>
</div>
</div>
</div>
<div id="centre_column">
<div id="content_header">
<div id="title_bar">
<div id="page_name">
<h1>perlperf</h1>
</div>
<div id="perl_version">
Perl 5 version 18.2 documentation
</div>
<div class="page_links" id="page_links_top">
<a href="#" onClick="toolbar.goToTop();return false;">Go to top</a>
</div>
<div class="page_links" id="page_links_bottom">
<a href="#" id="page_index_toggle">Show page index</a> •
<a href="#" id="recent_pages_toggle">Show recent pages</a>
</div>
<div id="search_form">
<form action="search.html" method="GET" id="search">
<input type="text" name="q" id="search_box" alt="Search">
</form>
</div>
</div>
<div id="breadcrumbs">
<a href="index.html">Home</a> >
<a href="index-language.html">Language reference</a> >
perlperf
</div>
</div>
<div id="content_body">
<!--[if lt IE 7]>
<div class="noscript">
<p>
<strong>It looks like you're using Internet Explorer 6. This is a very old
browser which does not offer full support for modern websites.</strong>
</p>
<p>
Unfortunately this means that this website will not work on
your computer.
</p>
<p>
Don't miss out though! To view the site (and get a better experience from
many other websites), simply upgrade to
<a href="http://www.microsoft.com/windows/Internet-explorer/default.aspx">Internet
Explorer 8</a>
or download an alternative browser such as
<a href="http://www.mozilla.com/en-US/firefox/firefox.html">Firefox</a>,
<a href="http://www.apple.com/safari/download/">Safari</a>, or
<a href="http://www.google.co.uk/chrome">Google Chrome</a>.
</p>
<p>
All of these browsers are free. If you're using a PC at work, you may
need to contact your IT administrator.
</p>
</div>
<![endif]-->
<noscript>
<div class="noscript">
<p>
<strong>Please note: Many features of this site require JavaScript. You appear to have JavaScript disabled,
or are running a non-JavaScript capable web browser.</strong>
</p>
<p>
To get the best experience, please enable JavaScript or download a modern web browser such as <a href="http://www.microsoft.com/windows/Internet-explorer/default.aspx">Internet Explorer 8</a>, <a href="http://www.mozilla.com/en-US/firefox/firefox.html">Firefox</a>, <a href="http://www.apple.com/safari/download/">Safari</a>, or <a href="http://www.google.co.uk/chrome">Google Chrome</a>.
</p>
</div>
</noscript>
<div id="recent_pages" class="hud_container">
<div id="recent_pages_header" class="hud_header">
<div id="recent_pages_close" class="hud_close"><a href="#" onClick="recentPages.hide();return false;"></a></div>
<div id="recent_pages_title" class="hud_title"><span class="hud_span_top">Recently read</span></div>
<div id="recent_pages_topright" class="hud_topright"></div>
</div>
<div id="recent_pages_content" class="hud_content">
</div>
<div id="recent_pages_footer" class="hud_footer">
<div id="recent_pages_bottomleft" class="hud_bottomleft"></div>
<div id="recent_pages_bottom" class="hud_bottom"><span class="hud_span_bottom"></span></div>
<div id="recent_pages_resize" class="hud_resize"></div>
</div>
</div>
<div id="from_search"></div>
<h1>perlperf</h1>
<!-- -->
<ul><li><a href="#NAME">NAME</a><li><a href="#DESCRIPTION">DESCRIPTION</a><li><a href="#OVERVIEW">OVERVIEW</a><ul><li><a href="#ONE-STEP-SIDEWAYS">ONE STEP SIDEWAYS</a><li><a href="#ONE-STEP-FORWARD">ONE STEP FORWARD</a><li><a href="#ANOTHER-STEP-SIDEWAYS">ANOTHER STEP SIDEWAYS</a></ul><li><a href="#GENERAL-GUIDELINES">GENERAL GUIDELINES</a><li><a href="#BENCHMARKS">BENCHMARKS</a><ul><li><a href="#Assigning-and-Dereferencing-Variables.">Assigning and Dereferencing Variables.</a><li><a href="#Search-and-replace-or-tr">Search and replace or tr</a></ul><li><a href="#PROFILING-TOOLS">PROFILING TOOLS</a><ul><li><a href="#Devel%3a%3aDProf">Devel::DProf</a><li><a href="#Devel%3a%3aProfiler">Devel::Profiler</a><li><a href="#Devel%3a%3aSmallProf">Devel::SmallProf</a><li><a href="#Devel%3a%3aFastProf">Devel::FastProf</a><li><a href="#Devel%3a%3aNYTProf">Devel::NYTProf</a></ul><li><a href="#SORTING">SORTING</a><li><a href="#LOGGING">LOGGING</a><ul><li><a href="#Logging-if-DEBUG-(constant)">Logging if DEBUG (constant)</a></ul><li><a href="#POSTSCRIPT">POSTSCRIPT</a><li><a href="#SEE-ALSO">SEE ALSO</a><ul><li><a href="#PERLDOCS">PERLDOCS</a><li><a href="#MAN-PAGES">MAN PAGES</a><li><a href="#MODULES">MODULES</a><li><a href="#URLS">URLS</a></ul><li><a href="#AUTHOR">AUTHOR</a></ul><a name="NAME"></a><h1>NAME</h1>
<p>perlperf - Perl Performance and Optimization Techniques</p>
<a name="DESCRIPTION"></a><h1>DESCRIPTION</h1>
<p>This is an introduction to the use of performance and optimization techniques
which can be used with particular reference to perl programs. While many perl
developers have come from other languages, and can use their prior knowledge
where appropriate, there are many other people who might benefit from a few
perl specific pointers. If you want the condensed version, perhaps the best
advice comes from the renowned Japanese Samurai, Miyamoto Musashi, who said:</p>
<pre class="verbatim"><ol><li> <span class="q">"Do Not Engage in Useless Activity"</span></li></ol></pre><p>in 1645.</p>
<a name="OVERVIEW"></a><h1>OVERVIEW</h1>
<p>Perhaps the most common mistake programmers make is to attempt to optimize
their code before a program actually does anything useful - this is a bad idea.
There's no point in having an extremely fast program that doesn't work. The
first job is to get a program to <i>correctly</i> do something <b>useful</b>, (not to
mention ensuring the test suite is fully functional), and only then to consider
optimizing it. Having decided to optimize existing working code, there are
several simple but essential steps to consider which are intrinsic to any
optimization process.</p>
<a name="ONE-STEP-SIDEWAYS"></a><h2>ONE STEP SIDEWAYS</h2>
<p>Firstly, you need to establish a baseline time for the existing code, which
timing needs to be reliable and repeatable. You'll probably want to use the
<code class="inline"><span class="w">Benchmark</span></code>
or <code class="inline"><span class="w">Devel::NYTProf</span></code>
modules, or something similar, for this step,
or perhaps the Unix system <code class="inline"><a class="l_k" href="functions/time.html">time</a></code> utility, whichever is appropriate. See the
base of this document for a longer list of benchmarking and profiling modules,
and recommended further reading.</p>
<a name="ONE-STEP-FORWARD"></a><h2>ONE STEP FORWARD</h2>
<p>Next, having examined the program for <i>hot spots</i>, (places where the code
seems to run slowly), change the code with the intention of making it run
faster. Using version control software, like <code class="inline"><span class="w">subversion</span></code>
, will ensure no
changes are irreversible. It's too easy to fiddle here and fiddle there -
don't change too much at any one time or you might not discover which piece of
code <b>really</b> was the slow bit.</p>
<a name="ANOTHER-STEP-SIDEWAYS"></a><h2>ANOTHER STEP SIDEWAYS</h2>
<p>It's not enough to say: "that will make it run faster", you have to check it.
Rerun the code under control of the benchmarking or profiling modules, from the
first step above, and check that the new code executed the <b>same task</b> in
<i>less time</i>. Save your work and repeat...</p>
<a name="GENERAL-GUIDELINES"></a><h1>GENERAL GUIDELINES</h1>
<p>The critical thing when considering performance is to remember there is no such
thing as a <code class="inline"><span class="w">Golden</span> <span class="w">Bullet</span></code>
, which is why there are no rules, only guidelines.</p>
<p>It is clear that inline code is going to be faster than subroutine or method
calls, because there is less overhead, but this approach has the disadvantage
of being less maintainable and comes at the cost of greater memory usage -
there is no such thing as a free lunch. If you are searching for an element in
a list, it can be more efficient to store the data in a hash structure, and
then simply look to see whether the key is defined, rather than to loop through
the entire array using grep() for instance. substr() may be (a lot) faster
than grep() but not as flexible, so you have another trade-off to access. Your
code may contain a line which takes 0.01 of a second to execute which if you
call it 1,000 times, quite likely in a program parsing even medium sized files
for instance, you already have a 10 second delay, in just one single code
location, and if you call that line 100,000 times, your entire program will
slow down to an unbearable crawl.</p>
<p>Using a subroutine as part of your sort is a powerful way to get exactly what
you want, but will usually be slower than the built-in <i>alphabetic</i> <code class="inline">cmp</code>
and
<i>numeric</i> <code class="inline"><=></code>
sort operators. It is possible to make multiple
passes over your data, building indices to make the upcoming sort more
efficient, and to use what is known as the <code class="inline"><span class="w">OM</span></code>
(Orcish Maneuver) to cache the
sort keys in advance. The cache lookup, while a good idea, can itself be a
source of slowdown by enforcing a double pass over the data - once to setup the
cache, and once to sort the data. Using <code class="inline"><a class="l_k" href="functions/pack.html">pack()</a></code> to extract the required sort
key into a consistent string can be an efficient way to build a single string
to compare, instead of using multiple sort keys, which makes it possible to use
the standard, written in <code class="inline"><span class="w">c</span></code>
and fast, perl <code class="inline"><a class="l_k" href="functions/sort.html">sort()</a></code> function on the output,
and is the basis of the <code class="inline"><span class="w">GRT</span></code>
(Guttman Rossler Transform). Some string
combinations can slow the <code class="inline"><span class="w">GRT</span></code>
down, by just being too plain complex for it's
own good.</p>
<p>For applications using database backends, the standard <code class="inline"><span class="w">DBIx</span></code>
namespace has
tries to help with keeping things nippy, not least because it tries to <i>not</i>
query the database until the latest possible moment, but always read the docs
which come with your choice of libraries. Among the many issues facing
developers dealing with databases should remain aware of is to always use
<code class="inline"><span class="w">SQL</span></code>
placeholders and to consider pre-fetching data sets when this might
prove advantageous. Splitting up a large file by assigning multiple processes
to parsing a single file, using say <code class="inline"><span class="w">POE</span></code>
, <code class="inline"><span class="w">threads</span></code>
or <code class="inline"><a class="l_k" href="functions/fork.html">fork</a></code> can also be a
useful way of optimizing your usage of the available <code class="inline"><span class="w">CPU</span></code>
resources, though
this technique is fraught with concurrency issues and demands high attention to
detail.</p>
<p>Every case has a specific application and one or more exceptions, and there is
no replacement for running a few tests and finding out which method works best
for your particular environment, this is why writing optimal code is not an
exact science, and why we love using Perl so much - TMTOWTDI.</p>
<a name="BENCHMARKS"></a><h1>BENCHMARKS</h1>
<p>Here are a few examples to demonstrate usage of Perl's benchmarking tools.</p>
<a name="Assigning-and-Dereferencing-Variables."></a><h2>Assigning and Dereferencing Variables.</h2>
<p>I'm sure most of us have seen code which looks like, (or worse than), this:</p>
<pre class="verbatim"><ol><li> if ( $obj->{_ref}->{_myscore} >= $obj->{_ref}->{_yourscore} ) {</li><li> ...</li></ol></pre><p>This sort of code can be a real eyesore to read, as well as being very
sensitive to typos, and it's much clearer to dereference the variable
explicitly. We're side-stepping the issue of working with object-oriented
programming techniques to encapsulate variable access via methods, only
accessible through an object. Here we're just discussing the technical
implementation of choice, and whether this has an effect on performance. We
can see whether this dereferencing operation, has any overhead by putting
comparative code in a file and running a <code class="inline"><span class="w">Benchmark</span></code>
test.</p>
<p># dereference</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Benchmark</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$ref</span> = <span class="s">{</span></li><li> <span class="q">'ref'</span> <span class="cm">=></span> <span class="s">{</span></li><li> <span class="w">_myscore</span> <span class="cm">=></span> <span class="q">'100 + 1'</span><span class="cm">,</span></li><li> <span class="w">_yourscore</span> <span class="cm">=></span> <span class="q">'102 - 1'</span><span class="cm">,</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="s">}</span><span class="sc">;</span></li><li></li><li> <span class="i">timethese</span><span class="s">(</span><span class="n">1000000</span><span class="cm">,</span> <span class="s">{</span></li><li> <span class="q">'direct'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$x</span> = <span class="i">$ref</span>->{<span class="w">ref</span>}->{<span class="w">_myscore</span>} . <span class="i">$ref</span>->{<span class="w">ref</span>}->{<span class="w">_yourscore</span>} <span class="sc">;</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="q">'dereference'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$ref</span> = <span class="i">$ref</span>->{<span class="w">ref</span>}<span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$myscore</span> = <span class="i">$ref</span>->{<span class="w">_myscore</span>}<span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$yourscore</span> = <span class="i">$ref</span>->{<span class="w">_yourscore</span>}<span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$x</span> = <span class="i">$myscore</span> . <span class="i">$yourscore</span><span class="sc">;</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="s">}</span><span class="s">)</span><span class="sc">;</span></li></ol></pre><p>It's essential to run any timing measurements a sufficient number of times so
the numbers settle on a numerical average, otherwise each run will naturally
fluctuate due to variations in the environment, to reduce the effect of
contention for <code class="inline"><span class="w">CPU</span></code>
resources and network bandwidth for instance. Running
the above code for one million iterations, we can take a look at the report
output by the <code class="inline"><span class="w">Benchmark</span></code>
module, to see which approach is the most effective.</p>
<pre class="verbatim"><ol><li> $> perl dereference</li><li></li><li> Benchmark: timing 1000000 iterations of dereference, direct...</li><li> dereference: 2 wallclock secs ( 1.59 usr + 0.00 sys = 1.59 CPU) @ 628930.82/s (n=1000000)</li><li> direct: 1 wallclock secs ( 1.20 usr + 0.00 sys = 1.20 CPU) @ 833333.33/s (n=1000000)</li></ol></pre><p>The difference is clear to see and the dereferencing approach is slower. While
it managed to execute an average of 628,930 times a second during our test, the
direct approach managed to run an additional 204,403 times, unfortunately.
Unfortunately, because there are many examples of code written using the
multiple layer direct variable access, and it's usually horrible. It is,
however, minusculy faster. The question remains whether the minute gain is
actually worth the eyestrain, or the loss of maintainability.</p>
<a name="Search-and-replace-or-tr"></a><h2>Search and replace or tr</h2>
<p>If we have a string which needs to be modified, while a regex will almost
always be much more flexible, <code class="inline"><a class="l_k" href="functions/tr.html">tr</a></code>, an oft underused tool, can still be a
useful. One scenario might be replace all vowels with another character. The
regex solution might look like this:</p>
<pre class="verbatim"><ol><li> <span class="i">$str</span> =~ <span class="q">s/[aeiou]/x/g</span></li></ol></pre><p>The <code class="inline"><a class="l_k" href="functions/tr.html">tr</a></code> alternative might look like this:</p>
<pre class="verbatim"><ol><li> <span class="i">$str</span> =~ <span class="q">tr/aeiou/xxxxx/</span></li></ol></pre><p>We can put that into a test file which we can run to check which approach is
the fastest, using a global <code class="inline"><span class="i">$STR</span></code>
variable to assign to the <code class="inline"><a class="l_k" href="functions/my.html">my</a> <span class="i">$str</span></code>
variable so as to avoid perl trying to optimize any of the work away by
noticing it's assigned only the once.</p>
<p># regex-transliterate</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Benchmark</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$STR</span> = <span class="q">"$$-this and that"</span><span class="sc">;</span></li><li></li><li> <span class="i">timethese</span><span class="s">(</span> <span class="n">1000000</span><span class="cm">,</span> <span class="s">{</span></li><li> <span class="q">'sr'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span> <a class="l_k" href="functions/my.html">my</a> <span class="i">$str</span> = <span class="i">$STR</span><span class="sc">;</span> <span class="i">$str</span> =~ <span class="q">s/[aeiou]/x/g</span><span class="sc">;</span> <a class="l_k" href="functions/return.html">return</a> <span class="i">$str</span><span class="sc">;</span> <span class="s">}</span><span class="cm">,</span></li><li> <span class="q">'tr'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span> <a class="l_k" href="functions/my.html">my</a> <span class="i">$str</span> = <span class="i">$STR</span><span class="sc">;</span> <span class="i">$str</span> =~ <span class="q">tr/aeiou/xxxxx/</span><span class="sc">;</span> <a class="l_k" href="functions/return.html">return</a> <span class="i">$str</span><span class="sc">;</span> <span class="s">}</span><span class="cm">,</span></li><li> <span class="s">}</span><span class="s">)</span><span class="sc">;</span></li></ol></pre><p>Running the code gives us our results:</p>
<pre class="verbatim"><ol><li> $> perl regex-transliterate</li><li></li><li> Benchmark: timing 1000000 iterations of sr, tr...</li><li> sr: 2 wallclock secs ( 1.19 usr + 0.00 sys = 1.19 CPU) @ 840336.13/s (n=1000000)</li><li> tr: 0 wallclock secs ( 0.49 usr + 0.00 sys = 0.49 CPU) @ 2040816.33/s (n=1000000)</li></ol></pre><p>The <code class="inline"><a class="l_k" href="functions/tr.html">tr</a></code> version is a clear winner. One solution is flexible, the other is
fast - and it's appropriately the programmer's choice which to use.</p>
<p>Check the <code class="inline"><span class="w">Benchmark</span></code>
docs for further useful techniques.</p>
<a name="PROFILING-TOOLS"></a><h1>PROFILING TOOLS</h1>
<p>A slightly larger piece of code will provide something on which a profiler can
produce more extensive reporting statistics. This example uses the simplistic
<code class="inline"><span class="w">wordmatch</span></code>
program which parses a given input file and spews out a short
report on the contents.</p>
<p># wordmatch</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> =<span class="w">head1</span> <span class="w">NAME</span></li><li></li><li> <span class="w">filewords</span> - <span class="w">word</span> <span class="w">analysis</span> <span class="w">of</span> <span class="w">input</span> <span class="w">file</span></li><li></li><li> =<span class="w">head1</span> <span class="w">SYNOPSIS</span></li><li></li><li> <span class="w">filewords</span> -f <span class="w">inputfilename</span> <span class="s">[</span>-d<span class="s">]</span></li><li></li><li> =<span class="w">head1</span> <span class="w">DESCRIPTION</span></li><li></li><li> <span class="w">This</span> <span class="w">program</span> <span class="w">parses</span> <span class="w">the</span> given <span class="w">filename</span><span class="cm">,</span> <span class="w">specified</span> <span class="w">with</span> <span class="w">C</span><span class="q"><-f></span><span class="cm">,</span> and <span class="w">displays</span> <span class="w">a</span></li><li> <span class="w">simple</span> <span class="w">analysis</span> <span class="w">of</span> <span class="w">the</span> <span class="w">words</span> <span class="w">found</span> <span class="w">therein</span>. <span class="w">Use</span> <span class="w">the</span> <span class="w">C</span><span class="q"><-d></span> switch <span class="w">to</span> <span class="w">enable</span></li><li> <span class="w">debugging</span> <span class="w">messages</span>.</li><li></li><li> =<span class="w">cut</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">FileHandle</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Getopt::Long</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$debug</span> = <span class="n">0</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$file</span> = <span class="q">''</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$result</span> = <span class="w">GetOptions</span> <span class="s">(</span></li><li> <span class="q">'debug'</span> <span class="cm">=></span> \<span class="i">$debug</span><span class="cm">,</span></li><li> <span class="q">'file=s'</span> <span class="cm">=></span> \<span class="i">$file</span><span class="cm">,</span></li><li> <span class="s">)</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/die.html">die</a><span class="s">(</span><span class="q">"invalid args"</span><span class="s">)</span> unless <span class="i">$result</span><span class="sc">;</span></li><li></li><li> unless <span class="s">(</span> -f <span class="i">$file</span> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/die.html">die</a><span class="s">(</span><span class="q">"Usage: $0 -f filename [-d]"</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$FH</span> = <span class="w">FileHandle</span><span class="w">->new</span><span class="s">(</span><span class="q">"< $file"</span><span class="s">)</span> or <a class="l_k" href="functions/die.html">die</a><span class="s">(</span><span class="q">"unable to open file($file): $!"</span><span class="s">)</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$i_LINES</span> = <span class="n">0</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$i_WORDS</span> = <span class="n">0</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">%count</span> = <span class="s">(</span><span class="s">)</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">@lines</span> = <span class="q"><$FH></span><span class="sc">;</span></li><li> foreach <a class="l_k" href="functions/my.html">my</a> <span class="i">$line</span> <span class="s">(</span> <span class="i">@lines</span> <span class="s">)</span> <span class="s">{</span></li><li> <span class="i">$i_LINES</span>++<span class="sc">;</span></li><li> <span class="i">$line</span> =~ <span class="q">s/\n//</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">@words</span> = <a class="l_k" href="functions/split.html">split</a><span class="s">(</span><span class="q">/ +/</span><span class="cm">,</span> <span class="i">$line</span><span class="s">)</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$i_words</span> = <a class="l_k" href="functions/scalar.html">scalar</a><span class="s">(</span><span class="i">@words</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="i">$i_WORDS</span> = <span class="i">$i_WORDS</span> + <span class="i">$i_words</span><span class="sc">;</span></li><li> <span class="i">debug</span><span class="s">(</span><span class="q">"line: $i_LINES supplying $i_words words: @words"</span><span class="s">)</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$i_word</span> = <span class="n">0</span><span class="sc">;</span></li><li> foreach <a class="l_k" href="functions/my.html">my</a> <span class="i">$word</span> <span class="s">(</span> <span class="i">@words</span> <span class="s">)</span> <span class="s">{</span></li><li> <span class="i">$i_word</span>++<span class="sc">;</span></li><li> <span class="i">$count</span>{<span class="i">$i_LINES</span>}{<span class="w">spec</span>} += <span class="i">matches</span><span class="s">(</span><span class="i">$i_word</span><span class="cm">,</span> <span class="i">$word</span><span class="cm">,</span> <span class="q">'[^a-zA-Z0-9]'</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="i">$count</span>{<span class="i">$i_LINES</span>}{<span class="w">only</span>} += <span class="i">matches</span><span class="s">(</span><span class="i">$i_word</span><span class="cm">,</span> <span class="i">$word</span><span class="cm">,</span> <span class="q">'^[^a-zA-Z0-9]+$'</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="i">$count</span>{<span class="i">$i_LINES</span>}{<span class="w">cons</span>} += <span class="i">matches</span><span class="s">(</span><span class="i">$i_word</span><span class="cm">,</span> <span class="i">$word</span><span class="cm">,</span> <span class="q">'^[(?i:bcdfghjklmnpqrstvwxyz)]+$'</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="i">$count</span>{<span class="i">$i_LINES</span>}{<span class="w">vows</span>} += <span class="i">matches</span><span class="s">(</span><span class="i">$i_word</span><span class="cm">,</span> <span class="i">$word</span><span class="cm">,</span> <span class="q">'^[(?i:aeiou)]+$'</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="i">$count</span>{<span class="i">$i_LINES</span>}{<span class="w">caps</span>} += <span class="i">matches</span><span class="s">(</span><span class="i">$i_word</span><span class="cm">,</span> <span class="i">$word</span><span class="cm">,</span> <span class="q">'^[(A-Z)]+$'</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li> <span class="s">}</span></li><li></li><li> <a class="l_k" href="functions/print.html">print</a> <span class="i">report</span><span class="s">(</span> <span class="i">%count</span> <span class="s">)</span><span class="sc">;</span></li><li></li><li><a name="matches"></a> sub <span class="m">matches</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$i_wd</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$word</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$regex</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$has</span> = <span class="n">0</span><span class="sc">;</span></li><li></li><li> if <span class="s">(</span> <span class="i">$word</span> =~ <span class="q">/($regex)/</span> <span class="s">)</span> <span class="s">{</span></li><li> <span class="i">$has</span>++ if <span class="i">$1</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li></li><li> <span class="i">debug</span><span class="s">(</span><span class="q">"word: $i_wd "</span>.<span class="s">(</span><span class="i">$has</span> ? <span class="q">'matches'</span> <span class="co">:</span> <span class="q">'does not match'</span><span class="s">)</span>.<span class="q">" chars: /$regex/"</span><span class="s">)</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/return.html">return</a> <span class="i">$has</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li></li><li><a name="report"></a> sub <span class="m">report</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">%report</span> = <span class="i">@_</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">%rep</span><span class="sc">;</span></li><li></li><li> foreach <a class="l_k" href="functions/my.html">my</a> <span class="i">$line</span> <span class="s">(</span> <a class="l_k" href="functions/keys.html">keys</a> <span class="i">%report</span> <span class="s">)</span> <span class="s">{</span></li><li> foreach <a class="l_k" href="functions/my.html">my</a> <span class="i">$key</span> <span class="s">(</span> <a class="l_k" href="functions/keys.html">keys</a> <span class="i">%</span>{ <span class="i">$report</span>{<span class="i">$line</span>} } <span class="s">)</span> <span class="s">{</span></li><li> <span class="i">$rep</span>{<span class="i">$key</span>} += <span class="i">$report</span>{<span class="i">$line</span>}{<span class="i">$key</span>}<span class="sc">;</span></li><li> <span class="s">}</span></li><li> <span class="s">}</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$report</span> = <span class="q">qq|</span></li><li> <span class="q"> $0 report for $file:</span></li><li> <span class="q"> lines in file: $i_LINES</span></li><li> <span class="q"> words in file: $i_WORDS</span></li><li> <span class="q"> words with special (non-word) characters: $i_spec</span></li><li> <span class="q"> words with only special (non-word) characters: $i_only</span></li><li> <span class="q"> words with only consonants: $i_cons</span></li><li> <span class="q"> words with only capital letters: $i_caps</span></li><li> <span class="q"> words with only vowels: $i_vows</span></li><li> <span class="q"> |</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/return.html">return</a> <span class="i">$report</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li></li><li><a name="debug"></a> sub <span class="m">debug</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$message</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></li><li></li><li> if <span class="s">(</span> <span class="i">$debug</span> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/print.html">print</a> <span class="i">STDERR</span> <span class="q">"DBG: $message\n"</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li> <span class="s">}</span></li><li></li><li> <a class="l_k" href="functions/exit.html">exit</a> <span class="n">0</span><span class="sc">;</span></li></ol></pre><a name="Devel%3a%3aDProf"></a><h2>Devel::DProf</h2>
<p>This venerable module has been the de-facto standard for Perl code profiling
for more than a decade, but has been replaced by a number of other modules
which have brought us back to the 21st century. Although you're recommended to
evaluate your tool from the several mentioned here and from the CPAN list at
the base of this document, (and currently <a href="http://search.cpan.org/perldoc/Devel::NYTProf">Devel::NYTProf</a> seems to be the
weapon of choice - see below), we'll take a quick look at the output from
<a href="http://search.cpan.org/perldoc/Devel::DProf">Devel::DProf</a> first, to set a baseline for Perl profiling tools. Run the
above program under the control of <code class="inline"><span class="w">Devel::DProf</span></code>
by using the <code class="inline">-d</code>
switch on
the command-line.</p>
<pre class="verbatim"><ol><li> $> perl -d:DProf wordmatch -f perl5db.pl</li><li></li><li> <...multiple lines snipped...></li><li></li><li> wordmatch report for perl5db.pl:</li><li> lines in file: 9428</li><li> words in file: 50243</li><li> words with special (non-word) characters: 20480</li><li> words with only special (non-word) characters: 7790</li><li> words with only consonants: 4801</li><li> words with only capital letters: 1316</li><li> words with only vowels: 1701</li></ol></pre><p><code class="inline"><span class="w">Devel::DProf</span></code>
produces a special file, called <i>tmon.out</i> by default, and
this file is read by the <code class="inline"><span class="w">dprofpp</span></code>
program, which is already installed as part
of the <code class="inline"><span class="w">Devel::DProf</span></code>
distribution. If you call <code class="inline"><span class="w">dprofpp</span></code>
with no options,
it will read the <i>tmon.out</i> file in the current directory and produce a human
readable statistics report of the run of your program. Note that this may take
a little time.</p>
<pre class="verbatim"><ol><li> $> dprofpp</li><li></li><li> Total Elapsed Time = 2.951677 Seconds</li><li> User+System Time = 2.871677 Seconds</li><li> Exclusive Times</li><li> %Time ExclSec CumulS #Calls sec/call Csec/c Name</li><li> 102. 2.945 3.003 251215 0.0000 0.0000 main::matches</li><li> 2.40 0.069 0.069 260643 0.0000 0.0000 main::debug</li><li> 1.74 0.050 0.050 1 0.0500 0.0500 main::report</li><li> 1.04 0.030 0.049 4 0.0075 0.0123 main::BEGIN</li><li> 0.35 0.010 0.010 3 0.0033 0.0033 Exporter::as_heavy</li><li> 0.35 0.010 0.010 7 0.0014 0.0014 IO::File::BEGIN</li><li> 0.00 - -0.000 1 - - Getopt::Long::FindOption</li><li> 0.00 - -0.000 1 - - Symbol::BEGIN</li><li> 0.00 - -0.000 1 - - Fcntl::BEGIN</li><li> 0.00 - -0.000 1 - - Fcntl::bootstrap</li><li> 0.00 - -0.000 1 - - warnings::BEGIN</li><li> 0.00 - -0.000 1 - - IO::bootstrap</li><li> 0.00 - -0.000 1 - - Getopt::Long::ConfigDefaults</li><li> 0.00 - -0.000 1 - - Getopt::Long::Configure</li><li> 0.00 - -0.000 1 - - Symbol::gensym</li></ol></pre><p><code class="inline"><span class="w">dprofpp</span></code>
will produce some quite detailed reporting on the activity of the
<code class="inline"><span class="w">wordmatch</span></code>
program. The wallclock, user and system, times are at the top of
the analysis, and after this are the main columns defining which define the
report. Check the <code class="inline"><span class="w">dprofpp</span></code>
docs for details of the many options it supports.</p>
<p>See also <code class="inline"><span class="w">Apache::DProf</span></code>
which hooks <code class="inline"><span class="w">Devel::DProf</span></code>
into <code class="inline"><span class="w">mod_perl</span></code>
.</p>
<a name="Devel%3a%3aProfiler"></a><h2>Devel::Profiler</h2>
<p>Let's take a look at the same program using a different profiler:
<code class="inline"><span class="w">Devel::Profiler</span></code>
, a drop-in Perl-only replacement for <code class="inline"><span class="w">Devel::DProf</span></code>
. The
usage is very slightly different in that instead of using the special <code class="inline">-d:</code>
flag, you pull <code class="inline"><span class="w">Devel::Profiler</span></code>
in directly as a module using <code class="inline">-M</code>
.</p>
<pre class="verbatim"><ol><li> $> perl -MDevel::Profiler wordmatch -f perl5db.pl</li><li></li><li> <...multiple lines snipped...></li><li></li><li> wordmatch report for perl5db.pl:</li><li> lines in file: 9428</li><li> words in file: 50243</li><li> words with special (non-word) characters: 20480</li><li> words with only special (non-word) characters: 7790</li><li> words with only consonants: 4801</li><li> words with only capital letters: 1316</li><li> words with only vowels: 1701</li></ol></pre><p><code class="inline"><span class="w">Devel::Profiler</span></code>
generates a tmon.out file which is compatible with the
<code class="inline"><span class="w">dprofpp</span></code>
program, thus saving the construction of a dedicated statistics
reader program. <code class="inline"><span class="w">dprofpp</span></code>
usage is therefore identical to the above example.</p>
<pre class="verbatim"><ol><li> $> dprofpp</li><li></li><li> Total Elapsed Time = 20.984 Seconds</li><li> User+System Time = 19.981 Seconds</li><li> Exclusive Times</li><li> %Time ExclSec CumulS #Calls sec/call Csec/c Name</li><li> 49.0 9.792 14.509 251215 0.0000 0.0001 main::matches</li><li> 24.4 4.887 4.887 260643 0.0000 0.0000 main::debug</li><li> 0.25 0.049 0.049 1 0.0490 0.0490 main::report</li><li> 0.00 0.000 0.000 1 0.0000 0.0000 Getopt::Long::GetOptions</li><li> 0.00 0.000 0.000 2 0.0000 0.0000 Getopt::Long::ParseOptionSpec</li><li> 0.00 0.000 0.000 1 0.0000 0.0000 Getopt::Long::FindOption</li><li> 0.00 0.000 0.000 1 0.0000 0.0000 IO::File::new</li><li> 0.00 0.000 0.000 1 0.0000 0.0000 IO::Handle::new</li><li> 0.00 0.000 0.000 1 0.0000 0.0000 Symbol::gensym</li><li> 0.00 0.000 0.000 1 0.0000 0.0000 IO::File::open</li></ol></pre><p>Interestingly we get slightly different results, which is mostly because the
algorithm which generates the report is different, even though the output file
format was allegedly identical. The elapsed, user and system times are clearly
showing the time it took for <code class="inline"><span class="w">Devel::Profiler</span></code>
to execute its own run, but
the column listings feel more accurate somehow than the ones we had earlier
from <code class="inline"><span class="w">Devel::DProf</span></code>
. The 102% figure has disappeared, for example. This is
where we have to use the tools at our disposal, and recognise their pros and
cons, before using them. Interestingly, the numbers of calls for each
subroutine are identical in the two reports, it's the percentages which differ.
As the author of <code class="inline"><span class="w">Devel::Proviler</span></code>
writes:</p>
<pre class="verbatim"><ol><li> ...running HTML::Template's test suite under Devel::DProf shows output()</li><li> taking NO time but Devel::Profiler shows around 10% of the time is in output().</li><li> I don't know which to trust but my gut tells me something is wrong with</li><li> Devel::DProf. HTML::Template::output() is a big routine that's called for</li><li> every test. Either way, something needs fixing.</li></ol></pre><p>YMMV.</p>
<p>See also <code class="inline"><span class="w">Devel::Apache::Profiler</span></code>
which hooks <code class="inline"><span class="w">Devel::Profiler</span></code>
into <code class="inline"><span class="w">mod_perl</span></code>
.</p>
<a name="Devel%3a%3aSmallProf"></a><h2>Devel::SmallProf</h2>
<p>The <code class="inline"><span class="w">Devel::SmallProf</span></code>
profiler examines the runtime of your Perl program and
produces a line-by-line listing to show how many times each line was called,
and how long each line took to execute. It is called by supplying the familiar
<code class="inline">-d</code>
flag to Perl at runtime.</p>
<pre class="verbatim"><ol><li> $> perl -d:SmallProf wordmatch -f perl5db.pl</li><li></li><li> <...multiple lines snipped...></li><li></li><li> wordmatch report for perl5db.pl:</li><li> lines in file: 9428</li><li> words in file: 50243</li><li> words with special (non-word) characters: 20480</li><li> words with only special (non-word) characters: 7790</li><li> words with only consonants: 4801</li><li> words with only capital letters: 1316</li><li> words with only vowels: 1701</li></ol></pre><p><code class="inline"><span class="w">Devel::SmallProf</span></code>
writes it's output into a file called <i>smallprof.out</i>, by
default. The format of the file looks like this:</p>
<pre class="verbatim"><ol><li> <num> <time> <ctime> <line>:<text></li></ol></pre><p>When the program has terminated, the output may be examined and sorted using
any standard text filtering utilities. Something like the following may be
sufficient:</p>
<pre class="verbatim"><ol><li> $> cat smallprof.out | grep \d*: | sort -k3 | tac | head -n20</li><li></li><li> 251215 1.65674 7.68000 75: if ( $word =~ /($regex)/ ) {</li><li> 251215 0.03264 4.40000 79: debug("word: $i_wd ".($has ? 'matches' :</li><li> 251215 0.02693 4.10000 81: return $has;</li><li> 260643 0.02841 4.07000 128: if ( $debug ) {</li><li> 260643 0.02601 4.04000 126: my $message = shift;</li><li> 251215 0.02641 3.91000 73: my $has = 0;</li><li> 251215 0.03311 3.71000 70: my $i_wd = shift;</li><li> 251215 0.02699 3.69000 72: my $regex = shift;</li><li> 251215 0.02766 3.68000 71: my $word = shift;</li><li> 50243 0.59726 1.00000 59: $count{$i_LINES}{cons} =</li><li> 50243 0.48175 0.92000 61: $count{$i_LINES}{spec} =</li><li> 50243 0.00644 0.89000 56: my $i_cons = matches($i_word, $word,</li><li> 50243 0.48837 0.88000 63: $count{$i_LINES}{caps} =</li><li> 50243 0.00516 0.88000 58: my $i_caps = matches($i_word, $word, '^[(A-</li><li> 50243 0.00631 0.81000 54: my $i_spec = matches($i_word, $word, '[^a-</li><li> 50243 0.00496 0.80000 57: my $i_vows = matches($i_word, $word,</li><li> 50243 0.00688 0.80000 53: $i_word++;</li><li> 50243 0.48469 0.79000 62: $count{$i_LINES}{only} =</li><li> 50243 0.48928 0.77000 60: $count{$i_LINES}{vows} =</li><li> 50243 0.00683 0.75000 55: my $i_only = matches($i_word, $word, '^[^a-</li></ol></pre><p>You can immediately see a slightly different focus to the subroutine profiling
modules, and we start to see exactly which line of code is taking the most
time. That regex line is looking a bit suspicious, for example. Remember that
these tools are supposed to be used together, there is no single best way to
profile your code, you need to use the best tools for the job.</p>
<p>See also <code class="inline"><span class="w">Apache::SmallProf</span></code>
which hooks <code class="inline"><span class="w">Devel::SmallProf</span></code>
into <code class="inline"><span class="w">mod_perl</span></code>
.</p>
<a name="Devel%3a%3aFastProf"></a><h2>Devel::FastProf</h2>
<p><code class="inline"><span class="w">Devel::FastProf</span></code>
is another Perl line profiler. This was written with a view
to getting a faster line profiler, than is possible with for example
<code class="inline"><span class="w">Devel::SmallProf</span></code>
, because it's written in <code class="inline"><span class="w">C</span></code>
. To use <code class="inline"><span class="w">Devel::FastProf</span></code>
,
supply the <code class="inline">-d</code>
argument to Perl:</p>
<pre class="verbatim"><ol><li> $> perl -d:FastProf wordmatch -f perl5db.pl</li><li></li><li> <...multiple lines snipped...></li><li></li><li> wordmatch report for perl5db.pl:</li><li> lines in file: 9428</li><li> words in file: 50243</li><li> words with special (non-word) characters: 20480</li><li> words with only special (non-word) characters: 7790</li><li> words with only consonants: 4801</li><li> words with only capital letters: 1316</li><li> words with only vowels: 1701</li></ol></pre><p><code class="inline"><span class="w">Devel::FastProf</span></code>
writes statistics to the file <i>fastprof.out</i> in the current
directory. The output file, which can be specified, can be interpreted by using
the <code class="inline"><span class="w">fprofpp</span></code>
command-line program.</p>
<pre class="verbatim"><ol><li> $> fprofpp | head -n20</li><li></li><li> # fprofpp output format is:</li><li> # filename:line time count: source</li><li> wordmatch:75 3.93338 251215: if ( $word =~ /($regex)/ ) {</li><li> wordmatch:79 1.77774 251215: debug("word: $i_wd ".($has ? 'matches' : 'does not match')." chars: /$regex/");</li><li> wordmatch:81 1.47604 251215: return $has;</li><li> wordmatch:126 1.43441 260643: my $message = shift;</li><li> wordmatch:128 1.42156 260643: if ( $debug ) {</li><li> wordmatch:70 1.36824 251215: my $i_wd = shift;</li><li> wordmatch:71 1.36739 251215: my $word = shift;</li><li> wordmatch:72 1.35939 251215: my $regex = shift;</li></ol></pre><p>Straightaway we can see that the number of times each line has been called is
identical to the <code class="inline"><span class="w">Devel::SmallProf</span></code>
output, and the sequence is only very
slightly different based on the ordering of the amount of time each line took
to execute, <code class="inline">if ( $debug ) { </code> and <code class="inline"><a class="l_k" href="functions/my.html">my</a> <span class="i">$message</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></code>
, for example. The
differences in the actual times recorded might be in the algorithm used
internally, or it could be due to system resource limitations or contention.</p>
<p>See also the <a href="http://search.cpan.org/perldoc/DBIx::Profile">DBIx::Profile</a> which will profile database queries running
under the <code class="inline"><span class="w">DBIx::</span>*</code>
namespace.</p>
<a name="Devel%3a%3aNYTProf"></a><h2>Devel::NYTProf</h2>
<p><code class="inline"><span class="w">Devel::NYTProf</span></code>
is the <b>next generation</b> of Perl code profiler, fixing many
shortcomings in other tools and implementing many cool features. First of all it
can be used as either a <i>line</i> profiler, a <i>block</i> or a <i>subroutine</i>
profiler, all at once. It can also use sub-microsecond (100ns) resolution on
systems which provide <code class="inline"><span class="i">clock_gettime</span><span class="s">(</span><span class="s">)</span></code>
. It can be started and stopped even
by the program being profiled. It's a one-line entry to profile <code class="inline"><span class="w">mod_perl</span></code>
applications. It's written in <code class="inline"><span class="w">c</span></code>
and is probably the fastest profiler
available for Perl. The list of coolness just goes on. Enough of that, let's
see how to it works - just use the familiar <code class="inline">-d</code>
switch to plug it in and run
the code.</p>
<pre class="verbatim"><ol><li> $> perl -d:NYTProf wordmatch -f perl5db.pl</li><li></li><li> wordmatch report for perl5db.pl:</li><li> lines in file: 9427</li><li> words in file: 50243</li><li> words with special (non-word) characters: 20480</li><li> words with only special (non-word) characters: 7790</li><li> words with only consonants: 4801</li><li> words with only capital letters: 1316</li><li> words with only vowels: 1701</li></ol></pre><p><code class="inline"><span class="w">NYTProf</span></code>
will generate a report database into the file <i>nytprof.out</i> by
default. Human readable reports can be generated from here by using the
supplied <code class="inline"><span class="w">nytprofhtml</span></code>
(HTML output) and <code class="inline"><span class="w">nytprofcsv</span></code>
(CSV output) programs.
We've used the Unix system <code class="inline"><span class="w">html2text</span></code>
utility to convert the
<i>nytprof/index.html</i> file for convenience here.</p>
<pre class="verbatim"><ol><li> $> html2text nytprof/index.html</li><li></li><li> Performance Profile Index</li><li> For wordmatch</li><li> Run on Fri Sep 26 13:46:39 2008</li><li> Reported on Fri Sep 26 13:47:23 2008</li><li></li><li> Top 15 Subroutines -- ordered by exclusive time</li><li> |Calls |P |F |Inclusive|Exclusive|Subroutine |</li><li> | | | |Time |Time | |</li><li> |251215|5 |1 |13.09263 |10.47692 |main:: |matches |</li><li> |260642|2 |1 |2.71199 |2.71199 |main:: |debug |</li><li> |1 |1 |1 |0.21404 |0.21404 |main:: |report |</li><li> |2 |2 |2 |0.00511 |0.00511 |XSLoader:: |load (xsub) |</li><li> |14 |14|7 |0.00304 |0.00298 |Exporter:: |import |</li><li> |3 |1 |1 |0.00265 |0.00254 |Exporter:: |as_heavy |</li><li> |10 |10|4 |0.00140 |0.00140 |vars:: |import |</li><li> |13 |13|1 |0.00129 |0.00109 |constant:: |import |</li><li> |1 |1 |1 |0.00360 |0.00096 |FileHandle:: |import |</li><li> |3 |3 |3 |0.00086 |0.00074 |warnings::register::|import |</li><li> |9 |3 |1 |0.00036 |0.00036 |strict:: |bits |</li><li> |13 |13|13|0.00032 |0.00029 |strict:: |import |</li><li> |2 |2 |2 |0.00020 |0.00020 |warnings:: |import |</li><li> |2 |1 |1 |0.00020 |0.00020 |Getopt::Long:: |ParseOptionSpec|</li><li> |7 |7 |6 |0.00043 |0.00020 |strict:: |unimport |</li><li></li><li> For more information see the full list of 189 subroutines.</li></ol></pre><p>The first part of the report already shows the critical information regarding
which subroutines are using the most time. The next gives some statistics
about the source files profiled.</p>
<pre class="verbatim"><ol><li> Source Code Files -- ordered by exclusive time then name</li><li> |Stmts |Exclusive|Avg. |Reports |Source File |</li><li> | |Time | | | |</li><li> |2699761|15.66654 |6e-06 |line . block . sub|wordmatch |</li><li> |35 |0.02187 |0.00062|line . block . sub|IO/Handle.pm |</li><li> |274 |0.01525 |0.00006|line . block . sub|Getopt/Long.pm |</li><li> |20 |0.00585 |0.00029|line . block . sub|Fcntl.pm |</li><li> |128 |0.00340 |0.00003|line . block . sub|Exporter/Heavy.pm |</li><li> |42 |0.00332 |0.00008|line . block . sub|IO/File.pm |</li><li> |261 |0.00308 |0.00001|line . block . sub|Exporter.pm |</li><li> |323 |0.00248 |8e-06 |line . block . sub|constant.pm |</li><li> |12 |0.00246 |0.00021|line . block . sub|File/Spec/Unix.pm |</li><li> |191 |0.00240 |0.00001|line . block . sub|vars.pm |</li><li> |77 |0.00201 |0.00003|line . block . sub|FileHandle.pm |</li><li> |12 |0.00198 |0.00016|line . block . sub|Carp.pm |</li><li> |14 |0.00175 |0.00013|line . block . sub|Symbol.pm |</li><li> |15 |0.00130 |0.00009|line . block . sub|IO.pm |</li><li> |22 |0.00120 |0.00005|line . block . sub|IO/Seekable.pm |</li><li> |198 |0.00085 |4e-06 |line . block . sub|warnings/register.pm|</li><li> |114 |0.00080 |7e-06 |line . block . sub|strict.pm |</li><li> |47 |0.00068 |0.00001|line . block . sub|warnings.pm |</li><li> |27 |0.00054 |0.00002|line . block . sub|overload.pm |</li><li> |9 |0.00047 |0.00005|line . block . sub|SelectSaver.pm |</li><li> |13 |0.00045 |0.00003|line . block . sub|File/Spec.pm |</li><li> |2701595|15.73869 | |Total |</li><li> |128647 |0.74946 | |Average |</li><li> | |0.00201 |0.00003|Median |</li><li> | |0.00121 |0.00003|Deviation |</li><li></li><li> Report produced by the NYTProf 2.03 Perl profiler, developed by Tim Bunce and</li><li> Adam Kaplan.</li></ol></pre><p>At this point, if you're using the <i>html</i> report, you can click through the
various links to bore down into each subroutine and each line of code. Because
we're using the text reporting here, and there's a whole directory full of
reports built for each source file, we'll just display a part of the
corresponding <i>wordmatch-line.html</i> file, sufficient to give an idea of the
sort of output you can expect from this cool tool.</p>
<pre class="verbatim"><ol><li> $> html2text nytprof/wordmatch-line.html</li><li></li><li> Performance Profile -- -block view-.-line view-.-sub view-</li><li> For wordmatch</li><li> Run on Fri Sep 26 13:46:39 2008</li><li> Reported on Fri Sep 26 13:47:22 2008</li><li></li><li> File wordmatch</li><li></li><li> Subroutines -- ordered by exclusive time</li><li> |Calls |P|F|Inclusive|Exclusive|Subroutine |</li><li> | | | |Time |Time | |</li><li> |251215|5|1|13.09263 |10.47692 |main::|matches|</li><li> |260642|2|1|2.71199 |2.71199 |main::|debug |</li><li> |1 |1|1|0.21404 |0.21404 |main::|report |</li><li> |0 |0|0|0 |0 |main::|BEGIN |</li><li></li><li></li><li> |Line|Stmts.|Exclusive|Avg. |Code |</li><li> | | |Time | | |</li><li> |1 | | | |#!/usr/bin/perl |</li><li> |2 | | | | |</li><li> | | | | |use strict; |</li><li> |3 |3 |0.00086 |0.00029|# spent 0.00003s making 1 calls to strict:: |</li><li> | | | | |import |</li><li> | | | | |use warnings; |</li><li> |4 |3 |0.01563 |0.00521|# spent 0.00012s making 1 calls to warnings:: |</li><li> | | | | |import |</li><li> |5 | | | | |</li><li> |6 | | | |=head1 NAME |</li><li> |7 | | | | |</li><li> |8 | | | |filewords - word analysis of input file |</li><li> <...snip...></li><li> |62 |1 |0.00445 |0.00445|print report( %count ); |</li><li> | | | | |# spent 0.21404s making 1 calls to main::report|</li><li> |63 | | | | |</li><li> | | | | |# spent 23.56955s (10.47692+2.61571) within |</li><li> | | | | |main::matches which was called 251215 times, |</li><li> | | | | |avg 0.00005s/call: # 50243 times |</li><li> | | | | |(2.12134+0.51939s) at line 57 of wordmatch, avg|</li><li> | | | | |0.00005s/call # 50243 times (2.17735+0.54550s) |</li><li> |64 | | | |at line 56 of wordmatch, avg 0.00005s/call # |</li><li> | | | | |50243 times (2.10992+0.51797s) at line 58 of |</li><li> | | | | |wordmatch, avg 0.00005s/call # 50243 times |</li><li> | | | | |(2.12696+0.51598s) at line 55 of wordmatch, avg|</li><li> | | | | |0.00005s/call # 50243 times (1.94134+0.51687s) |</li><li> | | | | |at line 54 of wordmatch, avg 0.00005s/call |</li><li> | | | | |sub matches { |</li><li> <...snip...></li><li> |102 | | | | |</li><li> | | | | |# spent 2.71199s within main::debug which was |</li><li> | | | | |called 260642 times, avg 0.00001s/call: # |</li><li> | | | | |251215 times (2.61571+0s) by main::matches at |</li><li> |103 | | | |line 74 of wordmatch, avg 0.00001s/call # 9427 |</li><li> | | | | |times (0.09628+0s) at line 50 of wordmatch, avg|</li><li> | | | | |0.00001s/call |</li><li> | | | | |sub debug { |</li><li> |104 |260642|0.58496 |2e-06 |my $message = shift; |</li><li> |105 | | | | |</li><li> |106 |260642|1.09917 |4e-06 |if ( $debug ) { |</li><li> |107 | | | |print STDERR "DBG: $message\n"; |</li><li> |108 | | | |} |</li><li> |109 | | | |} |</li><li> |110 | | | | |</li><li> |111 |1 |0.01501 |0.01501|exit 0; |</li><li> |112 | | | | |</li></ol></pre><p>Oodles of very useful information in there - this seems to be the way forward.</p>
<p>See also <code class="inline"><span class="w">Devel::NYTProf::Apache</span></code>
which hooks <code class="inline"><span class="w">Devel::NYTProf</span></code>
into <code class="inline"><span class="w">mod_perl</span></code>
.</p>
<a name="SORTING"></a><h1>SORTING</h1>
<p>Perl modules are not the only tools a performance analyst has at their
disposal, system tools like <code class="inline"><a class="l_k" href="functions/time.html">time</a></code> should not be overlooked as the next
example shows, where we take a quick look at sorting. Many books, theses and
articles, have been written about efficient sorting algorithms, and this is not
the place to repeat such work, there's several good sorting modules which
deserve taking a look at too: <code class="inline"><span class="w">Sort::Maker</span></code>
, <code class="inline"><span class="w">Sort::Key</span></code>
spring to mind.
However, it's still possible to make some observations on certain Perl specific
interpretations on issues relating to sorting data sets and give an example or
two with regard to how sorting large data volumes can effect performance.
Firstly, an often overlooked point when sorting large amounts of data, one can
attempt to reduce the data set to be dealt with and in many cases <code class="inline"><a class="l_k" href="functions/grep.html">grep()</a></code> can
be quite useful as a simple filter:</p>
<pre class="verbatim"><ol><li> <span class="i">@data</span> = <a class="l_k" href="functions/sort.html">sort</a> <a class="l_k" href="functions/grep.html">grep</a> <span class="s">{</span> <span class="q">/$filter/</span> <span class="s">}</span> <span class="i">@incoming</span></li></ol></pre><p>A command such as this can vastly reduce the volume of material to actually
sort through in the first place, and should not be too lightly disregarded
purely on the basis of its simplicity. The <code class="inline"><span class="w">KISS</span></code>
principle is too often
overlooked - the next example uses the simple system <code class="inline"><a class="l_k" href="functions/time.html">time</a></code> utility to
demonstrate. Let's take a look at an actual example of sorting the contents of
a large file, an apache logfile would do. This one has over a quarter of a
million lines, is 50M in size, and a snippet of it looks like this:</p>
<p># logfile</p>
<pre class="verbatim"><ol><li> 188.209-65-87.adsl-dyn.isp.belgacom.be - - [08/Feb/2007:12:57:16 +0000] "GET /favicon.ico HTTP/1.1" 404 209 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)"</li><li> 188.209-65-87.adsl-dyn.isp.belgacom.be - - [08/Feb/2007:12:57:16 +0000] "GET /favicon.ico HTTP/1.1" 404 209 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)"</li><li> 151.56.71.198 - - [08/Feb/2007:12:57:41 +0000] "GET /suse-on-vaio.html HTTP/1.1" 200 2858 "http://www.linux-on-laptops.com/sony.html" "Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8.1.1) Gecko/20061204 Firefox/2.0.0.1"</li><li> 151.56.71.198 - - [08/Feb/2007:12:57:42 +0000] "GET /data/css HTTP/1.1" 404 206 "http://www.rfi.net/suse-on-vaio.html" "Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8.1.1) Gecko/20061204 Firefox/2.0.0.1"</li><li> 151.56.71.198 - - [08/Feb/2007:12:57:43 +0000] "GET /favicon.ico HTTP/1.1" 404 209 "-" "Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8.1.1) Gecko/20061204 Firefox/2.0.0.1"</li><li> 217.113.68.60 - - [08/Feb/2007:13:02:15 +0000] "GET / HTTP/1.1" 304 - "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)"</li><li> 217.113.68.60 - - [08/Feb/2007:13:02:16 +0000] "GET /data/css HTTP/1.1" 404 206 "http://www.rfi.net/" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)"</li><li> debora.to.isac.cnr.it - - [08/Feb/2007:13:03:58 +0000] "GET /suse-on-vaio.html HTTP/1.1" 200 2858 "http://www.linux-on-laptops.com/sony.html" "Mozilla/5.0 (compatible; Konqueror/3.4; Linux) KHTML/3.4.0 (like Gecko)"</li><li> debora.to.isac.cnr.it - - [08/Feb/2007:13:03:58 +0000] "GET /data/css HTTP/1.1" 404 206 "http://www.rfi.net/suse-on-vaio.html" "Mozilla/5.0 (compatible; Konqueror/3.4; Linux) KHTML/3.4.0 (like Gecko)"</li><li> debora.to.isac.cnr.it - - [08/Feb/2007:13:03:58 +0000] "GET /favicon.ico HTTP/1.1" 404 209 "-" "Mozilla/5.0 (compatible; Konqueror/3.4; Linux) KHTML/3.4.0 (like Gecko)"</li><li> 195.24.196.99 - - [08/Feb/2007:13:26:48 +0000] "GET / HTTP/1.0" 200 3309 "-" "Mozilla/5.0 (Windows; U; Windows NT 5.1; fr; rv:1.8.0.9) Gecko/20061206 Firefox/1.5.0.9"</li><li> 195.24.196.99 - - [08/Feb/2007:13:26:58 +0000] "GET /data/css HTTP/1.0" 404 206 "http://www.rfi.net/" "Mozilla/5.0 (Windows; U; Windows NT 5.1; fr; rv:1.8.0.9) Gecko/20061206 Firefox/1.5.0.9"</li><li> 195.24.196.99 - - [08/Feb/2007:13:26:59 +0000] "GET /favicon.ico HTTP/1.0" 404 209 "-" "Mozilla/5.0 (Windows; U; Windows NT 5.1; fr; rv:1.8.0.9) Gecko/20061206 Firefox/1.5.0.9"</li><li> crawl1.cosmixcorp.com - - [08/Feb/2007:13:27:57 +0000] "GET /robots.txt HTTP/1.0" 200 179 "-" "voyager/1.0"</li><li> crawl1.cosmixcorp.com - - [08/Feb/2007:13:28:25 +0000] "GET /links.html HTTP/1.0" 200 3413 "-" "voyager/1.0"</li><li> fhm226.internetdsl.tpnet.pl - - [08/Feb/2007:13:37:32 +0000] "GET /suse-on-vaio.html HTTP/1.1" 200 2858 "http://www.linux-on-laptops.com/sony.html" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)"</li><li> fhm226.internetdsl.tpnet.pl - - [08/Feb/2007:13:37:34 +0000] "GET /data/css HTTP/1.1" 404 206 "http://www.rfi.net/suse-on-vaio.html" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)"</li><li> 80.247.140.134 - - [08/Feb/2007:13:57:35 +0000] "GET / HTTP/1.1" 200 3309 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; .NET CLR 1.1.4322)"</li><li> 80.247.140.134 - - [08/Feb/2007:13:57:37 +0000] "GET /data/css HTTP/1.1" 404 206 "http://www.rfi.net" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; .NET CLR 1.1.4322)"</li><li> pop.compuscan.co.za - - [08/Feb/2007:14:10:43 +0000] "GET / HTTP/1.1" 200 3309 "-" "www.clamav.net"</li><li> livebot-207-46-98-57.search.live.com - - [08/Feb/2007:14:12:04 +0000] "GET /robots.txt HTTP/1.0" 200 179 "-" "msnbot/1.0 (+http://search.msn.com/msnbot.htm)"</li><li> livebot-207-46-98-57.search.live.com - - [08/Feb/2007:14:12:04 +0000] "GET /html/oracle.html HTTP/1.0" 404 214 "-" "msnbot/1.0 (+http://search.msn.com/msnbot.htm)"</li><li> dslb-088-064-005-154.pools.arcor-ip.net - - [08/Feb/2007:14:12:15 +0000] "GET / HTTP/1.1" 200 3309 "-" "www.clamav.net"</li><li> 196.201.92.41 - - [08/Feb/2007:14:15:01 +0000] "GET / HTTP/1.1" 200 3309 "-" "MOT-L7/08.B7.DCR MIB/2.2.1 Profile/MIDP-2.0 Configuration/CLDC-1.1"</li></ol></pre><p>The specific task here is to sort the 286,525 lines of this file by Response
Code, Query, Browser, Referring Url, and lastly Date. One solution might be to
use the following code, which iterates over the files given on the
command-line.</p>
<p># sort-apache-log</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl -n</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">@data</span><span class="sc">;</span></li><li></li><li> <span class="j">LINE:</span></li><li> while <span class="s">(</span> <> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$line</span> = <span class="i">$_</span><span class="sc">;</span></li><li> if <span class="s">(</span></li><li> <span class="i">$line</span> =~ <span class="q">m/^(</span></li><li> <span class="q"> ([\w\.\-]+) # client</span></li><li> <span class="q"> \s*-\s*-\s*\[</span></li><li> <span class="q"> ([^]]+) # date</span></li><li> <span class="q"> \]\s*"\w+\s*</span></li><li> <span class="q"> (\S+) # query</span></li><li> <span class="q"> [^"]+"\s*</span></li><li> <span class="q"> (\d+) # status</span></li><li> <span class="q"> \s+\S+\s+"[^"]*"\s+"</span></li><li> <span class="q"> ([^"]*) # browser</span></li><li> <span class="q"> "</span></li><li> <span class="q"> .*</span></li><li> <span class="q"> )$/x</span></li><li> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">@chunks</span> = <a class="l_k" href="functions/split.html">split</a><span class="s">(</span><span class="q">/ +/</span><span class="cm">,</span> <span class="i">$line</span><span class="s">)</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$ip</span> = <span class="i">$1</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$date</span> = <span class="i">$2</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$query</span> = <span class="i">$3</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$status</span> = <span class="i">$4</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$browser</span> = <span class="i">$5</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/push.html">push</a><span class="s">(</span><span class="i">@data</span><span class="cm">,</span> <span class="s">[</span><span class="i">$ip</span><span class="cm">,</span> <span class="i">$date</span><span class="cm">,</span> <span class="i">$query</span><span class="cm">,</span> <span class="i">$status</span><span class="cm">,</span> <span class="i">$browser</span><span class="cm">,</span> <span class="i">$line</span><span class="s">]</span><span class="s">)</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li> <span class="s">}</span></li><li></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">@sorted</span> = <a class="l_k" href="functions/sort.html">sort</a> <span class="s">{</span></li><li> <span class="i">$a</span>->[<span class="n">3</span>] cmp <span class="i">$b</span>->[<span class="n">3</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">2</span>] cmp <span class="i">$b</span>->[<span class="n">2</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">0</span>] cmp <span class="i">$b</span>->[<span class="n">0</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">1</span>] cmp <span class="i">$b</span>->[<span class="n">1</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">4</span>] cmp <span class="i">$b</span>->[<span class="n">4</span>]</li><li> <span class="s">}</span> <span class="i">@data</span><span class="sc">;</span></li><li></li><li> foreach <a class="l_k" href="functions/my.html">my</a> <span class="i">$data</span> <span class="s">(</span> <span class="i">@sorted</span> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/print.html">print</a> <span class="i">$data</span>->[<span class="n">5</span>]<span class="sc">;</span></li><li> <span class="s">}</span></li><li></li><li> <a class="l_k" href="functions/exit.html">exit</a> <span class="n">0</span><span class="sc">;</span></li></ol></pre><p>When running this program, redirect <code class="inline"><span class="w">STDOUT</span></code>
so it is possible to check the
output is correct from following test runs and use the system <code class="inline"><a class="l_k" href="functions/time.html">time</a></code> utility
to check the overall runtime.</p>
<pre class="verbatim"><ol><li> $> time ./sort-apache-log logfile > out-sort</li><li></li><li> real 0m17.371s</li><li> user 0m15.757s</li><li> sys 0m0.592s</li></ol></pre><p>The program took just over 17 wallclock seconds to run. Note the different
values <code class="inline"><a class="l_k" href="functions/time.html">time</a></code> outputs, it's important to always use the same one, and to not
confuse what each one means.</p>
<ul>
<li><a name="Elapsed-Real-Time"></a><b>Elapsed Real Time</b>
<p>The overall, or wallclock, time between when <code class="inline"><a class="l_k" href="functions/time.html">time</a></code> was called, and when it
terminates. The elapsed time includes both user and system times, and time
spent waiting for other users and processes on the system. Inevitably, this is
the most approximate of the measurements given.</p>
</li>
<li><a name="User-CPU-Time"></a><b>User CPU Time</b>
<p>The user time is the amount of time the entire process spent on behalf of the
user on this system executing this program.</p>
</li>
<li><a name="System-CPU-Time"></a><b>System CPU Time</b>
<p>The system time is the amount of time the kernel itself spent executing
routines, or system calls, on behalf of this process user.</p>
</li>
</ul>
<p>Running this same process as a <code class="inline"><span class="w">Schwarzian</span> <span class="w">Transform</span></code>
it is possible to
eliminate the input and output arrays for storing all the data, and work on the
input directly as it arrives too. Otherwise, the code looks fairly similar:</p>
<p># sort-apache-log-schwarzian</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl -n</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/print.html">print</a></li><li></li><li> <a class="l_k" href="functions/map.html">map</a> <span class="i">$_</span>->[<span class="n">0</span>] <span class="cm">=></span></li><li></li><li> <a class="l_k" href="functions/sort.html">sort</a> <span class="s">{</span></li><li> <span class="i">$a</span>->[<span class="n">4</span>] cmp <span class="i">$b</span>->[<span class="n">4</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">3</span>] cmp <span class="i">$b</span>->[<span class="n">3</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">1</span>] cmp <span class="i">$b</span>->[<span class="n">1</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">2</span>] cmp <span class="i">$b</span>->[<span class="n">2</span>]</li><li> ||</li><li> <span class="i">$a</span>->[<span class="n">5</span>] cmp <span class="i">$b</span>->[<span class="n">5</span>]</li><li> <span class="s">}</span></li><li> <a class="l_k" href="functions/map.html">map</a> <span class="s">[</span> <span class="i">$_</span><span class="cm">,</span> <span class="q">m/^(</span></li><li> <span class="q"> ([\w\.\-]+) # client</span></li><li> <span class="q"> \s*-\s*-\s*\[</span></li><li> <span class="q"> ([^]]+) # date</span></li><li> <span class="q"> \]\s*"\w+\s*</span></li><li> <span class="q"> (\S+) # query</span></li><li> <span class="q"> [^"]+"\s*</span></li><li> <span class="q"> (\d+) # status</span></li><li> <span class="q"> \s+\S+\s+"[^"]*"\s+"</span></li><li> <span class="q"> ([^"]*) # browser</span></li><li> <span class="q"> "</span></li><li> <span class="q"> .*</span></li><li> <span class="q"> )$/xo</span> <span class="s">]</span></li><li></li><li> <span class="cm">=></span> <><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/exit.html">exit</a> <span class="n">0</span><span class="sc">;</span></li></ol></pre><p>Run the new code against the same logfile, as above, to check the new time.</p>
<pre class="verbatim"><ol><li> $> time ./sort-apache-log-schwarzian logfile > out-schwarz</li><li></li><li> real 0m9.664s</li><li> user 0m8.873s</li><li> sys 0m0.704s</li></ol></pre><p>The time has been cut in half, which is a respectable speed improvement by any
standard. Naturally, it is important to check the output is consistent with
the first program run, this is where the Unix system <code class="inline"><span class="w">cksum</span></code>
utility comes in.</p>
<pre class="verbatim"><ol><li> $> cksum out-sort out-schwarz</li><li> 3044173777 52029194 out-sort</li><li> 3044173777 52029194 out-schwarz</li></ol></pre><p>BTW. Beware too of pressure from managers who see you speed a program up by 50%
of the runtime once, only to get a request one month later to do the same again
(true story) - you'll just have to point out your only human, even if you are a
Perl programmer, and you'll see what you can do...</p>
<a name="LOGGING"></a><h1>LOGGING</h1>
<p>An essential part of any good development process is appropriate error handling
with appropriately informative messages, however there exists a school of
thought which suggests that log files should be <i>chatty</i>, as if the chain of
unbroken output somehow ensures the survival of the program. If speed is in
any way an issue, this approach is wrong.</p>
<p>A common sight is code which looks something like this:</p>
<pre class="verbatim"><ol><li> <span class="w">logger</span><span class="w">->debug</span><span class="s">(</span> <span class="q">"A logging message via process-id: $$ INC: "</span> . <span class="i">Dumper</span><span class="s">(</span>\<span class="i">%INC</span><span class="s">)</span> <span class="s">)</span></li></ol></pre><p>The problem is that this code will always be parsed and executed, even when the
debug level set in the logging configuration file is zero. Once the debug()
subroutine has been entered, and the internal <code class="inline"><span class="i">$debug</span></code>
variable confirmed to
be zero, for example, the message which has been sent in will be discarded and
the program will continue. In the example given though, the \%INC hash will
already have been dumped, and the message string constructed, all of which work
could be bypassed by a debug variable at the statement level, like this:</p>
<pre class="verbatim"><ol><li> <span class="w">logger</span><span class="w">->debug</span><span class="s">(</span> <span class="q">"A logging message via process-id: $$ INC: "</span> . <span class="i">Dumper</span><span class="s">(</span>\<span class="i">%INC</span><span class="s">)</span> <span class="s">)</span> if <span class="i">$DEBUG</span><span class="sc">;</span></li></ol></pre><p>This effect can be demonstrated by setting up a test script with both forms,
including a <code class="inline"><span class="i">debug</span><span class="s">(</span><span class="s">)</span></code>
subroutine to emulate typical <code class="inline"><span class="i">logger</span><span class="s">(</span><span class="s">)</span></code>
functionality.</p>
<p># ifdebug</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Benchmark</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Data::Dumper</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$DEBUG</span> = <span class="n">0</span><span class="sc">;</span></li><li></li><li><a name="debug"></a> sub <span class="m">debug</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$msg</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></li><li></li><li> if <span class="s">(</span> <span class="i">$DEBUG</span> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/print.html">print</a> <span class="q">"DEBUG: $msg\n"</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li> <span class="s">}</span><span class="sc">;</span></li><li></li><li> <span class="i">timethese</span><span class="s">(</span><span class="n">100000</span><span class="cm">,</span> <span class="s">{</span></li><li> <span class="q">'debug'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span></li><li> <span class="i">debug</span><span class="s">(</span> <span class="q">"A $0 logging message via process-id: $$"</span> . <span class="i">Dumper</span><span class="s">(</span>\<span class="i">%INC</span><span class="s">)</span> <span class="s">)</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="q">'ifdebug'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span></li><li> <span class="i">debug</span><span class="s">(</span> <span class="q">"A $0 logging message via process-id: $$"</span> . <span class="i">Dumper</span><span class="s">(</span>\<span class="i">%INC</span><span class="s">)</span> <span class="s">)</span> if <span class="i">$DEBUG</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="s">}</span><span class="s">)</span><span class="sc">;</span></li></ol></pre><p>Let's see what <code class="inline"><span class="w">Benchmark</span></code>
makes of this:</p>
<pre class="verbatim"><ol><li> $> perl ifdebug</li><li> Benchmark: timing 100000 iterations of constant, sub...</li><li> ifdebug: 0 wallclock secs ( 0.01 usr + 0.00 sys = 0.01 CPU) @ 10000000.00/s (n=100000)</li><li> (warning: too few iterations for a reliable count)</li><li> debug: 14 wallclock secs (13.18 usr + 0.04 sys = 13.22 CPU) @ 7564.30/s (n=100000)</li></ol></pre><p>In the one case the code, which does exactly the same thing as far as
outputting any debugging information is concerned, in other words nothing,
takes 14 seconds, and in the other case the code takes one hundredth of a
second. Looks fairly definitive. Use a <code class="inline"><span class="i">$DEBUG</span></code>
variable BEFORE you call the
subroutine, rather than relying on the smart functionality inside it.</p>
<a name="Logging-if-DEBUG-(constant)"></a><h2>Logging if DEBUG (constant)</h2>
<p>It's possible to take the previous idea a little further, by using a compile
time <code class="inline"><span class="w">DEBUG</span></code>
constant.</p>
<p># ifdebug-constant</p>
<pre class="verbatim"><ol><li> <span class="c">#!/usr/bin/perl</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">strict</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">warnings</span><span class="sc">;</span></li><li></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Benchmark</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">Data::Dumper</span><span class="sc">;</span></li><li> <a class="l_k" href="functions/use.html">use</a> <span class="w">constant</span></li><li> <span class="i">DEBUG</span> <span class="cm">=></span> <span class="n">0</span></li><li> <span class="sc">;</span></li><li></li><li><a name="debug"></a> sub <span class="m">debug</span> <span class="s">{</span></li><li> if <span class="s">(</span> <span class="i">DEBUG</span> <span class="s">)</span> <span class="s">{</span></li><li> <a class="l_k" href="functions/my.html">my</a> <span class="i">$msg</span> = <a class="l_k" href="functions/shift.html">shift</a><span class="sc">;</span></li><li> <a class="l_k" href="functions/print.html">print</a> <span class="q">"DEBUG: $msg\n"</span><span class="sc">;</span></li><li> <span class="s">}</span></li><li> <span class="s">}</span><span class="sc">;</span></li><li></li><li> <span class="i">timethese</span><span class="s">(</span><span class="n">100000</span><span class="cm">,</span> <span class="s">{</span></li><li> <span class="q">'debug'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span></li><li> <span class="i">debug</span><span class="s">(</span> <span class="q">"A $0 logging message via process-id: $$"</span> . <span class="i">Dumper</span><span class="s">(</span>\<span class="i">%INC</span><span class="s">)</span> <span class="s">)</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="q">'constant'</span> <span class="cm">=></span> <a class="l_k" href="functions/sub.html">sub</a> <span class="s">{</span></li><li> <span class="i">debug</span><span class="s">(</span> <span class="q">"A $0 logging message via process-id: $$"</span> . <span class="i">Dumper</span><span class="s">(</span>\<span class="i">%INC</span><span class="s">)</span> <span class="s">)</span> if <span class="i">DEBUG</span></li><li> <span class="s">}</span><span class="cm">,</span></li><li> <span class="s">}</span><span class="s">)</span><span class="sc">;</span></li></ol></pre><p>Running this program produces the following output:</p>
<pre class="verbatim"><ol><li> $> perl ifdebug-constant</li><li> Benchmark: timing 100000 iterations of constant, sub...</li><li> constant: 0 wallclock secs (-0.00 usr + 0.00 sys = -0.00 CPU) @ -7205759403792793600000.00/s (n=100000)</li><li> (warning: too few iterations for a reliable count)</li><li> sub: 14 wallclock secs (13.09 usr + 0.00 sys = 13.09 CPU) @ 7639.42/s (n=100000)</li></ol></pre><p>The <code class="inline"><span class="w">DEBUG</span></code>
constant wipes the floor with even the <code class="inline"><span class="i">$debug</span></code>
variable,
clocking in at minus zero seconds, and generates a "warning: too few iterations
for a reliable count" message into the bargain. To see what is really going
on, and why we had too few iterations when we thought we asked for 100000, we
can use the very useful <code class="inline"><span class="w">B::Deparse</span></code>
to inspect the new code:</p>
<pre class="verbatim"><ol><li> $> perl -MO=Deparse ifdebug-constant</li><li></li><li> use Benchmark;</li><li> use Data::Dumper;</li><li> use constant ('DEBUG', 0);</li><li> sub debug {</li><li> use warnings;</li><li> use strict 'refs';</li><li> 0;</li><li> }</li><li> use warnings;</li><li> use strict 'refs';</li><li> timethese(100000, {'sub', sub {</li><li> debug "A $0 logging message via process-id: $$" . Dumper(\%INC);</li><li> }</li><li> , 'constant', sub {</li><li> 0;</li><li> }</li><li> });</li><li> ifdebug-constant syntax OK</li></ol></pre><p>The output shows the constant() subroutine we're testing being replaced with
the value of the <code class="inline"><span class="w">DEBUG</span></code>
constant: zero. The line to be tested has been
completely optimized away, and you can't get much more efficient than that.</p>
<a name="POSTSCRIPT"></a><h1>POSTSCRIPT</h1>
<p>This document has provided several way to go about identifying hot-spots, and
checking whether any modifications have improved the runtime of the code.</p>
<p>As a final thought, remember that it's not (at the time of writing) possible to
produce a useful program which will run in zero or negative time and this basic
principle can be written as: <i>useful programs are slow</i> by their very
definition. It is of course possible to write a nearly instantaneous program,
but it's not going to do very much, here's a very efficient one:</p>
<pre class="verbatim"><ol><li> $> perl -e 0</li></ol></pre><p>Optimizing that any further is a job for <code class="inline"><span class="w">p5p</span></code>
.</p>
<a name="SEE-ALSO"></a><h1>SEE ALSO</h1>
<p>Further reading can be found using the modules and links below.</p>
<a name="PERLDOCS"></a><h2>PERLDOCS</h2>
<p>For example: <code class="inline"><span class="w">perldoc</span> -f <a class="l_k" href="functions/sort.html">sort</a></code>
.</p>
<p><a href="perlfaq4.html">perlfaq4</a>.</p>
<p><a href="perlfork.html">perlfork</a>, <a href="perlfunc.html">perlfunc</a>, <a href="perlretut.html">perlretut</a>, <a href="perlthrtut.html">perlthrtut</a>.</p>
<p><a href="threads.html">threads</a>.</p>
<a name="MAN-PAGES"></a><h2>MAN PAGES</h2>
<p><code class="inline"><a class="l_k" href="functions/time.html">time</a></code>.</p>
<a name="MODULES"></a><h2>MODULES</h2>
<p>It's not possible to individually showcase all the performance related code for
Perl here, naturally, but here's a short list of modules from the CPAN which
deserve further attention.</p>
<pre class="verbatim"><ol><li> <span class="w">Apache::DProf</span></li><li> <span class="w">Apache::SmallProf</span></li><li> <span class="w">Benchmark</span></li><li> <span class="w">DBIx::Profile</span></li><li> <span class="w">Devel::AutoProfiler</span></li><li> <span class="w">Devel::DProf</span></li><li> <span class="w">Devel::DProfLB</span></li><li> <span class="w">Devel::FastProf</span></li><li> <span class="w">Devel::GraphVizProf</span></li><li> <span class="w">Devel::NYTProf</span></li><li> <span class="w">Devel::NYTProf::Apache</span></li><li> <span class="w">Devel::Profiler</span></li><li> <span class="w">Devel::Profile</span></li><li> <span class="w">Devel::Profit</span></li><li> <span class="w">Devel::SmallProf</span></li><li> <span class="w">Devel::WxProf</span></li><li> <span class="w">POE::Devel::Profiler</span></li><li> <span class="w">Sort::Key</span></li><li> <span class="w">Sort::Maker</span></li></ol></pre><a name="URLS"></a><h2>URLS</h2>
<p>Very useful online reference material:</p>
<pre class="verbatim"><ol><li> http://www.ccl4.org/~nick/P/Fast_Enough/</li><li></li><li> http://www-128.ibm.com/developerworks/library/l-optperl.html</li><li></li><li> http://perlbuzz.com/2007/11/bind-output-variables-in-dbi-for-speed-and-safety.html</li><li></li><li> http://en.wikipedia.org/wiki/Performance_analysis</li><li></li><li> http://apache.perl.org/docs/1.0/guide/performance.html</li><li></li><li> http://perlgolf.sourceforge.net/</li><li></li><li> http://www.sysarch.com/Perl/sort_paper.html</li></ol></pre><a name="AUTHOR"></a><h1>AUTHOR</h1>
<p>Richard Foley <richard.foley@rfi.net> Copyright (c) 2008</p>
<div id="page_index" class="hud_container">
<div id="page_index_header" class="hud_header">
<div id="page_index_close" class="hud_close"><a href="#" onClick="pageIndex.hide();return false;"></a></div>
<div id="page_index_title" class="hud_title"><span class="hud_span_top">Page index</span></div>
<div id="page_index_topright" class="hud_topright"></div>
</div>
<div id="page_index_content" class="hud_content">
<ul><li><a href="#NAME">NAME</a><li><a href="#DESCRIPTION">DESCRIPTION</a><li><a href="#OVERVIEW">OVERVIEW</a><ul><li><a href="#ONE-STEP-SIDEWAYS">ONE STEP SIDEWAYS</a><li><a href="#ONE-STEP-FORWARD">ONE STEP FORWARD</a><li><a href="#ANOTHER-STEP-SIDEWAYS">ANOTHER STEP SIDEWAYS</a></ul><li><a href="#GENERAL-GUIDELINES">GENERAL GUIDELINES</a><li><a href="#BENCHMARKS">BENCHMARKS</a><ul><li><a href="#Assigning-and-Dereferencing-Variables.">Assigning and Dereferencing Variables.</a><li><a href="#Search-and-replace-or-tr">Search and replace or tr</a></ul><li><a href="#PROFILING-TOOLS">PROFILING TOOLS</a><ul><li><a href="#Devel%3a%3aDProf">Devel::DProf</a><li><a href="#Devel%3a%3aProfiler">Devel::Profiler</a><li><a href="#Devel%3a%3aSmallProf">Devel::SmallProf</a><li><a href="#Devel%3a%3aFastProf">Devel::FastProf</a><li><a href="#Devel%3a%3aNYTProf">Devel::NYTProf</a></ul><li><a href="#SORTING">SORTING</a><li><a href="#LOGGING">LOGGING</a><ul><li><a href="#Logging-if-DEBUG-(constant)">Logging if DEBUG (constant)</a></ul><li><a href="#POSTSCRIPT">POSTSCRIPT</a><li><a href="#SEE-ALSO">SEE ALSO</a><ul><li><a href="#PERLDOCS">PERLDOCS</a><li><a href="#MAN-PAGES">MAN PAGES</a><li><a href="#MODULES">MODULES</a><li><a href="#URLS">URLS</a></ul><li><a href="#AUTHOR">AUTHOR</a></ul>
</div>
<div id="page_index_footer" class="hud_footer">
<div id="page_index_bottomleft" class="hud_bottomleft"></div>
<div id="page_index_bottom" class="hud_bottom"><span class="hud_span_bottom"></span></div>
<div id="page_index_resize" class="hud_resize"></div>
</div>
</div>
</div>
<div id="content_footer">
</div>
</div>
<div class="clear"></div>
</div>
<div id="footer">
<div id="footer_content">
<div id="footer_strapline">
perldoc.perl.org - Official documentation for the Perl programming language
</div>
<div id="footer_links">
<div id="address">
<p class="name">Contact details</p>
<p class="address">
Site maintained by <a href="mailto:jj@jonallen.info">Jon Allen (JJ)</a><br>
</p>
<p class="contact">
Documentation maintained by the <a href="http://lists.cpan.org/showlist.cgi?name=perl5-porters">Perl 5 Porters</a>
</p>
</div>
<ul class="f1">
<li>Manual
<ul class="f2">
<li><a href="index-overview.html">Overview</a>
<li><a href="index-tutorials.html">Tutorials</a>
<li><a href="index-faq.html">FAQs</a>
<li><a href="index-history.html">Changes</a>
</ul>
<li>Reference
<ul class="f2">
<li><a href="index-language.html">Language</a>
<li><a href="index-functions.html">Functions</a>
<li><a href="perlop.html">Operators</a>
<li><a href="perlvar.html">Variables</a>
</ul>
<li>Modules
<ul class="f2">
<li><a href="index-modules-A.html">Modules</a>
<li><a href="index-pragmas.html">Pragmas</a>
<li><a href="index-utilities.html">Utilities</a>
</ul>
<li>Misc
<ul class="f2">
<li><a href="index-licence.html">License</a>
<li><a href="index-internals.html">Internals</a>
<li><a href="index-platforms.html">Platforms</a>
</ul> </ul>
<div class="clear"></div>
</div>
</div>
<div id="footer_end">
</div>
</div>
</div>
<script language="JavaScript" type="text/javascript" src="static/exploreperl.js"></script>
<script language="JavaScript" src="static/combined-20100403.js" type="text/javascript"></script>
<script language="JavaScript" type="text/javascript">
perldoc.setPath(0);
perldoc.pageName = 'perlperf';
perldoc.pageAddress = 'perlperf.html';
perldoc.contentPage = 1;
explorePerl.render();
explorePerl.addEvents('explore_anchor');
</script>
</body>
</html>
|