Perl Frequently Asked Questions List

August 16, 1995

This article details the contents of the most Frequently Asked Questions in comp.lang.perl.misc, a Usenet newsgroup devoted to the Perl programming language. There are five pieces following this, the general information questions in part1, part2 and part3 and the programming ones in part4 and part5.

If this is your first exposure to Perl, please read this document and the perl(1) man page before asking questions in comp.lang.perl.misc. If you're using v4 perl, that page contains all you need to know (or at least enough to get started). If you're using v5 perl, that page will show you where to look for specific information. When we refer to perlmod(1), it means the "perlmod" man page in section "1" of the manual, just as Foo(3pm), that means it's the "Foo" man page in section "3pm" (perl modules) of the library. The perl install does NOT automatically install the module man pages for you, however.

Hopefully the questions herein are asked enough that considerable net bandwidth can be saved by looking here before asking. Also, hopefully there is enough information contained here that someone who has never heard of Perl can read this and at least have some sort of idea as to what Perl is.

Some questions in this group aren't really about Perl, but rather about system-specific issues. You might also consult the Most Frequently Asked Questions list in comp.unix.questions for answers to this type of question.

The current version of perl is 5.001, perl 5.000 emerged into the world on 16 October, 1994. The previous non-beta version was 4.036 (version 4, patchlevel 36). Many of these questions were written for perl4, however a lot of perl5 information has also been added. Perl5 only features will be clearly marked as such, so as not to cause confusion for those still using perl4. You should upgrade to perl5 as soon as possible though (see below).

This list was initially written, and still hacked upon, by Tom Christiansen*. However, due to his erratic schedule, it is currently maintained by Stephen P Potter*. First person singular pronouns, when not in quoted postings, generally are Tom talking.

This document, and all its parts, are Copyright (c) 1994/1995, Stephen P Potter and Tom Christiansen, All rights reserved. HTML by Tom Christiansen. Permisson to distribute this collection, in part or full, via electronic means (emailed, posted or archived) or printed copy are granted providing that no charges are involved, reasonable attempt is made to use the most current version, and all credits and copyright notices are retained. Requests for other distribution rights, including incorporation in commercial products, such as books, magazine articles, or CD-ROMs should be made to

This FAQ is archived on [] in the file pub/perl/doc/FAQ, as well as on [] in /pub/usenet/comp.lang.perl.*. If you have any suggested additions or corrections to this article, please send them to .

General Information and Availability

1.1) What is Perl?
1.2) What are perl4 and perl5, are there any differences?
1.3) What features does perl5 provide over perl4?
1.4) Where can I get docs on perl5?
1.5) Will perl5 break my perl4 scripts?
1.6) When will Perl stabilize?
1.7) What's the difference between "perl" and "Perl"?
1.8) Is it a perl program or a perl script?
1.9) Is perl difficult to learn?
1.10) Should I program everything in perl?
1.11) How does perl compare with other scripting languages, like Tcl, Python or REXX?
1.12) Where can I get Perl over the Internet (FTP)?
1.13) How can I get Perl via email?
1.14) How can I get Perl via UUCP?
1.15) Are there other ways of getting perl?
1.16) Has perl been ported to machine FOO?
1.17) How do I get perl to compile on Solaris?
1.18) How do I get perl to compile on a NeXT?
1.19) What extensions are available for Perl and where can I get them?
1.20) What is dbperl and where can I get it?
1.21) Which DBM should I use?
1.22) Is there an SNMP aware perl?
1.23) Is there an ISO or ANSI certified version of Perl?

1.1) What is Perl?

Perl is a compiled scripting language written by Larry Wall*.

Here's the beginning of the description from the perl(1) man page:

Perl is an interpreted language optimized for scanning arbitrary text files, extracting information from those text files, and printing reports based on that information. It's also a good language for many system management tasks. The language is intended to be practical (easy to use, efficient, complete) rather than beautiful (tiny, elegant, minimal). It combines (in the author's opinion, anyway) some of the best features of C, sed, awk, and sh, so people familiar with those languages should have little difficulty with it. (Language historians will also note some vestiges of csh, Pascal, and even BASIC-PLUS.) Expression syntax corresponds quite closely to C expression syntax. Unlike most Unix utilities, perl does not arbitrarily limit the size of your data--if you've got the memory, perl can slurp in your whole file as a single string. Recursion is of unlimited depth. And the hash tables used by associative arrays grow as necessary to prevent degraded performance. Perl uses sophisticated pattern matching techniques to scan large amounts of data very quickly. Although optimized for scanning text, perl can also deal with binary data, and can make dbm files look like associative arrays (where dbm is available). Setuid perl scripts are safer than C programs through a dataflow tracing mechanism which prevents many stupid security holes. If you have a problem that would ordinarily use sed or awk or sh, but it exceeds their capabilities or must run a little faster, and you don't want to write the silly thing in C, then perl may be for you. There are also translators to turn your sed and awk scripts into perl scripts. OK, enough hype.

1.2) What are perl4 and perl5, are there any differences?

Perl4 and perl5 are different versions of the language. Perl4 was the previous release, and perl5 is ``Perl: The Next Generation.'' Perl5 is, essentially, a complete rewrite of the perl source code from the ground up. It has been modularized, object oriented, tweaked, trimmed, and optimized until it almost doesn't look like the old code. However, the interface is mostly the same, and compatibility with previous releases is very high.

1.3) What features does perl5 provide over perl4?

If you get the newest source (from any of the main FTP sites), you will find a directory full of man pages (possibly to be installed as section 1p and 3pm) that discuss the differences, new features, old incompatibilies and much more. Here, however, are some highlights as to the new features and old incompatibilities. (Thanks to Tom Christiansen* for this section)

1.4) Where can I get docs on perl5?

The complete perl documentation is available with the Perl distribution, or can be accessed from the following sites. Note that the PerlDoc ps file is 240 pages long!!
Marked Up (HTML) format:                       (Europe)
PostScript:  (Europe)  (Europe)  (Oz)
        unable to access as of 7/15/95
TeXinfo (Emacs) Format:

1.5) Will perl5 break my perl4 scripts?

In general, no. However, certain bad old practices have become highly frowned upon. The following are the most important of the known incompatibilities between perl4 and perl5. See perltrap(1) for more details.

1.6) When will Perl stabilize?

When asked at what point the Perl code would be frozen, Larry answered:
Part of the redesign of Perl is to allow us to more or less freeze the language itself. It won't totally freeze, of course, but I think the rate of change of the core of the language is asymptotically approaching 0. In fact, as time goes on, now that we have an official extension mechanism, some of the things that are currently in the core of the language may move out (transparently) as extensions. This has already happened to dbmopen().

I've also been continuously reminding myself of what Henry Spencer calls ``second system syndrome'', in which everything under the sun gets added, resulting in a colossal kludge, like OS 360. You'll find that the new features in Perl 5 are all pretty minimalistic. The object-oriented features in particular added only one new piece of syntax, a C++-style method call.

The whole idea behind Perl is to be a fast text-processing, system-maintenance, zero-startup time language. If it gets to be so large and complicated that it isn't fast-running and easy to use, it won't be to anyone's benefit.
My motto from the start has been, ``If it ain't broke, don't fix it.'' I've been trying very hard not to remove those features from Perl that make it what it is. At the same time, a lot of streamlining has gone into the syntax. The new yacc file is about half the size of the old one, and the number of official reserved words has been cut by 2/3. All built-in functions have been unified (dualified?) as either list operators or unary operators.
I really like a lot of the features in Perl, but in order for Perl to be useful on a long term basis, those features have to stay put. I bought the Camel book less than a year ago and it sounds like within another year it will be obsolete.
The parts of Perl that the Camel book covers have not changed all that much. Most old scripts still run. Many scripts from Perl version 1.0 still run. We'll certainly be revising the Camel, but the new man pages are split up such that it's pretty easy to ferret out the new info when you want it.

We did break a few misfeatures in going to Perl 5. It seemed like the first and last chance to do so. There's a list of the incompatibilities in the documentation.

Not only is it a lot of work to recompile Perl on 20+ machines periodically, but it's hard to write scripts that are useful in the long term if the guts of the language keep changing. (And if I keep having to buy new books. I keep hearing about new features of Perl 5 that aren't documented in any of the perl 5 documentation that I can find.)
I think you'll find a lot of folks who think that 4.036 has been a pretty stable platform.

Perl 5 is a special case. I've been working on it for years. (This is part of the reason 4.036 has been so stable!) There are many changes, most of them for the better, I hope. I don't expect the transition to be without pain. But that's why I stuck numbered versions out in your bin directory, so that you can upgrade piecemeal if you like. And that's why I made the -w switch warn about many of the incompatibilities.

And overriding all that, I've tried to keep it so that you don't have to know much about the new stuff to use the old stuff. You can upgrade your knowledge piecemeal too.

The extension mechanism is designed to take over most of the evolutionary role from now on. And it's set up so that, if you don't have a particular extension, you know it right up at the front.

Are there any plans to write a Perl compiler? While interpreted Perl is great for many applications, it would also be cool to be able to precompile many scripts. (Yes, I know you can undump things, but undump isn't provided with Perl and I haven't found a copy.) The creation of a perl library and dynamically-loadable modules seems like a step in that direction.
Yes, part of the design of Perl 5 was to make it possible to write a compiler for it. It could even be done as an extension module, I suppose. Anyone looking for a master's thesis topic?

In summary, almost every concern that you might think of has already been (at least) thought about. In a perfect world, every concern could be addressed perfectly. But in this world we just have to slog through.

1.7) What's the difference between ``perl'' and ``Perl''?

    32!  [ ord('p') - ord('P') ]

Larry now uses ``Perl'' to signify the language proper and ``perl'' the implementation of it, i.e. the current interpreter. Hence Tom's quip that ``Nothing but perl can parse Perl.''

On the other hand, the aesthetic value of casewise parallelism in ``awk'', ``sed'', and ``perl'' as much require the lower-case version as ``C'', ``Pascal'', and ``Perl'' require the upper-case version. It's also easier to type ``Perl'' in typeset print than to be constantly switching in Courier. :-)

In other words, it doesn't matter much, especially if all you're doing is hearing someone talk about the language; case is hard to distinguish aurally.

1.8) Is it a perl program or a perl script?

It depends on whether you are talking about the perl binary or something that you wrote using perl. And, actually, even this isn't necessarily true.

``Standard'' UNIX terminology is (roughly) this: programs are compiled into machine code once and run multiple times, scripts are translated (by a program) each time they are used. However, some say that a program is anything written which is executed on a computer system. Larry considers it a program if it is set in stone and you can't change it, whereas if you can go in and hack at it, it's a script. Of course, if you have the source code, that makes just about anything a script. ;)

In general, it probably doesn't really matter. The terms are used interchangeably. If you particularly like one or the other, use it. If you want to call yourself a perl programmer, call them programs. If you want to call yourself a perl scripter, call them scripts. Randal* and I (at least) will call them hacks. (See question 2.10 ;)

Larry says that a script is what you give an actor, but a program is what you give an audience.

1.9) Is perl difficult to learn?

Not at all. Many people find Perl extremely easy to learn. There are at least three main reasons for this.

The first reason is that most of Perl has been derived from standard utilities, tools, and languages that you are (probably) already familiar with. If you have any knowledge of the C programming language and standard C library, the Unix Shell, sed and awk, Perl should be simple and fun for you to learn.

The second reason that Perl is easy to learn is that you only have to know a very small subset of Perl to be able to get useful results. In fact, once you can master

    print "Hello, world\n";
you can start writing Perl scripts. In fact, you will probably never have to (or be able to) know everything about Perl. As you feel the need or desire to use more sophisticated features (such as C structures or networking), you can learn these as you go. The learning curve for Perl is not a steep one, especially if you have the headstart of having a background in UNIX. Rather, its learning curve is gentle and gradual, but it is admittedly rather long.

The third reason is that you can get immediate results from your scripts. Unlike a normal compiled language (like C or Pascal, for example), you don't have to continually recompile your program every time you change one little thing. Perl allows you to experiment and test/debug quickly and easily. This ease of experimentation flattens the learning curve even more.

If you don't know C or UNIX at all, it'll be a steeper learning curve, but what you then learn from Perl will carry over into other areas, like using the C library, UNIX system calls, regular expressions, and associative arrays, just to name a few. To know Perl is to know UNIX, and vice versa.

1.10) Should I program everything in Perl?

Most definitely. In fact, you should delete the binaries for sed, awk, cc, gcc, grep, rm, ls, cat... well, just delete your /bin directory.

But seriously, of course you shouldn't. As with any job, you should use the appropriate tool for the task at hand. Just because a hammer will put screws into a piece of board, you probably don't want to do that.

While it's true that the answer to the question ``Can I do (some arbitrary task) in Perl?'' is almost always ``yes'', that doesn't mean this is necessarily a good thing to do. For many people, Perl serves as a great replacement for shell programming. For a few people, it also serves as a replacement for most of what they'd do in C. But for some things, Perl just isn't the optimal choice.

1.11) How does Perl compare with other scripting languages, like Tcl, Python or REXX?

REXX is an interpreted programming language first seen on IBM systems. Python is an interpreted programming language by Guido van Rossum*. TCL is John Ousterhout*'s embeddable command language, designed just for embedded command extensions, but lately used for larger applications. TCL's most intriguing feature for many people is the tcl/tk toolset that allows for interpreted X-based tools. Others use it for its ``expect'' extension.

To avoid any flamage, if you really want to know the answer to this question, probably the best thing to do is try to write equivalent code to do a set of tasks. All three have their own newsgroups in which you can learn about (but hopefully not argue about) these languages.

To find out more about these or other languages, you might also check out David Muir Sharnoff*'s posting ``Catalog of Compilers, Interpreters, and Other Language Tools'' which he posts to comp.lang.misc, comp.sources.d, comp.archives.admin, and news.answers newsgroups. It's a comprehensive treatment of many different languages. (Caveat lector: he considers Perl's syntax ``unappealing''.)

1.12) How can I get Perl over the Internet?

Perl is available from any comp.sources.misc archive. You can use an archie server (see the alt.sources FAQ in news.answers) to find these if you want.

Version 4:
Volume  Issues  Patchlevel and Notes
------  ------  ------------------------------------------------
  18    19-54   Patchlevel 3, Initial posting.
  20    56-62   Patches 4-10

Version 5:
Volume  Issues  Patchlevel and Notes
------  ------  -----------------------------------------------
  45    64-128  Initial Posting, patchlevel 0.
Since 1993, a number of archives have sprung up specifically for Perl and Perl related items. Larry maintains the official distribution site (for both perl4.036 and perl5) at netlabs. Probably the largest archive is at the University of Florida. In order of probability these sites will have the sources.

Site    Directory and notes                         IP
---------------------------------------------       -------

North America:               
        not current as of 7/15/95          
        not current as of 7/15/95        
        not current as of 7/15/95                



South America (mirror of           

If there is a site in Asia or Japan, please tell us about it. Thanks!

You can also retrieve perl via non-ftp methods:   

1.13) How can I get Perl via Email?

The following is a list of known ftpmail sites. Please attempt to use the site closest to you with the ftp archive closest to it. Many of these sites already have perl on them. For information on how to use one of these sites, send email containing the word ``help'' to the address.

    United States:
	New Jersey:
	North Carolina:


Henk P Penning* suggests that if you are in Europe you should try the following (if you are in Germany or the UK, you should probably use one of the servers listed above):
    Email: Send a message to '' containing:
     path your_email_address
     send help
     send PERL/perl5.0/INDEX
    The path-line may be omitted if your message contains a normal
    From:-line.  You will receive a help-file and an index of the
    directory that contains the Perl stuff.

If all else fails, mail to Larry usually suffices.

1.14) How can I get Perl via UUCP?

There currently is no way of getting Perl via UUCP. If anyone knows of a way, please contact me. The OSU site has discontinued the service.

1.15) Are there other ways of getting perl?

Another possibility is to use UUNET, although they charge you for it. You have been duly warned. Here's the advertisement:

Anonymous Access to UUNET's Source Archives

UUNET now provides access to its extensive collection of UNIX related sources to non- subscribers. By calling 1-900-468-7727 and using the login ``uucp'' with no password, anyone may uucp any of UUNET's on line source collection. Callers will be charged 40 cents per minute. The charges will appear on their next tele- phone bill.

The file uunet!/info/help contains instructions. The file uunet!/index//ls-lR.Z contains a complete list of the files available and is updated daily. Files ending in Z need to be uncompressed before being used. The file uunet!~/compress.tar is a tar archive containing the C sources for the uncompress program.

This service provides a cost effective way of obtaining current releases of sources without having to maintain accounts with UUNET or some other service. All modems connected to the 900 number are Telebit T2500 modems. These modems support all standard modem speeds including PEP, V.32 (9600), V.22bis (2400), Bell 212a (1200), and Bell 103 (300). Using PEP or V.32, a 1.5 megabyte file such as the GNU C compiler would cost $10 in con- nect charges. The entire 55 megabyte X Window system V11 R4 would cost only $370 in connect time. These costs are less than the official tape distribution fees and they are available now via modem.

UUNET Communications Services
3110 Fairview Park Drive, Suite 570
Falls Church, VA 22042
+1 703 876 5050 (voice)
+1 703 876 5059 (fax)

1.16) Has perl been ported to machine FOO?

Perl runs on virtually all Unix machines simply by following the hints file and instructions in the Configure script. This auto-configuration script allows Perl to compile on a wide variety of platforms by modifying the machine specific parts of the code. For most Unix systems, or VMS systems for v5 perl, no porting is required. Try to compile Perl on your machine. If you have problems, examine the README file carefully. If all else fails, send a message to comp.lang.perl.misc and crosspost to comp.sys.[whatever], there's probably someone out there that has already solved your problem and will be able to help you out.

Perl4.036 has been ported to many non-Unix systems, although currently there are only a few (beta) v5 ports. All of the following are mirrored at The following are the (known) official distribution points. Please contact the porters directly (when possible) in case of questions on these ports.

1.17) How do I get Perl to compile on Solaris?

The following directions are for perl, version 4. Perl, version 5, should compile more easily. If not, send mail to The Perl Porters Mailing List (

John Lees* reports:

I have built perl on Solaris 2.1, 2.2 beta, and 2.2 FCS. Take /usr/ucb out of your path and do not use any BSD/UCB libraries. Only -lsocket, -lnsl, and -lm are needed. You can use the hint for Solaris 2.0, but the one for 2.1 is wrong. Do not use vfork. Do not use -I/usr/ucbinclude. The result works fine for me, but of couse does not support a couple of BSDism's.

Casper H.S. Dik* reports

You must remove all the references to /usr/ucblib AND /usr/ucbinclude. And ignore the Solaris_2.1 hints. They are wrong. The undefining of vfork() probably has to do with the confusion it gives to the compilers. If you use cc, you mustn't compile util.c/tutil.c with -O. I only used the following libs: -lsocket -lnsl -lm (there is a problem with -lmalloc)

Michael D'Errico* reports:

If you are using Solaris 2.x, the signal handling is broken. If you set up a signal handler such as 'ripper' it will be forgotten after the first time the signal is caught. To fix this, you need to recompile Perl. Just add '#define signal(x,y) sigset((x),(y))' after the '#include ' directive in each file that it occurs, then make it again.

1.18) How do I get Perl to compile on a Next?

According to Andreas Koenig*, under NeXTstep 3.2, both perl4.036 and perl5.000 compile with the supplied hints file.

However, Bill Eldridge* provides this message to help get perl4.036 on NeXTstep 3.0 to work:

To get perl to compile on NeXTs, you need to combine the ANSI and BSD headers:

    cd /usr/include
    mkdir ansibsd
    cd ansibsd
    ln -s ../ansi
    ln -s ../bsd

Then, follow the configuration instructions for NeXTs, replacing all mention of -I/usr/include/ansi or -I/usr/include/bsd with -I/usr/include/ansibsd.

1.19) What extensions are available from Perl and where can I get them?

Some of the more popular extensions include those for windowing, graphics, or data base work. Most of the major sites contain an archive of the extensions, usually in the ext directory. Since the list of available extensions changes so often, I have opted to list only the sites and directories, not the individual extensions, please check the closest archive for more information

1.20) What is dbperl and where can I get it?

Many database-oriented extensions to Perl have been written. Basically, these use the usub mechanism (see the usub/ subdirectory) in the source distribution) to link in a database library, allowing embedded calls to Informix, Ingres, Interbase, Oracle and Sybase.

Here are the authors of the various extensions:

What            Target DB       Who
--------        -----------     ----------------------------------------
?Infoperl       Informix        Kurt Andersen (
Ingperl         Ingres          Tim Bunce ( and Ted Lemon
Interperl       Interbase       Buzz Moschetti (
Isqlperl        Informix        William Hails
Oraperl         Oracle          Kevin Stock (
Pgperl          Postgres        Igor Metz (
*Sqlperl        Ingres          Ted Lemon (
Sybperl         Sybase          Michael Peppler (
Uniperl         Unify 5.0       Rick Wargo (

    ? Does this one still exist?
*Sqlperl appears to have been subsumed by Ingperl

Buzz Moschetti* has organized a project to create a higher level interface to allow you to write your queries in a database-independent fashion. If this type of project interests you, send mail to <> and asked to be placed on the ``perldb-interest'' mailing lists.

Here's a bit of advertising from Buzz:

Perl is an interpreted language with powerful string, scalar, and array processing features developed by Larry Wall that ``nicely bridges the functionality gap between sh(1) and C.'' Since relational DB operations are typically textually oriented, perl is particularly well-suited to manage the data flows. The C source code, which is available free of charge and runs on many platforms, contains a user-defined function entry point that permits a developer to extend the basic function set of the language. The DBperl Group seeks to exploit this capability by creating a standardized set of perl function extensions (e.g. db_fetch(), db_attach()) based on the SQL model for manipulating a relational DB, thus providing a portable perl interface to a variety of popular RDMS engines including Sybase, Oracle, Ingres, Informix, and Interbase. In theory, any DB engine that implements a dynamic SQL interpreter in its HLI can be bolted onto the perl front end with predicatable results, although at this time backends exist only for the aforementioned five DB engines.

The official archive for DBperl extensions is It's the home of the evolving DBperl API Specification. Here's an extract from the updated README there:

DBI/        The home of the DBI archive. To join the DBI mailing list
	    send your request to

DBD/        Database Drivers for the DBI ...

Oracle/      By Tim Bunce (not yet ready!)
Ingres/      By Tim Bunce (not yet started!)

mod/           Other Perl 5 Modules and Extensions ...

Sybperl/    By Michael Peppler,

perl4/         Perl 4 extensions (using the usub C interface)

   oraperl/   ORACLE 6 & 7  By Kevin Stock,
   sybperl/   SYBASE 4      By Michael Peppler,
   ingperl/   INGRES        By Tim Bunce and Ted Lemon
   isqlperl/  INFORMIX      By William Hails,
   interperl/ INTERBASE     By Buzz Moschetti,
   oraperl/   ORACLE 6 & 7  By Kevin Stock (sadly no longer on the net)
   sybperl/   SYBASE 4      By Michael Peppler,
   ingperl/   INGRES        By Tim Bunce and Ted Lemon
   isqlperl/  INFORMIX      By William Hails,
   interperl/ INTERBASE     By Buzz Moschetti,
   uniperl/   UNIFY 5.0     By Rick Wargo,
   pgperl/    POSTGRES      By Igor Metz,

   btreeperl/ NDBM perl extensions.   By John Conover,
   ctreeperl/ C-Tree perl extensions. By John Conover,
   duaperl/   X.500 Directory User Agent. By Eric Douglas.

scripts/       Perl and shell scripts

   rdb/       RDB is a perl RDBMS for ASCII files. By Walt Hobbs,
   shql/      SHQL is an interactive SQL database engine.  Written as a
		shell script, SHQL interprets SQL commands and
		manipulates flat files based on those commands. By
		Bruce Momjian, root@candle.uucp
   xbase/     Perl scripts for accessing xBase style files (dBase III)

refinfo/       Reference information

   sqlsyntax/ Yacc and lex syntax and C source code for SQL1 and SQL2
	and a draft SQL3 syntax from Jeff Fried <>+
   formats/   Details of file formats such as Lotus 1-2-3 .WK1

There are also a number of non SQL database interfaces for perl available from These include:

Directory   Target System   Authors and notes
---------   -------------   -------------------------------------------
btreeperl   NDBM extension  John Conover (
ctreeperl   CTree extension John Conover (
duaperl     X.500 DUA       Eric Douglas
rdb         RDBMS           Walt Hobbs (
shql        SQL Engine      Bruce Momjian (root@candle.uucp)

1.21) Which DBM should I use?

As shipped, Perl (version 5) comes with interfaces for several DBM packages (SDBM, old DBM, NDBM, GDBM, Berkeley DBM) that are not supplied but either come with your system are readily accessible via FTP. SDBM is guaranteed to be there. For a comparison, see AnyDBM_File(3pm) and DB_File(3pm) .

1.22) Is there an SNMP aware Perl?

snmperl was written by Guy Streeter (, and was posted in late February 1993 to comp.protocols.snmp. It can be found archived at one of two (known) places:

Location: /pub/net/snmp
       FILE -rw-rw-r--       3407  Aug 11 1992  snmperl.README
       FILE -rw-r--r--      17678  Aug 11 1992  snmperl.tar.Z

Location: /pub/perl/scripts/snmp

Here is the gist of the README:

This directory contains the source code to add callable C subroutines to perl. The subroutines implement the SNMP functions ``get'', ``getnext'', and ``set''. They use the freely-distributable SNMP package (version 1.1b) from CMU.

USE: There are four subroutines defined in the callable interface: snmp_get, snmp_next, snmp_set, and snmp_error.

snmp_get and snmp_next implement the GET and GETNEXT operations, respectively. The first two calling arguments are the hostname and Community string. The IP address of the host, as a dotted-quad ASCII string, may be used as the hostname. The rest of the calling arguments are a list of variables. See the CMU package documentation for how variables may be specified.

snmp_set also takes hostname and Community string as arguments. The remaining arguments are a list of triples consisting of variable name, variable type, and value. The variable type is a string, such as ``INTEGER'' or ``IpAddress''.

snmp_get, snmp_next, and snmp_set return a list containing alternating variables and values. snmp_get and snmp_next will simply omit non-existent variables on return. snmp_set will fail completely if one of the specified variables does not exist (or is read-only).

snmp_error will return a text string containing some error information about the most recent snmp_get|next|set call, if it had an error.

I didn't find all the places where the CMU library writes to stderr or calls exit() directly.

The changes I made to mib.c involve the formatting of variable values for return to the caller. I took out the descriptive prefix so the string contains only the value.

Enumerated types are returned as a string containing the symbolic representation followed in parentheses by the numeric.


perl and the CMU SNMP package have their own statements. Read them. The work I've done is free and clear. Just don't say you wrote it if you didn't, and don't say I wrote it if you change it.

    Guy Streeter
    April 1, 1992 (not a joke!)

1.23) Is there an ISO or ANSI certified version of Perl?

No. Larry thinks it likely that he'll be certified before perl is.

Informational Sources

2.1) Is there a USENET group for perl?
2.2) Have any books or magazine articles been published about perl?
2.3) When will the Camel and Llama books be updated?
2.4) What FTP resources are available?
2.5) What WWW/gopher resources are available?
2.6) Can people who don't have access to USENET get comp.lang.perl.misc?
2.7) Are archives of comp.lang.perl.* available?
2.8) Is there a WAIS server for comp.lang.perl.*?
2.9) What other sources of information about Perl or training are available?
2.10) Where can I get training classes on Perl?
2.11) What companies ship or use perl?
2.12) Is there commercial, third-party support for perl?
2.13) What is a JAPH? What does "Will hack perl for ..." mean?
2.14) Where can I get a collection of Larry Wall witticisms?
2.15) What are the known bugs?
2.16) Where should I post bugs?
2.17) Where should I post source code?
2.18) Where can I learn about object-orienting Perl programming?
2.19) Where can I learn about linking C with Perl? [h2xs, xsubpp]
2.20) What is
2.21) What do the asterisks (*) throughout the FAQ stand for?

2.1) Is there a USENET group for Perl?

Yes there is: comp.lang.perl.misc. This group, which currently can get up to 150 messages per day, contains all kinds of discussions about Perl; everything from bug reports to new features to the history to humour and trivia. This is the best source of information about anything Perl related, especially what's new with Perl5. Because of its vast array of topics, it functions as both a comp.lang.* style newsgroup (providing technical information) and also as a rec.* style newsgroup, kind of a support group for Perl addicts (PerlAnon?). There is also the group comp.lang.perl.announce, a place specifically for announcements related to perl (new releases, the FAQ, new modules, etc).

Larry is a frequent poster to this group as well as most (all?) of the seasoned Perl programmers. Questions will be answered by some of the most knowledgable Perl Hackers, often within minutes of a question being posted (give or take distribution times).

2.2) Have any books or magazine articles been published about Perl?

There are a number of books either available or planned. Mostly chronologically, they are:

Programming Perl (the ``Camel Book''):

Author: Larry Wall and Randal Schwartz
Publisher: O'Reilly and Associates
ISBN 0-937175-64-1 (English)
ISBN 4-89052-384-7 (Japanese)
ISBN 3-446-17257-2 (German) (Programmieren in Perl)
(translator: Hanser Verlag)

This is probably the most well known and most useful book for 4.036 and earlier. This part of O'Reilly's hugely successful ``Nutshell Handbook'' series. Besides serving as a reference guide for Perl, it also contains tutorial material and is a great source of examples and cookbook procedures, as well as wit and wisdom, tricks and traps, pranks and pitfalls. The code examples contained therein are available from or Corrections and additions to the book can be found in the Perl4 man page right before the BUGS section under the heading ERRATA AND ADDENDA.

Learning Perl (the ``Llama Book''):

Another of O'Reilly's ``Nutshell Handbooks'', by Randal Schwartz. This book is a smaller, gentler introduction to perl and is based off of Randal's perl classes. While in general this is a good book for learning perl (like its title), early printings did contain many typos and don't cover some of the more interesting features of perl. Please check the errata sheet at, as well as the on-line examples. If you can't find these books in your local technical bookstore, they may be ordered directly from O'Reilly by calling 1-800-998-9938 if in North America and 1-707-829-0515 otherwise.

Johan Vromans* created a beautiful reference guide. The reference guide comes with the Camel book in a nice, glossy format. The LaTeX (source) and PostScript (ready to print) versions are available for FTP from in Europe or from in the United States. Obsolete versions in TeX or troff may still be available, but these versions don't print as nicely. See also:

Johan has also updated and released a reference guide based on version 5.000. This is available from the same places as the 4.036 guide. This version is also available from in the /pub/gnu section along with the perl5 source. It may be added to the standard perl5 distribution sometime after 5.002. If you are using version 5.000, you will want to get this version rather than the 4.036 version.

Larry routinely carries around a camel stamp to use when autographing copies of his book. If you can catch him at a conference you can usually get him to sign your book for you.

Prentice Hall also has two perl books.

The first is Perl by Example by Ellie Quigley. (385 pages, $26.96, ISBN 0-13-122839-0) A perl tutorial (perl4); every feature is presented via an annotated example and sample output. Reviews of this book have varied widely. Many new perl users have used this book with much success, while many ``veteran'' programmers have had many complaints about it.

The second book is called Software Engineering with Perl by Carl Dichter and Mark Pease. Randal Schwartz was a technical reviewer for this book and notes this:

SEwP is not meant as instruction in the Perl language, but rather as an example of how Perl may be used to assist in the semi-formal software engineering development cycles. There's a lot of Perl code that's fairly well commented, but most of the book describes software engineering methodologies. For the perl-challenged, there's a light treatment of the language as well, but they refer to the llama and the camel for the real meat.

SAMS Publishing also has a Perl book available, as part of their ``Teach Yourself in 21 Days'' series, called Teach Yourself Perl in 21 Days. ISBN 0-672-30586-0 Price: $29.95, 841 Pages. This book is the first book to have a section devoted to version 5.000, although it was written during an alpha stage and may not necessarily reflect current reality.

Please note that none of the above books are perfect, all have some inaccurances and typos. The two which Larry is directly associated with (the O'Reilly books) are probably the most technically correct, but also the most dated. Carefully looking over any book you are considering purchasing will save you much time, money, and frustration.

Starting in the March, 1995 edition of Unix Review. Randal Schwartz* has been authoring a bi-monthly Perl column. This has so far been an introductory tutorial.

Larry Wall has published a 3-part article on perl in Unix World (August through October of 1991), and Rob Kolstad also had a 3-parter in Unix Review (May through July of 1990). Tom Christiansen also has a brief overview article in the trade newsletter Unix Technology Advisor from November of 1989. You might also investigate ``The Wisdom of Perl'' by Gordon Galligher from SunExpert magazine; April 1991 Volume 2 Number 4. The Dec 92 Computer Language magazine also contains a cover article on Perl, ``Perl: the Programmers Toolbox''.

Many other articles on Perl have been recently published. If you have references, especially on-line copies, please mail them to the FAQ maintainer for inclusion is this notice.

The USENIX LISA (Large Installations Systems Administration) Conference have for several years now included many papers of tools written in Perl. Old proceedings of these conferences are available; look in your current issue of ``;login:'' or send mail to for further information.

Japan seems to be jumping with Perl books. If you can read japanese here are a few you might be interested in. Thanks to Jeffrey Friedl* and Ken Lunde* for this list (NOTE: my screen cannot handle japanese characters, so this is all in English for the moment NOTE2: These books are written in Japanese, these titles are just translations):

Title: Welcome to Perl Country	(Perl-no Kuni-he Youkoso)
Authors: Kaoru Maeda, Hiroshi Koyama, Yasushi Saito and Arihito
Pages: 268+9	    	    	Publisher: Science Company	
Pub. Date: April 25, 1993   	ISBN: 4-7819-0697-4 		
Price: 2472Y	    	    	Author Email:
Comments: Written during the time the Camel book was being 
translated.  A useful introduction, but uses jperl (Japanese Perl)
which is not necessarily compatible.

Title: How to Write Perl (Perl Shohou) Author: Toshiyuki Masui Pages: 352 Publisher: ASCII Corporation Pub. Date: July 1, 1993 ISBN: 4-7561-0281-6 Price: 3200Y Author Email: Comments: More advanced than ``Welcome..'' and not meant as an introduction. Uses the standard perl and has examples for handling Japanese text.

Title: Introduction to Perl (Nyuumon Perl) Author: Shinji Kono Pages: 203 Publisher: ASCII Corporation Date: July 11, 1994 ISBN: 4-7561-0292-1 Price: 1800Y Author Email: Comments: Uses the interactive Perl debugger to explain how things work.

Title: Perl Programming Authors: L Wall & R Schwartz Translator: Yoshiyuki Kondo Pages: 637+32 Publisher: Softbank Corporation Pub. Date: February 28, 1993 ISBN: 4-89052-384-7 Price: 4500Y Author Email: Comments: Official Japanese translation of the Camel book, ``Programming Perl''. Somewhat laced with translator notes to explain the humour. The most useful book. Also includes the Perl Quick Reference -- in Japanese!

2.3) When will the Camel and Llama books be updated?

As of August, 1995, ORA has contracted with Stephen to handle the Camel update. According to the accepted timeline, the first draft is to be finished by the end of April, 1996. The tutorial sections are being cut some, and the book will take on much more of a reference style. Don't worry, it will still contain its distinctive humor and flair.

There are no current plans to update the Llama. For the most part, it serves as a good introduction for both major versions of perl. There may be some minor editing to it, but probably nothing major. If anything, it is more likely that a third book (working title: Learning More Perl) will be written as a tutorial for the new perl5 paradigm.

2.4) What FTP resources are available?

Since 1993, several ftp sites have sprung up for Perl and Perl related items. The site with the biggest repository of Perl scripts right now seems to be [] in /pub/perl. The scripts directory has an INDEX with over 400 lines in it, each describing what the script does. The src directory has sources and/or binaries for a number of different perl ports, including MS-Dos, Macintosh and Windows/NT. This is maintained by the Computing Staff at UF*.

Note: European users please use the site [] in /pub/computing/programming/languages/perl/ The link speed would be a lot better for all. Contact for more information. It is updated daily.

There are also a number of other sites. I'll add more of them as I get information on them. [site maintainers: if you want to add a blurb here, especially if you have something unique, please let me know. -spp]

The Comprehensive Perl Archive Network (CPAN) is in heavy development. Once the main site and its mirrors are fully operational, this answer will change to reflect its existence.

2.5) What WWW/gopher resources are available?

The World Wide Web is exploding with new Perl sites all the time. Some of the more notable ones are:, which has a great section on Perl5., a great site for European and UK users.

2.6) Can people who don't have access to USENET get comp.lang.perl.misc?

``Perl-Users'' is the mailing list version of the comp.lang.perl.misc newsgroup. If you're not lucky enough to be on USENET you can post to comp.lang.perl.misc by sending to one of the following addresses. Which one will work best for you depends on which nets your site is hooked into. Ask your local network guru if you're not certain.


The Perl-Users list is bidirectionally gatewayed with the USENET newsgroup comp.lang.perl.misc. This means that VIRGINIA functions as a reflector. All traffic coming in from the non-USENET side is immediately posted to the newsgroup. Postings from the USENET side are periodically digested and mailed out to the Perl-Users mailing list. A digest is created and distributed at least once per day, more often if traffic warrants.

All requests to be added to or deleted from this list, problems, questions, etc., should be sent to:

Marc Rouleau <mer6g@VIRGINIA.EDU>

2.7) Are archives of comp.lang.perl.misc available?

Yes, there are.*/monthly has an almost complete collection dating back to 12/89 (missing 08/91 through 12/93). They are kept as one large file for each month.

A more sophisticated query and retrieval mechanism is desirable. Preferably one that allows you to retrieve article using a fast-access indices, keyed on at least author, date, subject, thread (as in ``trn'') and probably keywords. Right now, the MH pick command works for this, but it is very slow to select on 18000 articles.

If you have, or know where I can find, the missing sections, please let know.

2.8) Is there a WAIS server for comp.lang.perl.*?

Yes there is. Set your WAIS client to*. According to their introduction, they have a complete selection from 1989 on.

Bill Middleton <> offers this:

I have setup a perl script retrieval service and WaisSearch here at feenix. To check it out, just point your gopher at us, and select the appropriate menu option. The WaisSearch is of the iubio type, which means you can do boolean searching. Thus you might try something like:
ioctl and fcntl
grep and socket not curses

and other things to see examples of how other folks have done this or that. This service is still under construction, but I'd like to get feedback, if you have some time.

There's also a WaisSearch into all the RFC's and some other fairly nifty stuff.

2.9) What other sources of information about Perl or training are available?

There is a #Perl channel on IRC (Internet Relay Chat) where Tom and Randal have been known to hang out. Here you can get immediate answers to questions from some of the most well-known Perl Hackers.

The perl5-porters ( mailing list was created to aid in communication among the people working on perl5. However, it has overgrown this function and now also handles a good deal of traffic about perl internals.

2.10) Where can I get training classes on Perl?

USENIX, LISA, SUG, WCSAS, AUUG, FedUnix and Europen sponsor tutorials of varying lengths on Perl at the System Administration and General Conferences. These public classes are typically taught by Tom Christiansen*.

In part, Tom and Randal teach Perl to help keep bread on their tables long enough while they continue their pro bono efforts of documenting perl (Tom keeps writing more man pages for it :-) and expanding the perl toolkit through extension libraries, work which they enjoy doing as it's fun and helps out the whole world, but which really doesn't pay the bills. Such is the nature of free(ly available) software. Send mail to <> for details and availability.

Tom is also available to teach on-site classes, included courses on advanced perl and perl5. Classes run anywhere from one day to week long sessions and cover a wide range of subject matter. Classes can include lab time with exercises, a generally beneficial aspect. If you would like more information regarding Perl classes or when the next public appearances are, please contact Tom directly at 1.303.444.3212.

Randal Schwartz* provides a 2-day lecture-only and a 4-5 day lecture-lab course based on his popular book ``Learning Perl''. For details, contact Randal directly via email or at 1.503.777.0095.

Internet One provides a 2 day ``Introduction to Perl'' and 2 day ``Advanced Perl'' workshop. The 50% hands-on and 50% lecture format allow attendees to write several programs themselves. Supplied are the user manuals, reference copies of Larry Wall's ``Programming Perl'', and a UNIX directory of all training examples and labs. To obtain outlines, pricing, or scheduling information, use the following:

2.11) What companies use or ship Perl?

At this time, the known list of companies that ship Perl includes at least the following, although some have snuck it into /usr/contrib or its moral equivalent:

BSDI Comdisco Systems CONVEX Computer Corporation Crosspoint Solutions Data General Dell DRD Corporation IBM (SP systems) Intergraph Kubota Pacific Netlabs SGI (without taintperl) Univel

Some companies ship it on their ``User Contributed Software Tape'', such as DEC and HP. Apple Computer has shipped the MPW version of Macintosh Perl on one of their Developer CDs (EssentialsToolsObjects #11) (and they included it under ``Essentials'' :-)

Many other companies use Perl internally for purposes of tools development, systems administration, installation scripts, and test suites. Rumor has it that the large workstation vendors (the TLA set) are seriously looking into shipping Perl with their standard systems ``soon''.

People with support contracts with their vendors are actively encouraged to submit enhancement requests that Perl be shipped as part of their standard system. It would, at the very least, reduce the FTP load on the Internet. :-)

If you know of any others, please send them in.

2.12) Is there commercial, third-party support for Perl?

Not really. Although perl is included in the GNU distribution, at last check, Cygnus does not offer support for it. However, it's unclear whether they've ever been offered sufficient financial incentive to do so. Feel free to try.

On the other hand, you do have comp.lang.perl.misc as a totally gratis support mechanism. As long as you ask ``interesting'' questions, you'll probably get plenty of help. :-)

While some vendors do ship Perl with their platforms, that doesn't mean they support it on arbitrary other platforms. And in fact, all they'll probably do is forward any bug reports on to Larry. In practice, this is far better support than you could hope for from nearly any vendor.

If you purchase a product from Netlabs (the company Larry works for), you actually can get a support contract that includes Perl.

The companies who won't use something unless they can pay money for it will be left out. Often they're motivated by wanting someone whom they could sue. If all they want is someone to help them out with Perl problems, there's always the net. And if they really want to pay someone for that help, well, any of a number of the regular Perl ``dignitaries'' would appreciate the money. ;-)

If companies want ``commercial support'' for it badly enough, speak up -- something might be able to be arranged.

2.13) What is a JAPH? What does "Will hack perl for ..." mean?

These are the ``just another perl hacker'' signatures that some people sign their postings with. About 100 of the of the earlier ones are available from the various FTP sites.

When people started running out of tricky and interesting JAPHs, some of them turned to writing ``Will hack perl for ...'' quotes. While sometimes humourous, they just didn't have the flair of the JAPHs and have since almost completely vanished.

2.14) Where can I get a list of Larry Wall witticisms?

Over a hundred quips by Larry, from postings of his or source code, can be found in many of the FTP sites or through the World Wide Web at ""

2.15) What are the known bugs?

This is NOT a complete list, just some of the more common bugs that tend to bite people. There are in 5.001:

2.16) Where should I post bugs?

Before posting about a bug, please make sure that you are using the most recent versions of perl (currently 4.036 and 5.001) available. Please also check at the major archive sites to see if there are any development patches available (usually named something like perl5.001a.patch or patch5.001a - the patch itself, or perl5.001a.tar.gz - a prepatched distribution). If you are not using one of these versions, chances are you will be told to upgrade because the bug has already been fixed.

If you are reporting a bug in perl5, the best place to send your bug is <>, which is currently just an alias for <>. In the past, there have been problems with the perlbug address. If you have problems with it, please send your bug directly to <>. You may subscribe to the list in the customary fashion via mail to <>. Feel free to post your bugs to the comp.lang.perl.misc newsgroup as well, but do make sure they still go to the mailing list.

If you are posting a bug with a non-Unix port, a non-standard Module (such as Tk, Sx, etc) please see the documentation that came with it to determine the correct place to post bugs.

To enhance your chances of getting any bug you report fixed:

  1. Make sure you are using a production version of perl. Alpha and Beta version problems have probably already been reported and fixed.

  2. Try to narrow the problem down to as small a piece of code as possible. If you can get it down to 1 line of Perl then so much the better.

  3. Include a copy of the output from the myconfig script from the Perl source distribution in your posting.

2.17) Where should I post source code?

You should post source code to whichever group is most appropriate, but feel free to cross-post to comp.lang.perl.misc. If you want to cross-post to alt.sources, please make sure it follows their posting standards, including setting the Followups-To header line to NOT include alt.sources; see their FAQ for details.

2.18) Where can I learn about object-oriented Perl programming?

The perlobj(1) man page is a good place to start, and then you can check out the excellent perlbot(1) man page written by the dean of perl o-o himself, Dean Roehrich. Areas covered include the following:

Idx  Subsections in perlobj.1          Lines
 1   NAME                                  2
 2   DESCRIPTION                          16
 3   An Object is Simply a Reference      60
 4   A Class is Simply a Package          31
 5   A Method is Simply a Subroutine      34
 6   Method Invocation                    75
 7   Destructors                          14
 8   Summary                               7

Idx  Subsections in perlbot.1          Lines
 1   NAME                                  2
 2   INTRODUCTION                          9
 3   Instance Variables                   43
 4   Scalar Instance Variables            21
 5   Instance Variable Inheritance        35
 6   Object Relationships                 33
 7   Overriding Superclass Methods        49
 8   Using Relationship with Sdbm         45
 9   Thinking of Code Reuse              111


The section on instance variables should prove very helpful to those wondering how to get data inheritance in perl.

2.19) Where can I learn about linking C with Perl? [h2xs, xsubpp]

While it used to be deep magic, how to do this is now revealed in the perlapi(1) , perlguts(1) , and perlcall(1) man pages, which treat with this matter extensively. You should also check the many extensions that people have written (see question 1.19), many of which do this very thing.

2.20) What is is just Tom's domain name, registered as dedicated to ``Perl training and consulting''. While not a full ftp site (he hasn't got the bandwidth (yet)), it does have some interesting bits, most of which are replicated elsewhere. It serves as a clearinghouse for certain perl related mailing lists. The following aliases work:

perl-packrats:          The archivist list
perl-porters:           The porters list
perlbook:               The Camel/Llama/Alpaca writing committee
perlbugs:               The bug list (perl-porters for now)
perlclasses:            Info on Perl training
perlfaq:                Submissions/Errata to the Perl FAQ
			(Tom and Steve)
perlrefguide:           Submissions/Errata to the Perl RefGuide

2.21) What do the asterisks (*) throughout the FAQ stand for?

To keep from cluttering up the FAQ and for easy reference all email addresses have been collected in this location. For each person listed, I offer my thanks for their input and help.

Larry Wall  
Tom Christiansen
Stephen P Potter
Andreas König         k@franz.ww.TU-Berlin.DE
Bill Eldridge
Buzz Moschetti
Casper H.S. Dik
David Muir Sharnoff
Dean Roehrich
Dominic Giampaolo
Frédéric Chauveau
Gene Spafford
Guido van Rossum
Henk P Penning
Jeff Friedl 
Johan Vromans
John Dallman
John Lees   
John Ousterhout
Jon Biggar  
Ken Lunde   
Malcolm Beattie
Matthias Neeracher
Michael D'Errico
Nick Ing-Simmons
Randal Schwartz
Roberto Salama
Steven L Kunz
Thomas R. Kimpton
Tim Bunce   
Timothy Murphy
UF Computer Staff
William Setzer

Programming Aids

3.1) How do I use perl interactively?
3.2) Is there a perl profiler?
3.3) Is there a yacc for perl?
3.4) Is there a pretty printer (similar to indent(1)) for perl?
3.5) How can I convert my perl scripts directly to C or compile them into binary form?
3.6) Where can I get a perl mode for emacs?
3.7) Is there a perl shell?
3.8) How can I use curses with perl?
3.9) How can I use X or Tk with perl?
3.10) Can I dynamically load C user routines?
3.11) What is undump and where can I get it?
3.12) How can I get '#!perl' to work under MS-DOS?
3.13) Can I write useful perl programs on the command line?
3.14) What's a "closure"?

3.1) How can I use Perl interactively?

The easiest way to do this is to run Perl under its debugger. If you have no program to debug, you can invoke the debugger on an `empty' program like this:

perl -de 0

(The more positive hackers prefer perl -de 1. :-)

Now you can type in any legal Perl code, and it will be immediately evaluated. You can also examine the symbol table, get stack backtraces, check variable values, and if you want to, set breakpoints and do the other things you can do in a symbolic debugger.

3.2) Is there a Perl profiler?

While there isn't one included with the perl source distribution (yet) various folks have written packages that allow you to do at least some sort of profiling. The strategy usually includes modifying the perl debugger to handle profiling. Authors of these packages include

Wayne Thompson me@anywhere.EBay.Sun.COM Ray Lischner Kresten Krab Thorup

The original articles by these folks containing their profilers are available at

Recently, Dean Roehrich* has written a profiler for version 5 that likely will be distributed with the standard release. For now, it should be available through any of the extension archives as DProf.tar.gz.

3.3) Is there a yacc for Perl?

Yes!! It's a version of Berkeley yacc that outputs Perl code instead of C code! You can get this from, or send the author mail for details.

3.4) Is there a pretty-printer (similar to indent(1)) for Perl?

That depends on what you mean. If you want something that works like vgrind on Perl programs, then the answer is ``yes, nearly''. Here's a vgrind entry for perl:

    if for foreach unless until while continue else elsif \
    do eval require \
    die exit \
    defined delete reset \
    goto last redo next dump \
    local undef return  \
    write format  \
    sub package

It doesn't actually do everything right; in particular, things like $#, $', s#/foo##, and $foo'bar all confuse it.

David Levine uses this:

# perl 4.x                    David Levine <> 05 apr 1993
# Derived from Tom Christiansen's perl vgrindef.  I'd like to treat all  of
# perl's built-ins as  keywords,  but vgrind   fields are  limited  to 1024
# characters  and the built-ins overflow that (surprise  :-).  So, I didn't
# include the dbm*, end*, get*, msg*, sem*, set*,  and  shm* functions.   I
# couldn't come up with an easy way to  distinguish beginnings  of literals
# ('...') from package prefixes, so literals are not marked.
# Be sure to:
# 1) include whitespace between a subprogram name and its opening {
# 2) include whitespace before a comment (so that $# doesn't get
# interpreted as one).
	:kw=accept alarm atan2 bind binmode caller chdir chmod chop \
chown chroot close closedir connect continue cos crypt defined delete \
die do dump each else elsif eof eval exec exit exp fcntl fileno flock \
for foreach fork format getc gmtime goto grep hex if include index int \
ioctl join keys kill last length link listen local localtime log lstat \
m mkdir next oct open opendir ord pack package pipe pop print printf \
push q qq qx rand read readdir readlink recv redo rename require reset \
return reverse rewinddir rindex rmdir s scalar seek seekdir select send \
shift shutdown sin sleep socket socketpair sort splice split sprintf \
sqrt srand stat study sub substr symlink syscall sysread system \
syswrite tell telldir time times tr truncate umask undef unless unlink \
unpack unshift until utime values vec wait waitpid wantarray warn while \
write y: 

If what you mean is whether there is a program that will reformat the program much as indent(1) will do for C, then the answer is no. The complex feedback between the scanner and the parser (as in the things that confuse vgrind) make it challenging at best to write a stand-alone Perl parser.

Of course, if you follow the guidelines in perlstyle(1), you shouldn't need to reformat.

3.5) How can I convert my perl scripts directly to C or compile them into binary form?

The short answer is: ``No, you can't compile perl into C. Period.''

However, having said that, it is believed that it would be possible to write a perl to C translator, although it is a PhD thesis waiting to happen. Anyone need a good challenging thesis?

In the way of further, detailed explication, it seems that the reasons people want to do this usaully break down into one or more of the following:

  1. speed
  2. secrecy
  3. maintainability

  1. SPEED:

    1. You can't turn perl source code or perl intermediary code into native machine code to make it run faster, and saving the perl intermediary code doesn't really buy you as much as you'd like. If you really must, check out the undump and unexec alternatives. If your motivations are speed, then this may or may not help you much.

      You might also look into autoloading functions on the fly, which can greatly reduce start-up time.

    2. If you have a few routines that are bogging you down, you just possibly might wish to hand-translate just them into C, then dynamically load these in. See perlapi(1) for details. Most of the time, however, reorganizing your perl algorithm is the best way to address this.


    1. If you're trying to stop people from seeing what you're doing, you can shroud it, i.e. turn all the idents into silly stuff, rearrange strings, and remove redundant white space. There's a program out there called ShroudIt! that works on a number of languages, including Perl. Note that it is a commercial product though. Contact David Webber ( for more information.
    2. You might also look into the cryptswitch() stuff in the perl source, which would allow you to ship something in a form they can't read. This isn't particulary well-documented.

    3. If you're worried about them using your software without licence, you put some huge disclaimer at the top that says something like the following. This is actually the best solution, because only a legal solution will really work if legality is what you're worried about: trying to solve legal problems with technical solutions is not worth the effort, and too easily circumvented.
      This is UNPUBLISHED PROPRIETARY SOURCE CODE of XYZZY, Inc.; the contents of this file may not be disclosed to third parties, copied or duplicated in any form, in whole or in part, without the prior written permission of XYZZY, Inc.

      Permission is hereby granted soley to the licencee for use of this source code in its unaltered state. This source code may not be modified by licencee except under direction of XYZZY Inc. Neither may this source code be given under any circumstances to non-licensees in any form, including source or binary. Modification of this source constitutes breach of contract, which voids any potential pending support responsibilities by XYZZY Inc. Divulging the exact or paraphrased contents of this source code to unlicensed parties either directly or indirectly constitutes violation of federal and international copyright and trade secret laws, and will be duly prosecuted to the fullest extent permitted under law.

      This software is provided by XYZZY Inc. ``as is'' and any express or implied warranties, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose are disclaimed. In no event shall the regents or contributors be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, data, or profits; or business interruption) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this software, even if advised of the possibility of such damage.


    1. If you just want to stop people from changing it because you're concerned about support issues, you can put in a big disclaimer at the top that says that if they touch the file they void the warranty, and then make them give you a size, checksum, and version of the file before answering any questions about it.

      If you maintain a central site that distributes software to internal client machines, use rdist(1) to send around a proper version periodically, perhaps using the -y option on the install to flag destinations younger than the source.

      Let it be noted than in the many, many years that Perl's author has been releasing and supporting freely redistributable software, he has NEVER ONCE been bitten by a bogus bug report generated by someone breaking his code because they had access to it. Rather, he and many other open software provided (where open software means that for which the source is provided, the only truly open software) have saved themselves countless hours of labor thousands of times over because they've allowed people to inspect the source for themselves. Proprietary source-code hoarding is its own headache.

      Thus, obscurity for the sake of maintainability would seem to be a red herring.

    2. If you can't count on perl being installed at the destination customer, then by all means, merely ship it with your program. This is no hardship, since software providers are accustomed to shipping software in machine-specific binary form. The basic idea is as simple as:
      shar /usr/local/{lib,bin,man}/perl myprog
      Just don't overwrite their own Perl installation if they have one!

3.6) Where can I get a perl-mode for emacs?

Since Emacs version 19 patchlevel 22 or so, there has been both a perl-mode.el and support for the perl debugger built in. These should come with the standard Emacs 19 distribution.

In the perl source directory, you'll find a directory called ``emacs'', which contains several files that should help you.

Note that the perl-mode of emacs will have fits with main'foo (single quote), and mess up the indentation and hilighting. However, note that in perl5, you should be using main::foo. By the way, did we mention that you should upgrade?

3.7) Is there a Perl shell?

Not really. Perl is a programming language, not a command interpreter. There is a very simple one called ``perlsh'' included in the Perl source distribution. It just does this:
    $/ = '';        # set paragraph mode
    $SHlinesep = "\n";
    while ($SHcmd = <>) {
	$/ = $SHlinesep;
	eval $SHcmd; print $@ || "\n";
	$SHlinesep = $/; $/ = '';
Not very interesting, eh?

Daniel Smith <> is working on an interactive Perl shell called SoftList. It's currently at version 3.0b7a (beta). SoftList 3.0b7a has tcsh-like command line editing, can let you define a file of aliases so that you can run chunks of perl or UNIX commands, and so on. You can pick up a copy at in /pub/dls/SoftList-3.0b7a.gz.

3.8) How can I use curses with perl?

In release 4 of perl, the only way to do this was was to build a curseperl binary by linking in your C curses library as described in the usub subdirectory of the perl sources. This requires a modicum of work, but it will be reasonably fast since it's all in C (assuming you consider curses reasonably fast. :-) Programs written using this method require the modified curseperl, not vanilla perl, to run. While this is something of a disadvantage, experience indicates that it's better to use curseperl than to try to roll your own using termcap directly.

Fortunately, in version 5, Curses is a dynamically loaded extension by William Setzer*. You should be able to pick it up wherever you get Perl 5 from, or at least these places (expect that the version may change by the time you read this):

For a good example of using curses with Perl, you might want to pick up a copy of Steven L Kunz's* ``perl menus'' package (``'') via anonymous FTP from ``''. It's in the directory /pub/perl as: is a complete menu front-end for perl+curses and demonstrates a lot of things (plus it is useful to boot if you want full-screen menu selection ability). It provides full-screen menu selection ability for three menu styles (single-selection, multiple-selection, and ``radio-button''). The ``perl menus'' package also includes routines for full-screen data entry. A ``template'' concept is implemented to create a simple (yet flexible) perl interface for building data-entry screens for registration, database, or other record-oriented tasks. is supported on Perl4/curseperl and Perl5/Curses. Complete user documentation is provided along with several demos and ``beginner applications''. A menu utility module is provided that is a collection of useful Perl curses routines (such as "pop-up query boxes) that may be called from your applications.

Another possibility is to use Henk Penning's cterm package, a curses emulation library written in perl. cterm is actually a separate program with which you communicate via a pipe. It is available from [] via anonymous ftp. in the directory pub/PERL. You may also acquire the package via email in compressed, uuencoded form by sending a message to containing these lines:

    send PERL/cterm.shar.Z

See the question on retrieving perl via mail for more information on how to retrieve other items of interest from the mail server there.

3.9) How can I use X or Tk with Perl?

Right now, you have several choices. If you are still using perl4, use the WAFE or STDWIN packages, or try to make your own usub binding.

However, if you've upgraded to version 5, you have several exciting possibilities, with more popping up each day. Right now, Tk and Sx are the best known such extensions.

If you like the tk package, you should get the Tk extension kit, written by Nick Ing-Simmons*. The official distribution point is at

but many of the major archive sites now have it in their /ext{entions} directory also. Depending upon your location, you may be better off checking there. Also, understand that the version number may have changed by the time you read this.

This package replaced the tkperl5 project, by Malcolm Beattie*, which was based on an older version of Tk, 3.6 as compared to the current 4.X. This package was also known as nTk (new Tk) while it was in the alpha stages, but has been changed to just Tk now that it is in beta. Also, be advised that you need at least perl5.001 (preferably 5.002, when it becomes available) and the official unofficial patches.

You may also use the old Sx package, (Athena & Xlib), written by originally written by by Dominic Giampaolo*, then and rewritten for Sx by Frédéric Chauveau*. It's available from these sites:

STDWIN is a library written by Guido van Rossum* (author of the Python programming language) that is portable between Mac, Dos and X11. One could write a Perl agent to speak to this STDWIN server.

WAFE is a package that implements a symbolic interface to the Athena widgets (X11R5). A typical Wafe application consists in our framework of two parts: the front-end (we call it Wafe for Widget[Athena]front end) and an application program running typically as a separate process. The application program can be implemented in an arbitrary programming language and talks to the front-end via stdio. Since Wafe (the front-end) was developed using the extensible TCL shell (cite John Ousterhout), an application program can dynamically submit requests to the front-end to build up the graphical user interface; the application can even down-load application specific procedures into the front-end. The distribution contains sample application programs in Perl, GAWK, Prolog, TCL, and C talking to the same Wafe binary. Many of the demo applications are implemented in Perl. Wafe 0.9 can be obtained via anonymous ftp from[]:pub/src/X11/wafe-0.9.tar.Z

Alternatively, you could use wish from tcl.

#  An example of calling wish as a subshell under Perl and
#  interactively communicating with it through sockets.
#  The script is directly based on Gustaf Neumann's perlwafe script.
#  Dov Grobgeld
#  1993-05-17

    $wishbin = "/usr/local/bin/wish";

    die "socketpair unsuccessful: $!!\n" unless socketpair(W0,WISH,1,1,0);
    if ($pid=fork) {
	    select(WISH); $| = 1;

	# Create some TCL procedures
	    print WISH 'proc echo {s} {puts stdout $s; flush stdout}',"\n";

	# Create the widgets
	print WISH &;lt;&;lt;TCL;
	# This is a comment "inside" wish

	frame .f -relief raised -border 1 -bg green
	pack append . .f {top fill expand}

	button .f.button-pressme -text "Press me" -command {
	    echo "That's nice."
	button .f.button-quit -text quit -command {
	    echo "quit"
	pack append .f .f.button-pressme {top fill expand} \\
		       .f.button-quit {top expand}

	# Here is the main loop which receives and sends commands
	# to wish.
	while () {
	    print "Wish sais: <$_>\n";
	    if (/^quit/) { print WISH "destroy .\n"; last; }
    } elsif (defined $pid) {
	open(STDOUT, ">&W0");
	open(STDIN, ">&W0");
	select(STDOUT); $| = 1;
	exec "$wishbin --";
    } else {
	die "fork error: $!\n";

3.10) Can I dynamically load C user routines?

Yes -- dynamic loading comes with the distribution. That means that you no longer need 18 different versions of fooperl floating around. In fact, all of perl can be stuck into a library and then your /usr/local/bin/perl binary reduced to just 50k or so. See DynaLoader(3pm) for details.

In perl4, the answer was kinda. One package has been released that does this, by Roberto Salama*. He writes:

Here is a version of dylperl, dynamic linker for perl. The code here is based on Oliver Sharp's May 1993 article in Dr. Dobbs Journal (Dynamic Linking under Berkeley UNIX).

  dyl.c - code extracted from Oliver Sharp's article

  hash.c - Berkeley's hash functions, should use perl's but
	   could not be bothered

dylperl.c - perl usersubs
  user.c - userinit function

sample.c - sample code to be dyl'ed
sample2.c -          " - sample perl script that dyl's sample*.o

The Makefile assumes that uperl.o is in /usr/local/src/perl/... You will probably have to change this to reflect your installation. Other than that, just type 'make'...

The idea behind being able to dynamically link code into perl is that the linked code should become perl functions, i.e. they can be invoked as &foo(...). For this to happen, the incrementally loaded code must use the perl stack, look at sample.c to get a better idea.

The few functions that make up this package are outlined below.

&dyl("file.o"): dynamically link file.o. All functions 
   and non-static variables become visible from within perl. This
   function returns a pointer to an internal hash table corresponding
   to the symbol table of the newly loaded code.

eg: $ht = &dyl("sample.o")

This function can also be called with the -L and -l ld options.

   eg: $ht = &dyl("sample2.o", "-L/usr/lib", "-lm")
       will also pick up the math library if sample.o
       accesses any symbols there.

&dyl_find("func"): find symbol 'func' and return its symbol table entry

&dyl_functions($ht): print the contents of the internal hash table
&dyl_print_symbols($f): prints the contents of the symbol returned by

There is very little documentation, maybe something to do for a future release. The files sample.o, and sample2.o contain code to be incrementally loaded, is the test perl script.

Comments are welcome. I submit this code for public consumption and, basically, am not responsible for it in any way.

3.11) What is undump and where can I get it?

The undump program comes from the TeX distribution. If you have TeX, then you may have a working undump. If you don't, and you can't get one, AND you have a GNU emacs working on your machine that can clone itself, then you might try taking its unexec() function and compiling Perl with -DUNEXEC, which will make Perl call unexec() instead of abort(). You'll have to add unexec.o to the objects line in the Makefile. If you succeed, post to comp.lang.perl.misc about your experience so others can benefit from it.

If you have a version of undump that works with Perl, please submit its anon-FTP whereabouts to the FAQ maintainer.

3.12) How can I get #!perl to work under MS-DOS?

John Dallman* has written a program ``#!perl.exe" which will do this. It is available through anonymous ftp from in the directory /pub/msdos/perl/ This program works by finding the script and perl.exe, building a command line and running perl.exe as a child process. For more information on this, contact John directly.

3.13) Can I write useful perl programs on the command line?

Sure, if they're simple enough. Of course, for most programs, you'll enter them in a file and call perl on them from your shell. That way you can go into the hack/execute/debug cycle. But there are plenty of useful one-liner: see below. (Things marked perl5 need to be run from v5.000 or better, but the rest don't care.)

# what's octal value of random char (":" in this case)?
perl -e 'printf "%#o\n", ord(shift)' ":"

# sum first and last fields
perl -lane 'print $F[0] + $F[1]'

# strip high bits
perl -pe 'tr/\200-\377/\000-\177/'

# find text files
perl -le 'for(@ARGV) {print if -f && -T}' *

# trim newsrc
perl5 -i.old -pe 's/!.*?(\d+)$/! 1-$1/' ~/.newsrc

# cat a dbmfile
perl -e 'dbmopen(%f,shift,undef);while(($k,$v)=each%f){print "$k:\
	$v\n"}' /etc/aliases 

# remove comments from C program
perl5 -0777 -pe 's{/\*.*?\*/}{}gs' foo.c

# make file a month younger than today, defeating reaper daemons
perl -e '$X=24*60*60; utime(time(),time() + 30 * $X,@ARGV)' *

# find first unused uid
perl5 -le '$i++ while getpwuid($i); print $i'

# find first unused uid after 100, even with perl4
perl -le '$i = 100; $i++ while ($x) = getpwuid($i); print $i'

# detect pathetically insecurable systems
perl5 -le 'use POSIX; print "INSECURE" unless sysconf(_PC_CHOWN_RESTRICTED)'

# display reasonable manpath
echo $PATH | perl5 -nl -072 -e '

Ok, the last one was actually an obfuscate perl entry. :-)

3.14) What's a "closure"?

(Larry wrote) This is a notion out of the Lisp world that says if you define an anonymous function in a particular lexical context, it pretends to run in that context even when it's called outside of the context.

In human terms, it's a funny way of passing arguments to a subroutine when you define it as well as when you call it. It's useful for setting up little bits of code to run later, such as callbacks. You can even do object-oriented stuff with it, though Perl provides a different mechanism to do that already.

You can also think of it as a way to write a subroutine template without using eval.

Here's a small example of how this works:

    sub newprint {
	my $x = shift;
	return sub { my $y = shift; print "$x, $y!\n"; };
    $h = newprint("Howdy");
    $g = newprint("Greetings");

    # Time passes...


This prints:

    Howdy, world!
    Greetings, earthlings!

Note particularly that $x continues to refer to the value passed into newprint() despite the fact that the my $x has seemingly gone out of scope by the time the anonymous subroutine runs. That's what closure is all about.

This only applies to lexical variables, by the way. Dynamic variables continue to work as they have always worked. Closure is not something that most Perl programmers need trouble themselves about to begin with.

General Programming, Regexps, and I/O

4.1) What are all these $@%*<> signs and how do I know when to use them?
4.2) Why do Perl operators have different precedence than C operators?
4.3) What's the difference between dynamic and static (lexical) scoping?
4.4) What's the difference between deep and shallow binding?
4.5) How can I manipulate fixed-record-length files?
4.6) How can I make a file handle local to a subroutine?
4.7) How can I sleep or alarm for under a second?
4.8) How can I do an atexit() or setjmp()/longjmp()? (Exception handling)
4.9) How can I catch signals?
4.10) Why isn't my octal data interpretted correctly?
4.11) How can I compare two date strings?
4.12) How can I find the Julian Day?
4.13) Does perl have a round function? What about ceil() and floor()?
4.14) What's the fastest way to code up a given task in perl?
4.15) Do I always/never have to quote my strings or use semicolons?
4.16) What is variable suicide and how can I prevent it?
4.17) What does ``Malformed command links'' mean?
4.18) How can I set up a footer format to be used with write()?
4.19) Why does my program keep growing in size?
4.20) Can I do RPC?
4.21) Why doesn't my sockets program work under System V (Solaris)? What does the error message ``Protocol not supported'' mean?
4.22) How can I quote a variable to use in a regexp?
4.23) How can I change the first N letters of a string?
4.24) How can I count the number of occurrences of a substring within a string?
4.25) Can I use Perl regular expressions to match balanced text?
4.26) What does it mean that regexps are greedy? How can I get around it?
4.27) How do I use a regular expression to strip C style comments from a file?
4.28) How can I split a [character] delimited string except when inside [character]?
4.29) Why doesn't local($foo) = <FILE>; work right?
4.30) How can I detect keyboard input without reading it?
4.31) How can I read a single character from the keyboard under UNIX and DOS?
4.32) How can I get input from the keyboard without it echoing to the screen?
4.33) Is there any easy way to strip blank space from the beginning/end of a string?
4.34) How can I output my numbers with commas added?
4.35) How do I expand tags in a string?
4.36) What's wrong with grep in a void context?

4.1) What are all these $@%*<> signs and how do I know when to use them?

Those are type specifiers:

See the question on arrays of arrays for more about Perl pointers.

While there are a few places where you don't actually need these type specifiers, except for files, you should always use them. Note that <FILE> is NOT the type specifier for files; it's the equivalent of awk's getline function, that is, it reads a line from the handle FILE. When doing open, close, and other operations besides the getline function on files, do NOT use the brackets.

Beware of saying:

    $foo = BAR;
Which wil be interpreted as
    $foo = 'BAR';
and not as
    $foo = ;
If you always quote your strings, you'll avoid this trap.

Normally, files are manipulated something like this (with appropriate error checking added if it were production code):

    open (FILE, ">/tmp/foo.$$");
    print FILE "string\n";
    close FILE;
If instead of a filehandle, you use a normal scalar variable with file manipulation functions, this is considered an indirect reference to a filehandle. For example,
    $foo = "TEST01";
    open($foo, "file");
After the open, these two while loops are equivalent:

    while (<$foo>) {}
    while () {}
as are these two statements:
    close $foo;
    close TEST01;
but NOT to this:
    while (<$TEST01>) {} # error
    ^ note spurious dollar sign

This is another common novice mistake; often it's assumed that
    open($foo, "output.$$");
will fill in the value of $foo, which was previously undefined. This just isn't so -- you must set $foo to be the name of a filehandle before you attempt to open it.

Often people request:

How about changing perl syntax to be more like awk or C? I $$mean @less $-signs = &other *special \%characters?

Larry's answer is:

Then it would be less like the shell. :-)

You'll be pleased to know that I've been trying real hard to get rid of unnecessary punctuation in Perl 5. You'll be displeased to know that I don't think noun markers like $ and @ unnecessary. Not only do they function like case markers do in human language, but they are automatically distinguished within interpolative contexts, and the user doesn't have to worry about different syntactic treatments for variable references within or without such a context.

But the & prefix on verbs is now optional, just as ``do'' is in English. I do hope you do understand what I mean.

For example, you used to have to write this:

    &california || &bust;

It can now be written more cleanly like this:

    california or bust;

Strictly speaking, of course, $ and @ aren't case markers, but number markers. English has mandatory number markers, and people get upset when they doesn't agree.

It were just convenient in Perl (for the shellish interplative reasons mentioned above) to pull the markers out to the front of each noun phrase. Most people seems to like it that way. It certainly seem to make more sense than putting them on the end, like most varieties of BASIC does.

4.2) How come Perl operators have different precedence than C operators?

Actually, they don't; all C operators have the same precedence in Perl as they do in C. The problem is with a class of functions called list operators, e.g. print, chdir, exec, system, and so on. These are somewhat bizarre in that they have different precedence depending on whether you look on the left or right of them. Basically, they gobble up all things on their right. For example,

    unlink $foo, "bar", @names, "others";

will unlink all those file names. A common mistake is to write:

    unlink "a_file" || die "snafu";

The problem is that this gets interpreted as

    unlink("a_file" || die "snafu");

To avoid this problem, you can always make them look like function calls or use an extra level of parentheses:

    unlink("a_file")  || die "snafu";
    (unlink "a_file") || die "snafu";

In perl5, there are low precedence ``and'', ``or'', and ``not'' operators, which bind less tightly than comma. This allows you to write:

    unlink $foo, "bar", @names, "others"    or die "snafu";

Sometimes you actually do care about the return value:

    unless ($io_ok = print("some", "list")) { } 

Yes, print() returns I/O success. That means

    $io_ok = print(2+4) * 5;

returns 5 times whether printing (2+4) succeeded, and

    print(2+4) * 5;
returns the same 5*io_success value and tosses it.

See the perlop(1) man page's section on Precedence for more gory details, and be sure to use the -w flag to catch things like this.

One very important thing to be aware of is that if you start thinking of Perl's $, @, %, and & as just flavored versions of C's * operator, you're going to be sorry. They aren't really operators, per se, and even if you do think of them that way. In C, if you write

then the brackets will bind more tightly than the star, yielding
But in perl, they DO NOT! That's because the ${}, @{}, %{}, and &{} notations (and I suppose the *{} one as well for completeness) aren't actually operators. If they were, you'd be able to write them as *() and that's not feasible. Instead of operators whose precedence is easily understandable, they are instead figments of yacc's grammar. This means that:

is really


(by which I actually mean)


and not


See the difference? If not, check out perlref(1) for gory details.

4.3) What's the difference between dynamic and static (lexical) scoping?

What are my() and local()?
[NOTE: This question refers to perl5 only. There is no my() in perl4]
Scoping refers to visibility of variables. A dynamic variable is created via local() and is just a local value for a global variable, whereas a lexical variable created via my() is more what you're expecting from a C auto. (See also ``What's the difference between deep and shallow binding.'') In general, we suggest you use lexical variables wherever possible, as they're faster to access and easier to understand. The ``use strict vars'' pragma will enforce that all variables are either lexical, or full classified by package name. We strongly suggest that you develop your code with ``use strict;'' and the -w flag. (When using formats, however, you will still have to use dynamic variables.) Here's an example of the difference:

    $myvar = 10;
    $localvar = 10;
    print "Before the sub call - my: $myvar, local: $localvar\n";

    print "After the sub call - my: $myvar, local: $localvar\n";


    sub sub1 {
	my $myvar;
	local $localvar;

	$myvar = 5;     # Only in this block
	$localvar = 20; # Accessible to children

	print "Inside first sub call - my: $myvar, local: $localvar\n";

    sub sub2 {
	print "Inside second sub - my: $myvar, local: $localvar\n";


Notice that the variables declared with my() are visible only within the scope of the block which names them. They are not visible outside of this block, not even in routines or blocks that it calls. local() variables, on the other hand, are visible to routines that are called from the block where they are declared. Neither is visible after the end (the final closing curly brace) of the block at all.

Oh, lexical variables are only available in perl5. Have we mentioned yet that you might consider upgrading? :-)

4.4) What's the difference between deep and shallow binding?

5.000 answer:

This only matters when you're making subroutines yourself, at least so far. This will give you shallow binding:

      my $x = time;
      $coderef = sub { $x };

When you call &$coderef(), it will get whatever dynamic $x happens to be around when invoked. However, you can get the other behaviour this way:

      my $x = time;
      $coderef = eval "sub { \$x }";

Now you'll access the lexical variable $x which is set to the time the subroutine was created. Note that the difference in these two behaviours can be considered a bug, not a feature, so you should in particular not rely upon shallow binding, as it will likely go away in the future. See perlref(1) .

5.001 Answer:

Perl will always give deep binding to functions, so you don't need the eval hack anymore. Furthermore, functions and even formats lexically declared nested within another lexical scope have access to that scope.

    require 5.001;

    sub mkcounter {
    my $start = shift;
	return sub {
	    return ++$start;

    $f1 = mkcounter(10);
    $f2 = mkcounter(20);
    print &$f1(), &$f2(); 
11 21
    print &$f1(), &$f2(), &$f1();
12 22 13

See the question on ``What's a closure?''

4.5) How can I manipulate fixed-record-length files?

The most efficient way is using pack and unpack. This is faster than using substr. Here is a sample chunk of code to break up and put back together again some fixed-format input lines, in this case, from ps.

    # sample input line:
    #   15158 p5  T      0:00 perl /mnt/tchrist/scripts/now-what
    $ps_t = 'A6 A4 A7 A5 A*';
    open(PS, "ps|");
    $_ = ; print;
    while () {
	($pid, $tt, $stat, $time, $command) = unpack($ps_t, $_);
	for $var ('pid', 'tt', 'stat', 'time', 'command' ) {
	    print "$var: <", eval "\$$var", ">\n";
	print 'line=', pack($ps_t, $pid, $tt, $stat, $time, $command),  "\n";

4.6) How can I make a file handle local to a subroutine?

You must use the type-globbing *VAR notation. Here is some code to cat an include file, calling itself recursively on nested local include files (i.e. those with #include "file", not #include ):

sub cat_include {
    local($name) = @_;
    warn "\n";
    if (!open (FILE, $name)) {
	    warn "can't open $name: $!\n";
    while (<FILE>) {
	if (/^#\s*include "([^"]*)"/) {
	} else {
    close FILE;

4.7) How can I call alarm() or usleep() from Perl?

If you want finer granularity than 1 second (as usleep() provides) and have itimers and syscall() on your system, you can use the following. You could also use select().

It takes a floating-point number representing how long to delay until you get the SIGALRM, and returns a floating- point number representing how much time was left in the old timer, if any. Note that the C function uses integers, but this one doesn't mind fractional numbers.

# alarm; send me a SIGALRM in this many seconds (fractions ok)
# tom christiansen <>
sub alarm {
    require '';
    require 'sys/';

    local($ticks) = @_;
    local($isecs, $iusecs, $secs, $usecs);

    local($itimer_t) = 'L4'; # should be &itimer'typedef()

    $secs = int($ticks);
    $usecs = ($ticks - $secs) * 1e6;

    $out_timer = pack($itimer_t,0,0,0,0);  
    $in_timer  = pack($itimer_t,0,0,$secs,$usecs);

    syscall(&SYS_setitimer, &ITIMER_REAL, $in_timer, $out_timer)
    && die "alarm: setitimer syscall failed: $!";

    ($isecs, $iusecs, $secs, $usecs) = unpack($itimer_t,$out_timer);
    return $secs + ($usecs/1e6);

4.8) How can I do an atexit() or setjmp()/longjmp() in Perl? (Exception handling)

Perl's exception-handling mechanism is its eval operator. You can use eval as setjmp and die as longjmp. Here's an example of Larry's for timed-out input, which in C is often implemented using setjmp and longjmp:

    sub TIMEOUT { die "restart input\n" }
    do { eval { &realcode } } while $@ =~ /^restart input/;
    sub realcode {
      alarm 15;
      $ans = ;
      alarm 0;

Here's an example of Tom's for doing atexit() handling:

    sub atexit { push(@_exit_subs, @_) }
    sub _cleanup { unlink $tmp }
    eval <<'End_Of_Eval';  $here = __LINE__;
	# as much code here as you want

    $oops = $@;  # save error message

    # now call his stuff
    for (@_exit_subs) { &$_() }
    $oops && ($oops =~ s/\(eval\) line (\d+)/$0 .
	" line " . ($1+$here)/e, die $oops);

You can register your own routines via the &atexit function now. You might also want to use the &realcode method of Larry's rather than embedding all your code in the here-is document. Make sure to leave via die rather than exit, or write your own &exit routine and call that instead. In general, it's better for nested routines to exit via die rather than exit for just this reason.

In Perl5, it is easy to set this up because of the automatic processing of per-package END functions. These work much like they would in awk. See perlfunc(1) , perlmod(1) and perlrun(1) .

Eval is also quite useful for testing for system dependent features, like symlinks, or using a user-input regexp that might otherwise blowup on you.

4.9) How do I catch signals in perl?

Perl allows you to trap signals using the %SIG associative array. Using the signals you want to trap as the key, you can assign a subroutine to that signal. The %SIG array will only contain those values which the programmer defines. Therefore, you do not have to assign all signals. For example, to exit cleanly from a ^C:

    $SIG{'INT'} = 'CLEANUP';
    sub CLEANUP {
	print "\n\nCaught Interrupt (^C), Aborting\n";

There are two special ``routines'' for signals called DEFAULT and IGNORE. DEFAULT erases the current assignment, restoring the default value of the signal. IGNORE causes the signal to be ignored. In general, you don't need to remember these as you can emulate their functionality with standard programming features. DEFAULT can be emulated by deleting the signal from the array and IGNORE can be emulated by any undeclared subroutine.

In 5.001, the $SIG{__WARN__} and $SIG{__DIE__} handlers may be used to intercept die() and warn(). For example, here's how you could promote unitialized variables to trigger a fatal rather merely complaining:

    #!/usr/bin/perl -w 
    require 5.001;
    $SIG{__WARN__} = sub {
	if ($_[0] =~ /uninit/) {
	    die $@;
	} else {
	    warn $@;

4.10) Why doesn't Perl interpret my octal data octally?

Perl only understands octal and hex numbers as such when they occur as literals in your program. If they are read in from somewhere and assigned, then no automatic conversion takes place. You must explicitly use oct() or hex() if you want this kind of thing to happen. Actually, oct() knows to interpret both hex and octal numbers, while hex only converts hexadecimal ones. For example:

    print "What mode would you like? ";
    $mode = <STDIN>;
    $mode = oct($mode);
    unless ($mode) {
	print "You can't really want mode 0!\n";
    chmod $mode, $file;

Without the octal conversion, a requested mode of 755 would turn into 01363, yielding bizarre file permissions of --wxrw--wt.

If you want something that handles decimal, octal, and hex input, you could follow the suggestion in the man page and use:

    $val = oct($val) if $val =~ /^0/;

4.11) How can I compare two date strings?

If the dates are in an easily parsed, predetermined format, then you can break them up into their component parts and call &timelocal from the distributed perl library. If the date strings are in arbitrary formats, however, it's probably easier to use the getdate program from the Cnews distribution, since it accepts a wide variety of dates. Note that in either case the return values you will really be comparing will be the total time in seconds as returned by time(). Here's a getdate function for perl that's not very efficient; you can do better than this by sending it many dates at once or modifying getdate to behave better on a pipe. Beware the hardcoded pathname.

    sub getdate {
	local($_) = shift;
	s/-(\d{4})$/+$1/ || s/\+(\d{4})$/-$1/; 
	    # getdate has broken timezone sign reversal!
	$_ = `/usr/local/lib/news/newsbin/getdate '$_'`;

You can also get the GetDate extension module that's actually the C code linked into perl from wherever fine Perl extensions are given away. It's about 50x faster. If you can't find it elsewhere, I usually keep a copy on for ftp, since I (Tom) ported it.

Richard Ohnemus <Rick_Ohnemus@Sterling.COM> actually has a getdate.y for use with the Perl yacc (see question 3.3 "Is there a yacc for Perl?").

You might also consider using these:        - print dates how you want with the sysv +FORMAT method 
date.shar      - routines to manipulate and calculate dates
ftp-chat2.shar - updated version of ftpget. includes library and demo 
getdate.shar   - returns number of seconds since epoch for any given
ptime.shar     - print dates how you want with the sysv +FORMAT method 

You probably want 'getdate.shar'... these and other files can be ftp'd from the /pub/perl/scripts directory on See the README file in the /pub/perl directory for time and the European mirror site details.

4.12) How can I find the Julian Day?

Here's an example of a Julian Date function provided by Thomas R. Kimpton*.


@theJulianDate = ( 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334 );

#****   Return 1 if we are after the leap day in a leap year.       *****
sub leapDay                  
    my($year,$month,$day) = @_;

    if (year % 4) {

if (!(year % 100)) {             # years that are multiples of 100
				     # are not leap years
	if (year % 400) {            # unless they are multiples of 400
    if (month < 2) {
    } elsif ((month == 2) && (day < 29)) {
    } else {
#****   Pass in the date, in seconds, of the day you want the       *****
#****   julian date for.  If your localtime() returns the year day  *****
#****   return that, otherwise figure out the julian date.          *****
sub julianDate      
    my($dateInSeconds) = @_;
    my($sec, $min, $hour, $mday, $mon, $year, $wday, $yday);

    ($sec, $min, $hour, $mday, $mon, $year, $wday, $yday) =
    if (defined($yday)) {
    } else {
	return($theJulianDate[$mon] + $mday + &leapDay($year,$mon,$mday));


print "Today's julian date is: ",&julianDate(time),"\n";

4.13) Does perl have a round function? What about ceil() and floor()?

Perl does not have an explicit round function. However, it is very simple to create a rounding function. Since the int() function simply removes the decimal value and returns the integer portion of a number, you can use

sub round {
    my($number) = shift;
    return int($number + .5);

If you examine what this function is doing, you will see that any number greater than .5 will be increased to the next highest integer, and any number less than .5 will remain the current integer, which has the same effect as rounding.

A slightly better solution, one which handles negative numbers as well, might be to change the return (above) to:

    return int($number + .5 * ($number <=> 0));

which will modify the .5 to be either positive or negative, based on the number passed into it.

If you wish to round to a specific significant digit, you can use the printf function (or sprintf, depending upon the situation), which does proper rounding automatically. See the perlfunc man page for more information on the (s)printf function.

Version 5 includes a POSIX module which defines the standard C math library functions, including floor() and ceil(). floor($num) returns the largest integer not greater than $num, while ceil($num) returns the smallest integer not less than $num. For example:

    use POSIX qw(ceil floor);

    $num = 42.4;  # The Answer to the Great Question (on a Pentium)!

    print "Floor returns: ", floor($num), "\n";
    print "Ceil returns:  ", ceil($num), "\n";
Which prints:
    Floor returns: 42
    Ceil returns:  43

4.14) What's the fastest way to code up a given task in perl?

Post it to comp.lang.perl.misc and ask Tom or Randal a question about it. ;)

Because Perl so lends itself to a variety of different approaches for any given task, a common question is which is the fastest way to code a given task. Since some approaches can be dramatically more efficient that others, it's sometimes worth knowing which is best. Unfortunately, the implementation that first comes to mind, perhaps as a direct translation from C or the shell, often yields suboptimal performance. Not all approaches have the same results across different hardware and software platforms. Furthermore, legibility must sometimes be sacrificed for speed.

While an experienced perl programmer can sometimes eye-ball the code and make an educated guess regarding which way would be fastest, surprises can still occur. So, in the spirit of perl programming being an empirical science, the best way to find out which of several different methods runs the fastest is simply to code them all up and time them. For example:

$COUNT = 10_000; $| = 1;

print "method 1: ";

($u, $s) = times;
for ($i = 0; $i < $COUNT; $i++) {
    # code for method 1
($nu, $ns) = times;
printf "%8.4fu %8.4fs\n", ($nu - $u), ($ns - $s);

print "method 2: ";

($u, $s) = times;
for ($i = 0; $i < $COUNT; $i++) {
    # code for method 2
($nu, $ns) = times;
printf "%8.4fu %8.4fs\n", ($nu - $u), ($ns - $s);

Perl5 includes a new module called You can now simplify the code to use the Benchmarking, like so:

use Benchmark;

timethese($count, {
    Name1 => '...code for method 1...',
    Name2 => '...code for method 2...',
    ... });

It will output something that looks similar to this:

Benchmark: timing 100 iterations of Name1, Name2...
    Name1:  2 secs (0.50 usr 0.00 sys = 0.50 cpu)
    Name2:  1 secs (0.48 usr 0.00 sys = 0.48 cpu)

For example, the following code will show the time difference between three different ways of assigning the first character of a string to a variable:

use Benchmark;
timethese(100000, {
    'regex1' => '$str="ABCD"; $str =~ s/^(.)//; $ch = $1',
    'regex2' => '$str="ABCD"; $str =~ s/^.//; $ch = $&',
    'substr' => '$str="ABCD"; $ch=substr($str,0,1); substr($str,0,1)="",

The results will be returned like this:

Benchmark: timing 100000 iterations of regex1, regex2, substr...
   regex1: 11 secs (10.80 usr   0.00 sys =  10.80 cpu)
   regex2: 10 secs (10.23 usr   0.00 sys =  10.23 cpu)
   substr:  7 secs ( 5.62 usr   0.00 sys =   5.62 cpu)

For more specific tips, see the section on Efficiency in the ``Other Oddments'' chapter at the end of the Camel Book.

4.15) Do I always/never have to quote my strings or use semicolons?

You don't have to quote strings that can't mean anything else in the language, like identifiers with any upper-case letters in them. Therefore, it's fine to do this:

    $SIG{INT} = Timeout_Routine;
    @Days = (Sun, Mon, Tue, Wed, Thu, Fri, Sat, Sun);

but you can't get away with this:

    $foo{while} = until;

in place of

    $foo{'while'} = 'until';

The requirements on semicolons have been increasingly relaxed. You no longer need one at the end of a block, but stylistically, you're better to use them if you don't put the curly brace on the same line:

    for (1..10) { print }

is ok, as is

    @nlist = sort { $a <=> $b } @olist;

but you probably shouldn't do this:

    for ($i = 0; $i < @a; $i++) {
	print "i is $i\n"  # <-- oops!

because you might want to add lines later, and anyway, it looks funny. :-)

Actually, I lied. As of 5.001, there are two autoquoting contexts:

This                    is like this
------------            ---------------
$foo{line}              $foo{"line"}
bar => stuff            "bar" => stuff

4.16) What is variable suicide and how can I prevent it?

Variable suicide is a nasty side effect of dynamic scoping and the way variables are passed by reference. If you say

    $x = 17;
    sub munge {
	local($myvar) = $_[0];

Then you have just clobbered $_[0]! Why this is occurring is pretty heavy wizardry: the reference to $x stored in $_[0] was temporarily occluded by the previous local($x) statement (which, you're recall, occurs at run-time, not compile-time). The work around is simple, however: declare your formal parameters first:

    sub munge {
	local($myvar) = $_[0];

That doesn't help you if you're going to be trying to access @_ directly after the local()s. In this case, careful use of the package facility is your only recourse.

Another manifestation of this problem occurs due to the magical nature of the index variable in a foreach() loop.

    @num = 0 .. 4;
    print "num begin  @num\n";
    foreach $m (@num) { &ug }
    print "num finish @num\n";
    sub ug {
	local($m) = 42;
	print "m=$m  $num[0],$num[1],$num[2],$num[3]\n";

Which prints out the mysterious:
    num begin  0 1 2 3 4
    m=42  42,1,2,3
    m=42  0,42,2,3
    m=42  0,1,42,3
    m=42  0,1,2,42
    m=42  0,1,2,3
    num finish 0 1 2 3 4

What's happening here is that $m is an alias for each element of @num. Inside &ug, you temporarily change $m. Well, that means that you've also temporarily changed whatever $m is an alias to!! The only workaround is to be careful with global variables, using packages, and/or just be aware of this potential in foreach() loops.

The perl5 static autos via my() do not exhibit this problem.

4.17) What does ``Malformed command links'' mean?

This is a bug in 4.035. While in general it's merely a cosmetic problem, it often comanifests with a highly undesirable coredumping problem. Programs known to be affected by the fatal coredump include plum and pcops. This bug has been fixed since 4.036. It did not resurface in 5.001.

4.18) How can I set up a footer format to be used with write()?

While the $^ variable contains the name of the current header format, there is no corresponding mechanism to automatically do the same thing for a footer. Not knowing how big a format is going to be until you evaluate it is one of the major problems.

If you have a fixed-size footer, you can get footers by checking for line left on page ($-) before each write, and printing the footer yourself if necessary.

Another strategy is to open a pipe to yourself, using open(KID, "|-") and always write()ing to the KID, who then postprocesses its STDIN to rearrange headers and footers however you like. Not very convenient, but doable.

See the perlform(1) man page for other tricks.

4.19) Why does my Perl program keep growing in size?

This is caused by a strange occurrence that often dubbed ``feeping creaturism''. Larry is always adding one more feature, always getting Perl to handle one more problem. Hence, it keeps growing. Once you've worked with perl long enough, you will probably start to do the same thing. You will then notice this problem as you see your scripts becoming larger and larger.

Oh, wait... you meant a currently running program and its stack size. Mea culpa, I misunderstood you. ;) While there may be a real memory leak in the Perl source code or even whichever malloc() you're using, common causes are incomplete eval()s or local()s in loops.

An eval() which terminates in error due to a failed parsing will leave a bit of memory unusable.

A local() inside a loop:

    for (1..100) {

will build up 100 versions of @array before the loop is done. The work-around is:

    for (1..100) {
	undef @array;

This local array behaviour has been fixed for perl5, but a failed eval() still leaks.

One other possibility, due to the way reference counting works, is when you've introduced a circularity in a data structure that would normally go out of scope and be unreachable. For example:

    sub oops {
	my $x;
	$x = \$x;

When $x goes out of scope, the memory can't be reclaimed, because there's still something point to $x (itself, in this case). A full garbage collection system could solve this, but at the cost of a great deal of complexity in perl itself and some inevitable performance problems as well. If you're making a circular data structure that you want freed eventually, you'll have to break the self-reference links yourself.

4.20) Can I do RPC in Perl?

Yes, you can, since Perl has access to sockets. An example of the rup program written in Perl can be found in the script at the scripts archive on I warn you, however, that it's not a pretty sight, as it's used nothing from h2ph or c2ph, so everything is utterly hard-wired.

4.21) Why doesn't my sockets program work under System V (Solaris)? What does the error message ``Protocol not supported'' mean?

Some System V based systems, notably Solaris 2.X, redefined some of the standard socket constants. Since these were constant across all architectures, they were often hardwired into the perl code. The ``proper'' way to deal with this is to make sure that you run h2ph against sys/socket.h, require that file and use the symbolic names (SOCK_STREAM, SOCK_DGRAM, SOCK_RAW, SOCK_RDM, and SOCK_SEQPACKET).

Note that even though SunOS 4 and SunOS 5 are binary compatible, these values are different, and require a different for each OS.

Under version 5, you can also ``use Socket'' to get the proper values.

4.22) How can I quote a variable to use in a regexp?

From the manual:

    $pattern =~ s/(\W)/\\$1/g;

Now you can freely use /$pattern/ without fear of any unexpected meta- characters in it throwing off the search. If you don't know whether a pattern is valid or not, enclose it in an eval to avoid a fatal run- time error.

Perl5 provides a vastly improved way of doing this. Simply use the new quotemeta character (\Q) within your variable.

4.23) How can I change the first N letters of a string?

Remember that the substr() function produces an lvalue, that is, it may be assigned to. Therefore, to change the first character to an S, you could do this:

    substr($var,0,1) = 'S';

This assumes that $[ is 0; for a library routine where you can't know $[, you should use this instead:

    substr($var,$[,1) = 'S';
While it would be slower, you could in this case use a substitute:

    $var =~ s/^./S/;
But this won't work if the string is empty or its first character is a newline, which ``.'' will never match. So you could use this instead:

    $var =~ s/^[^\0]?/S/;

To do things like translation of the first part of a string, use substr, as in:

    substr($var, $[, 10) =~ tr/a-z/A-Z/;

If you don't know the length of what to translate, something like this works:

    /^(\S+)/ && substr($_,$[,length($1)) =~ tr/a-z/A-Z/;
For some things it's convenient to use the /e switch of the substitute operator:

    s/^(\S+)/($tmp = $1) =~ tr#a-z#A-Z#, $tmp/e

although in this case, it runs more slowly than does the previous example.

4.24) How can I count the number of occurrences of a substring within a string?

If you want a count of a certain character (X) within a string, you can use the tr/// function like so:

    $count = ($string =~ tr/X//);
    print "There are $count Xs in the string";

This is fine if you are just looking for a single character. However, if you are trying to count multiple character substrings within a larger string, tr/// won't work. What you can do is wrap a while loop around a pattern match.

    $string="-9 55 48 -2 23 -76 4 14 -44";
    $count++ while $string =~ /-\d+/g;
    print "There are $count negative numbers in the string";

4.25) Can I use Perl regular expressions to match balanced text?

No, or at least, not by themselves.

Regexps just aren't powerful enough. Although Perl's patterns aren't strictly regular because they do backreferencing (the \1 notation), you still can't do it. You need to employ auxiliary logic. A simple approach would involve keeping a bit of state around, something vaguely like this (although we don't handle patterns on the same line):

while(<>) {
    if (/pat1/) {
        if ($inpat++ > 0) { warn "already saw pat1" } 
    if (/pat2/) {
        if (--$inpat < 0) { warn "never saw pat1" } 

A rather more elaborate subroutine to pull out balanced and possibly nested single chars, like ` and ', { and }, or ( and ) can be found on in /pub/perl/scripts/pull_quotes.

4.26) What does it mean that regexps are greedy? How can I get around it?

The basic idea behind regexps being greedy is that they will match the maximum amount of data that they can, sometimes resulting in incorrect or strange answers.

For example, I recently came across something like this:

    $_="this (is) an (example) of multiple parens";
    while ( m#\((.*)\)#g ) {
	print "$1\n";

This code was supposed to match everything between a set of parentheses. The expected output was:


However, the backreference ($1) ended up containing "is) an (example", clearly not what was intended.

In perl4, the way to stop this from happening is to use a negated group. If the above example is rewritten as follows, the results are correct:

    while ( m#\(([^)]*)\)#g ) {

In perl5 there is a new minimal matching metacharacter, '?'. This character is added to the normal metacharacters to modify their behaviour, such as ``*?'', ``+?'', or even ``??''. The example would now be written in the following style:

    while (m#\((.*?)\)#g )

Hint: This new operator leads to a very elegant method of stripping comments from C code:


4.27) How do I use a regular expression to strip C style comments from a file?

Since we're talking about how to strip comments under perl5, now is a good time to talk about doing it in perl4. Since comments can be embedded in strings, or look like function prototypes, care must be taken to ignore these cases. Jeffrey Friedl* proposes the following two programs to strip C comments and C++ comments respectively: C comments:

$/ = undef;
$_ = <>; 


C++ comments:

$/ = undef;
$_ = <>;
s#//(.*)|/\*[^*]*\*+([^/*][^*]*\*+)*/|"(\\.|[^"\\])*"|'(\\.|[^'\\])*'|[^/"']+#  $1 ? "/*$1 */" : $& #ge;

(Yes, Jeffrey says, those are complete programs to strip comments correctly.)

4.28) How can I split a [character] delimited string except when inside [character]?

I'm trying to split a string that is comma delimited into its different fields. I could easily use split(/,/), except that I need to not split if the comma is inside quotes. For example, my data file has a line like this:

    SAR001,"","Cimetrix, Inc","Bob Smith","CAM",N,8,1,0,7,"Error, Core Dumped"

Due to the restriction of the quotes, this is a fairly complex solution. However, we thankfully have Jeff Friedl* to handle these for us. He suggests (assuming that your data is contained in the special variable $_):

    undef @field;
    push(@fields, defined($1) ? $1:$3) 
	while m/"([^"\\]*(\\.[^"\\]*)*)"|([^,]+)/g;

4.29) Why doesn't local($foo) = <FILE> work right?

Well, it does. The thing to remember is that local() provides an array context, and that the <FILE> syntax in an array context will read all the lines in a file. To work around this, use:

    $foo = <FILE>;

You can use the scalar() operator to cast the expression into a scalar context:

    local($foo) = scalar(<FILE>);

4.30) How can I detect keyboard input without reading it?

You should check out the Frequently Asked Questions list in comp.unix.* for things like this: the answer is essentially the same. It's very system dependent. Here's one solution that works on BSD systems:

    sub key_ready {
	local($rin, $nfd);
	vec($rin, fileno(STDIN), 1) = 1;
	return $nfd = select($rin,undef,undef,0);

Under perl5, you should look into getting the ReadKey extension from your regular perl archive.

4.31) How can I read a single character from the keyboard under UNIX and DOS?

A closely related question to the no-echo question below is how to input a single character from the keyboard. Again, this is a system dependent operation. As with the previous question, you probably want to get the ReadKey extension. The following code may or may not help you. It should work on both SysV and BSD flavors of UNIX:

$BSD = -f '/vmunix';
if ($BSD) {
    system "stty cbreak /dev/tty 2>&1";
else {
    system "stty", '-icanon',
    system "stty", 'eol', "\001"; 

$key = getc(STDIN);

if ($BSD) {
    system "stty -cbreak /dev/tty 2>&1";
else {
    system "stty", 'icanon';
    system "stty", 'eol', '^@'; # ascii null
print "\n";

You could also handle the stty operations yourself for speed if you're going to be doing a lot of them. This code works to toggle cbreak and echo modes on a BSD system:

sub set_cbreak { # &set_cbreak(1) or &set_cbreak(0)
    local($on) = $_[0];
    require 'sys/';
    $sgttyb_t   = 'C4 S' unless $sgttyb_t;  # c2ph: &sgttyb'typedef()
    ioctl(STDIN,&TIOCGETP,$sgttyb) || die "Can't ioctl TIOCGETP: $!";

    @ary = unpack($sgttyb_t,$sgttyb);
    if ($on) {
	$ary[4] |= &CBREAK;
	$ary[4] &= ~&ECHO;
    } else {
	$ary[4] &= ~&CBREAK;
	$ary[4] |= &ECHO;
    $sgttyb = pack($sgttyb_t,@ary);
    ioctl(STDIN,&TIOCSETP,$sgttyb) || die "Can't ioctl TIOCSETP: $!";

Note that this is one of the few times you actually want to use the getc() function; it's in general way too expensive to call for normal I/O. Normally, you just use the <FILE> syntax, or perhaps the read() or sysread() functions.

For perspectives on more portable solutions, use anon ftp to retrieve the file /pub/perl/info/keypress from

Under Perl5, with William Setzer's Curses module, you can call &Curses::cbreak() and &Curses::nocbreak() to turn cbreak mode on and off. You can then use getc() to read each character. This should work under both BSD and SVR systems. If anyone can confirm or deny (especially William), please contact the maintainers.

For DOS systems, Dan Carson <dbc@tc.fluke.COM> reports:

To put the PC in ``raw'' mode, use ioctl with some magic numbers gleaned from msdos.c (Perl source file) and Ralf Brown's interrupt list (comes across the net every so often):

    $old_ioctl = ioctl(STDIN,0,0);     # Gets device info
    $old_ioctl &= 0xff;
    ioctl(STDIN,1,$old_ioctl | 32);    # Writes it back, setting bit 5

Then to read a single character:

    sysread(STDIN,$c,1);               # Read a single character

And to put the PC back to ``cooked'' mode:

    ioctl(STDIN,1,$old_ioctl);         # Sets it back to cooked mode.

So now you have $c. If ord($c) == 0, you have a two byte code, which means you hit a special key. Read another byte with sysread(STDIN,$c,1), and that value tells you what combination it was according to this table:

# PC 2-byte keycodes = ^@ + the following:

# HEX     KEYS
# ---     ----
# 0F      SHF TAB
# 3B-44   F1-F10
# 47-49   HOME,UP,PgUp
# 4B      LEFT
# 4D      RIGHT
# 4F-53   END,DOWN,PgDn,Ins,Del
# 54-5D   SHF F1-F10
# 5E-67   CTR F1-F10
# 68-71   ALT F1-F10
# 78-83   ALT 1234567890-=
# 84      CTR PgUp

This is all trial and error I did a long time ago, I hope I'm reading the file that worked.

4.32) How can I get input from the keyboard without it echoing to the screen?

Terminal echoing is generally handled directly by the shell. Therefore, there is no direct way in perl to turn echoing on and off. However, you can call the command "stty [-]echo". The following will allow you to accept input without it being echoed to the screen, for example as a way to accept passwords (error checking deleted for brevity):

    print "Please enter your password: '';
    system("stty -echo");
    print "\n";
    system("stty echo");

Again, under perl 5, you can use Curses and call &Curses::noecho() and &Curses::echo() to turn echoing off and on. Or, there's always the ReadKey extension.

4.33) Is there any easy way to strip blank space from the beginning/end of a string?

Yes, there is. Using the substitution command, you can match the blanks and replace it with nothing. For example, if you have the string " String " you can use this:

    s/^\s*(.*?)\s*$/$1/;    # perl5 only!

    s/^\s+|\s+$//g;         # perl4 or perl5

or even

    s/^\s+//; s/\s+$//;

Note however that Jeffrey Friedl* says these are only good for shortish strings. For longer strings, and worse-case scenarios, they tend to break-down and become inefficient.

For the longer strings, he suggests using either

    $_ = $1 if m/^\s*((.*\S)?)/;

It should also be noted that for generally nice strings, these tend to be noticably slower than the simple ones above. It is suggested that you use whichever one will fit your situation best, understanding that the first examples will work in roughly ever situation known even if slow at times.

4.34) How can I print out a number with commas into it?

This one will do it for you:

sub commify {
    local($_) = shift;
    1 while s/^(-?\d+)(\d{3})/$1,$2/;
    return $_;

$n = 23659019423.2331;
print "GOT: ", &commify($n), "\n";

GOT: 23,659,019,423.2331

The reason you can't just do

Is that you have to put the comma in and then recalculate anything. Some substitutions need to work this way. See the question on expanding tabs for another such.

4.35) How do I expand tabs in a string?

    1 while s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e;

You could have written that

    while (s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e) {
	# spin, spin, spin, ....

Placed in a function:

sub tab_expand {
    local($_) = shift;
    1 while s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e;
    return $_;

This is especially important when you're working going to unpack an ascii string that might have tabs in it. Otherwise you'll be off on the byte count. For example:

$NG = "/usr/local/lib/news/newsgroups";
open(NG, "< $NG") || die "can't open $NG: $!";
while () {
    chop;  # chomp would be better, but it's only perl5
    # now for the darned tabs in the newsgroups file
    1 while s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e;
    ($ng, $desc) = unpack("A24 A*", $_);
    if (length($ng) == 24) {
        $desc =~ s/^(\S+)\s*//;
        $ng .= $1;

4.36) What's wrong with grep() or map() in a void context?

Well, nothing precisely, but it's not a good way to write maintainable code. It's just fine to use grep when you want an answer, like

    @bignums = grep ($_ > 100, @allnums);
    @triplist = map {$_ * 3} @allnums;

But using it in a void context like this:

    grep{ $_ *= 3, @nums);

Is using it for its side-effects, and side-effects can be mystifying. There's no void grep that's not better written as a for() loop:

    for (@nums) { $_ *= 3 } 

In the same way, a ?: in a void context is considered poor form:

    fork ? wait : exec $prog;

When you can write it this way:

    if (fork) {
    } else {
	exec $prog;
	die "can't exec $prog: $!";


Of course, using ?: in expressions is just what it's made for, and just fine (but try not to nest them.).

Remember that the most important things in almost any program are, and in this order:

  1. correctness
  2. maintainability
  3. efficiency
Notice at no point did cleverness enter the picture.

On the other hand, if you're just trying write JAPHs (aka Obfuscated Perl entries), or write ugly code, you would probably invert these :-)

  1. cleverness
  2. efficiency
  3. maintainability
  4. correctness

Arrays and Shell and External Program Interactions

5.1) What is the difference between $array[1] and @array[1]?
5.2) How can I make an array of arrays or other recursive data types?
5.3) How can I make an array of structures containing various data types?
5.4) How can I extract just the unique elements of an array?
5.5) How can I tell whether an array contains a certain element?
5.6) How can I sort an associative array by value instead of by key?
5.7) How can I know how many entries are in an associative array?
5.8) What's the difference between "delete" and "undef" with %arrays?
5.9) Why don't backticks work as they do in shells?
5.10) Why does my converted awk/sed/sh script run more slowly in perl?
5.11) How can I call my system's unique C functions from perl?
5.12) Where do I get the include files to do ioctl() or syscall()? [h2ph]
5.13) Why do setuid perl scripts complain about kernel problems?
5.14) How can I open a pipe both to and from a command?
5.15) How can I capture STDERR from an external command?
5.16) Why doesn't open() return an error when a pipe open fails?
5.17) Why can't my script read from STDIN after I gave it ^D (EOF)?
5.18) How can I translate tildes (~) in a filename?
5.19) How can I convert my shell script to perl?
5.20) Can I use perl to run a telnet or ftp session?
5.21) Why do I sometimes get an "Argument list to long" when I use <*>?
5.22) How do I do a "tail -f" in perl?
5.23) Is there a way to hide perl's command line from programs such as "ps"?
5.24) I {changed directory, modified my environment} in a perl script. How come the change disappeared when I exited the script? How do I get my changes to be visible?
5.25) How can I pass a filehandle to a function, or make a list of filehandles?
5.26) How can I open a file with a leading ">" or trailing blanks?
5.27) How can I tell if a variable is tainted?

5.1) What is the difference between $array[1] and @array[1]?

Always make sure to use a $ for single values and @ for multiple ones. Thus element 2 of the @foo array is accessed as $foo[1], not @foo[1], which is a list of length one (not a scalar), and is a fairly common novice mistake. Sometimes you can get by with @foo[1], but it's not really doing what you think it's doing for the reason you think it's doing it, which means one of these days, you'll shoot yourself in the foot; ponder for a moment what these will really do:

    @foo[0] = `cmd args`;
    @foo[1] = <FILE>
Just always say $foo[1] and you'll be happier.

This may seem confusing, but try to think of it this way: you use the character of the type which you want back. You could use @foo[1..3] for a slice of three elements of @foo, or even @foo{A,B,C} for a slice of of %foo. This is the same as using ($foo[1], $foo[2], $foo[3]) and ($foo{A}, $foo{B}, $foo{C}) respectively. In fact, you can even use lists to subscript arrays and pull out more lists, like @foo[@bar] or @foo{@bar}, where @bar is in both cases presumably a list of subscripts.

5.2) How can I make an array of arrays or other recursive data types?

In Perl5, it's quite easy to declare these things. For example

    @A = (
        [ 'ww' .. 'xx'  ], 
        [ 'xx' .. 'yy'  ], 
        [ 'yy' .. 'zz'  ], 
        [ 'zz' .. 'zzz' ], 

And now reference $A[2]->[0] to pull out ``yy''. These may also nest and mix with tables:

    %T = (
        key0, { k0, v0, k1, v1 },   
        key1, { k2, v2, k3, v3 },   
        key2, { k2, v2, k3, [ 'a' .. 'z' ] },    
    Allowing you to reference $T{key2}->{k3}->[3] to pull out 'd'.

Perl4 is infinitely more difficult. Remember that Perl[0..4] isn't about nested data structures. It's about flat ones, so if you're trying to do this, you may be going about it the wrong way or using the wrong tools. You might try parallel arrays with common subscripts.

But if you're bound and determined, you can use the multi-dimensional array emulation of $a{'x','y','z'}, or you can make an array of names of arrays and eval it.

For example, if @name contains a list of names of arrays, you can get at a the j-th element of the i-th array like so:

    $ary = $name[$i];
    $val = eval "\$$ary[$j]";

or in one line

    $val = eval "\$$name[$i][\$j]";

You could also use the type-globbing syntax to make an array of *name values, which will be more efficient than eval. Here @name hold a list of pointers, which we'll have to dereference through a temporary variable.

For example:

    { local(*ary) = $name[$i]; $val = $ary[$j]; }

In fact, you can use this method to make arbitrarily nested data structures. You really have to want to do this kind of thing badly to go this far, however, as it is notationally cumbersome.

Let's assume you just simply have to have an array of arrays of arrays. What you do is make an array of pointers to arrays of pointers, where pointers are *name values described above. You initialize the outermost array normally, and then you build up your pointers from there. For example:

    @w = ( 'ww' .. 'xx' );
    @x = ( 'xx' .. 'yy' );
    @y = ( 'yy' .. 'zz' );
    @z = ( 'zz' .. 'zzz' );

    @ww = reverse @w;
    @xx = reverse @x;
    @yy = reverse @y;
    @zz = reverse @z;

Now make a couple of arrays of pointers to these:

    @A = ( *w, *x, *y, *z );
    @B = ( *ww, *xx, *yy, *zz );

And finally make an array of pointers to these arrays:

    @AAA = ( *A, *B );

To access an element, such as AAA[i][j][k], you must do this:

    local(*foo) = $AAA[$i];
    local(*bar) = $foo[$j];
    $answer = $bar[$k];

Similar manipulations on associative arrays are also feasible.

You could take a look at package posted by Felix Lee*, which lets you simulate vectors and tables (lists and associative arrays) by using type glob references and some pretty serious wizardry.

In C, you're used to creating recursive datatypes for operations like recursive decent parsing or tree traversal. In Perl, these algorithms are best implemented using associative arrays. Take an array called %parent, and build up pointers such that $parent{$person} is the name of that person's parent. Make sure you remember that $parent{'adam'} is 'adam'. :-) With a little care, this approach can be used to implement general graph traversal algorithms as well.

5.3) How do I make an array of structures containing various data types?

This answer will work under perl5 only. Did we mention that you should upgrade? There is a perl4 solution, but you are using perl5 now, anyway, so there's no point in posting it. Right?

The best way to do this is to use an associative array to model your structure, then either a regular array (AKA list) or another associative array (AKA hash, table, or hash table) to store it.

    %foo = (
               'field1'        => "value1",
               'field2'        => "value2",
               'field3'        => "value3",

    @all = ( \%foo, \%bar, ... );

    print $all[0]{'field1'};

Or even

    @all = (
           'field1'        => "value1",
           'field2'        => "value2",
           'field3'        => "value3",
           'field1'        => "value1",
           'field2'        => "value2",
           'field3'        => "value3",

Note that if you want an associative array of lists, you'll want to make assignments like

    $t{$value} = [ @bar ];

And with lists of associative arrays, you'll use

    %{$a[$i]} = %old;

Study these for a while, and in an upcoming FAQ, we'll explain them fully:

    $table{'some key'}    =   @big_list_o_stuff;   # SCARY #0
    $table{'some key'}    =  \@big_list_o_stuff;   # SCARY #1
    @$table{'some key'}   =   @big_list_o_stuff;   # SCARY #2
    @{$table{'some key'}} =   @big_list_o_stuff;   # ICKY RANDALIAN CODE
    $table{'some key'}    = [ @big_list_o_stuff ]; # same, but NICE

And while you're at it, take a look at these:

    $table{"051"}         = $some_scalar;          # SCARY #3
    $table{"0x51"}        = $some_scalar;          # ditto
    $table{051}           = $some_scalar;          # ditto
    $table{0x51}          = $some_scalar;          # ditto
    $table{51}            = $some_scalar;          # ok, i guess
    $table{"51"}          = $some_scalar;          # better

    $table{\@x}           = $some_scalar;          # SCARY #4
    $table{[@x]}          = $some_scalar;          # ditto
    $table{@x}            = $some_scalar;          # SCARY #5 (cf #0)

See perlref(1) for details.

5.4) How can I extract just the unique elements of an array?

There are several possible ways, depending on whether the array is ordered and you wish to preserve the ordering.

a) If @in is sorted, and you want @out to be sorted:

    $prev = 'nonesuch';
    @out = grep($_ ne $prev && (($prev) = $_), @in);

This is nice in that it doesn't use much extra memory, simulating uniq's behavior of removing only adjacent duplicates.

b) If you don't know whether @in is sorted:

    undef %saw;
    @out = grep(!$saw{$_}++, @in);

c) Like (b), but @in contains only small integers:

    @out = grep(!$saw[$_]++, @in);

d) A way to do (b) without any loops or greps:

    undef %saw;
    @saw{@in} = ();
    @out = sort keys %saw;  # remove sort if undesired

e) Like (d), but @in contains only small positive integers:

    undef @ary;
    @ary[@in] = @in;
    @out = sort @ary;

5.5) How can I tell whether an array contains a certain element?

There are several ways to approach this. If you are going to make this query many times and the values are arbitrary strings, the fastest way is probably to invert the original array and keep an associative array lying about whose keys are the first array's values.

    @blues = ('turquoise', 'teal', 'lapis lazuli');
    undef %is_blue;
    for (@blues) { $is_blue{$_} = 1; }

Now you can check whether $is_blue{$some_color}. It might have been a good idea to keep the blues all in an assoc array in the first place.

If the values are all small integers, you could use a simple indexed array. This kind of an array will take up less space:

    @primes = (2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31);
    undef @is_tiny_prime;
    for (@primes) { $is_tiny_prime[$_] = 1; }

Now you check whether $is_tiny_prime[$some_number].

If the values in question are integers instead of strings, you can save quite a lot of space by using bit strings instead:

    @articles = ( 1..10, 150..2000, 2017 );
    undef $read;
    grep (vec($read,$_,1) = 1, @articles);
    Now check whether vec($read,$n,1) is true for some $n.

5.6) How do I sort an associative array by value instead of by key?

You have to declare a sort subroutine to do this, or use an inline function. Let's assume you want an ASCII sort on the values of the associative array %ary. You could do so this way:

    foreach $key (sort by_value keys %ary) {
        print $key, '=', $ary{$key}, "\n";
    sub by_value { $ary{$a} cmp $ary{$b}; }

If you wanted a descending numeric sort, you could do this:

    sub by_value { $ary{$b} <=> $ary{$a}; }

You can also inline your sort function, like this, at least if you have a relatively recent patchlevel of perl4 or are running perl5:

    foreach $key ( sort { $ary{$b} <=> $ary{$a} } keys %ary ) {
        print $key, '=', $ary{$key}, "\n";

If you wanted a function that didn't have the array name hard-wired into it, you could so this:

    foreach $key (&sort_by_value(*ary)) {
        print $key, '=', $ary{$key}, "\n";
    sub sort_by_value {
        local(*x) = @_;
        sub _by_value { $x{$a} cmp $x{$b}; } 
        sort _by_value keys %x;

If you want neither an alphabetic nor a numeric sort, then you'll have to code in your own logic instead of relying on the built-in signed comparison operators ``cmp'' and ``<=>''.

Note that if you're sorting on just a part of the value, such as a piece you might extract via split, unpack, pattern-matching, or substr, then rather than performing that operation inside your sort routine on each call to it, it is significantly more efficient to build a parallel array of just those portions you're sorting on, sort the indices of this parallel array, and then to subscript your original array using the newly sorted indices. This method works on both regular and associative arrays, since both @ary[@idx] and @ary{@idx} make sense. See page 245 in the Camel Book on ``Sorting an Array by a Computable Field'' for a simple example of this.

For example, here's an efficient case-insensitive comparison:

    @idx = ();
    for (@data) { push (@idx, "\U$_") }
    @sorted = @data[ sort { $idx[$a] cmp $idx[$b] } 0..$#data];

5.7) How can I know how many entries are in an associative array?

While the number of elements in a @foobar array is simply @foobar when used in a scalar, you can't figure out how many elements are in an associative array in an analogous fashion. That's because %foobar in a scalar context returns the ratio (as a string) of number of buckets filled versus the number allocated. For example, scalar(%ENV) might return ``20/32''. While perl could in theory keep a count, this would break down on associative arrays that have been bound to dbm files.

However, while you can't get a count this way, one thing you can use it for is to determine whether there are any elements whatsoever in the array, since ``if (%table)'' is guaranteed to be false if nothing has ever been stored in it.

As of perl4.035, you can says

    $count = keys %ARRAY;

keys() when used in a scalar context will return the number of keys, rather than the keys themselves.

5.8) What's the difference between "delete" and "undef" with %arrays?

Pictures help... here's the %ary table:

          keys  values
        |  a   |  3   |
        |  x   |  7   |
        |  d   |  0   |
        |  e   |  2   |

And these conditions hold

        $ary{'a'}                       is true
        $ary{'d'}                       is false
        defined $ary{'d'}               is true
        defined $ary{'a'}               is true
        exists $ary{'a'}                is true (perl5 only)
        grep ($_ eq 'a', keys %ary)     is true

If you now say

        undef $ary{'a'}

your table now reads:

          keys  values
        |  a   | undef|
        |  x   |  7   |
        |  d   |  0   |
        |  e   |  2   |

and these conditions now hold; changes in caps:

        $ary{'a'}                       is FALSE
        $ary{'d'}                       is false
        defined $ary{'d'}               is true
        defined $ary{'a'}               is FALSE
        exists $ary{'a'}                is true (perl5 only)
        grep ($_ eq 'a', keys %ary)     is true

Notice the last two: you have an undef value, but a defined key!

Now, consider this:

        delete $ary{'a'}

your table now reads:

          keys  values
        |  x   |  7   |
        |  d   |  0   |
        |  e   |  2   |

and these conditions now hold; changes in caps:

        $ary{'a'}                       is false
        $ary{'d'}                       is false
        defined $ary{'d'}               is true
        defined $ary{'a'}               is false
        exists $ary{'a'}                is FALSE (perl5 only)
        grep ($_ eq 'a', keys %ary)     is FALSE

See, the whole entry is gone!

5.9) Why don't backticks work as they do in shells?

Several reasons. One is because backticks do not interpolate within double quotes in Perl as they do in shells. Let's look at two common mistakes:

     $foo = "$bar is `wc $file`";  # WRONG

This should have been:

     $foo = "$bar is " . `wc $file`;

But you'll have an extra newline you might not expect. This does not work as expected:

$back = `pwd`; chdir($somewhere); chdir($back); # WRONG

Because backticks do not automatically eat trailing or embedded newlines. The chop() function will remove the last character from a string. This should have been:

      chop($back = `pwd`); chdir($somewhere); chdir($back);

You should also be aware that while in the shells, embedding single quotes will protect variables, in Perl, you'll need to escape the dollar signs.

    Shell: foo=`cmd 'safe $dollar'`
    Perl:  $foo=`cmd 'safe \$dollar'`;

5.10) How come my converted awk/sed/sh script runs more slowly in Perl?

The natural way to program in those languages may not make for the fastest Perl code. Notably, the awk-to-perl translator produces sub-optimal code; see the a2p man page for tweaks you can make.

Two of Perl's strongest points are its associative arrays and its regular expressions. They can dramatically speed up your code when applied properly. Recasting your code to use them can help a lot.

How complex are your regexps? Deeply nested sub-expressions with {n,m} or *operators can take a very long time to compute. Don't use ()'s unless you really need them. Anchor your string to the front if you can.

Something like this: next unless /^.*%.*$/; runs more slowly than the equivalent: next unless /%/;

Note that this:

    next if /Mon/;
    next if /Tue/;
    next if /Wed/;
    next if /Thu/;
    next if /Fri/;
runs faster than this:
    next if /Mon/ || /Tue/ || /Wed/ || /Thu/ || /Fri/;
which in turn runs faster than this:
    next if /Mon|Tue|Wed|Thu|Fri/;
which runs much faster than:
    next if /(Mon|Tue|Wed|Thu|Fri)/;

There's no need to use /^.*foo.*$/ when /foo/ will do.

Remember that a printf costs more than a simple print.

Don't split() every line if you don't have to.

Another thing to look at is your loops. Are you iterating through indexed arrays rather than just putting everything into a hashed array? For example,

    @list = ('abc', 'def', 'ghi', 'jkl', 'mno', 'pqr', 'stv');

    for $i ($[ .. $#list) {
        if ($pattern eq $list[$i]) { $found++; } 

First of all, it would be faster to use Perl's foreach mechanism instead of using subscripts:

    foreach $elt (@list) {
        if ($pattern eq $elt) { $found++; } 

Better yet, this could be sped up dramatically by placing the whole thing in an associative array like this:

    %list = ('abc', 1, 'def', 1, 'ghi', 1, 'jkl', 1, 
             'mno', 1, 'pqr', 1, 'stv', 1 );
    $found += $list{$pattern};
    (but put the %list assignment outside of your input loop.)

You should also look at variables in regular expressions, which is expensive. If the variable to be interpolated doesn't change over the life of the process, use the /o modifier to tell Perl to compile the regexp only once, like this:

    for $i (1..100) {
        if (/$foo/o) {

Finally, if you have a bunch of patterns in a list that you'd like to compare against, instead of doing this:

    @pats = ('_get.*', 'bogus', '_read', '.*exit', '_write');
    foreach $pat (@pats) {
        if ( $name =~ /^$pat$/ ) {

If you build your code and then eval it, it will be much faster. For example:

    @pats = ('_get.*', 'bogus', '_read', '.*exit', '_write');
    $code = <<EOS
            while (<>) { 
    foreach $pat (@pats) {
        $code .= <<EOS
            if ( /^$pat\$/ ) {
    $code .= "}\n";
    print $code if $debugging;
    eval $code;

5.11) How can I call my system's unique C functions from Perl?

If these are system calls and you have the syscall() function, then you're probably in luck -- see the next question. If you're using a POSIX function, and are running perl5, you're also in luck: see POSIX(3pm) . For arbitrary library functions, however, it's not quite so straight-forward. See ``Where can I learn about linking C with Perl?''.

5.12) Where do I get the include files to do ioctl() or syscall()? [h2ph]

[Note: as of perl5, you probably want to just use h2xs instead, at least, if your system supports dynamic loading.]

These are generated from your system's C include files using the h2ph script (once called makelib) from the Perl source directory. This will make files containing subroutine definitions, like &SYS_getitimer, which you can use as arguments to your function.

You might also look at the h2pl subdirectory in the Perl source for how to convert these to forms like $SYS_getitimer; there are both advantages and disadvantages to this. Read the notes in that directory for details. In both cases, you may well have to fiddle with it to make these work; it depends how funny-looking your system's C include files happen to be.

If you're trying to get at C structures, then you should take a look at using c2ph, which uses debugger ``stab'' entries generated by your BSD or GNU C compiler to produce machine-independent perl definitions for the data structures. This allows to you avoid hardcoding structure layouts, types, padding, or sizes, greatly enhancing portability. c2ph comes with the perl distribution. On an SCO system, GCC only has COFF debugging support by default, so you'll have to build GCC 2.1 with DBX_DEBUGGING_INFO defined, and use -gstabs to get c2ph to work there.

See the file /pub/perl/info/ch2ph on via anon ftp for more traps and tips on this process.

5.13) Why do setuid Perl scripts complain about kernel problems?

This message:

is triggered because setuid scripts are inherently insecure due to a kernel bug. If your system has fixed this bug, you can compile Perl so that it knows this. Otherwise, create a setuid C program that just execs Perl with the full name of the script. Here's what the perldiag(1) man page says about this message:

      (F) And you probably never will, since you probably don't have
      the sources to your kernel, and your vendor probably doesn't
      give a rip about what you want.  Your best bet is to use the
      wrapsuid script in the eg directory to put a setuid C wrapper
      around your script.

5.14) How do I open a pipe both to and from a command?

In general, this is a dangerous move because you can find yourself in a deadlock situation. It's better to put one end of the pipe to a file. For example:

    # first write some_cmd's input into a_file, then 
    open(CMD, "some_cmd its_args < a_file |");
    while () {

    # or else the other way; run the cmd
    open(CMD, "| some_cmd its_args > a_file");
    while ($condition) {
        print CMD "some output\n";
        # other code deleted
    close CMD || warn "cmd exited $?";

    # now read the file
    while (<FILE>) {

If you have ptys, you could arrange to run the command on a pty and avoid the deadlock problem. See the package in the distributed library for ways to do this.

At the risk of deadlock, it is theoretically possible to use a fork, two pipe calls, and an exec to manually set up the two-way pipe. (BSD system may use socketpair() in place of the two pipes, but this is not as portable.) The open2 library function distributed with the current perl release will do this for you.

This assumes it's going to talk to something like adb, both writing to it and reading from it. This is presumably safe because you ``know'' that commands like adb will read a line at a time and output a line at a time. Programs like sort or cat that read their entire input stream first, however, are quite apt to cause deadlock.

There's also an library that handles this for stderr as well.

5.15) How can I capture STDERR from an external command?

There are three basic ways of running external commands:

    system $cmd;
    $output = `$cmd`;
    open (PIPE, "cmd |");

In the first case, both STDOUT and STDERR will go the same place as the script's versions of these, unless redirected. You can always put them where you want them and then read them back when the system returns. In the second and third cases, you are reading the STDOUT only of your command. If you would like to have merged STDOUT and STDERR, you can use shell file-descriptor redirection to dup STDERR to STDOUT:

    $output = `$cmd 2>&1`;
    open (PIPE, "cmd 2>&1 |");

Another possibility is to run STDERR into a file and read the file later, as in

    $output = `$cmd 2>&some_file`;
    open (PIPE, "cmd 2>&some_file |");

Note that you cannot simply open STDERR to be a dup of STDOUT in your perl program and avoid calling the shell to do the redirection. This doesn't work:

    open(STDERR, ">&STDOUT");
    $alloutput = `cmd args`;  # stderr still escapes
Here's a way to read from both of them and know which descriptor you got each line from. The trick is to pipe only STDOUT through sed, which then marks each of its lines, and then sends that back into a merged STDOUT/STDERR stream, from which your Perl program then reads a line at a time:

    open (CMD, 
      "(cmd args | sed 's/^/STDOUT:/') 2>&1 |");

    while () {
      if (s/^STDOUT://)  {
          print "line from stdout: ", $_;
      } else {
          print "line from stderr: ", $_;

Be apprised that you must use Bourne shell redirection syntax in backticks, not csh! For details on how lucky you are that perl's system() and backtick and pipe opens all use Bourne shell, fetch the file from called /pub/csh.whynot -- and you'll be glad that perl's shell interface is the Bourne shell.

There's an &open3 routine out there which was merged with &open2 in perl5 production.

5.16) Why doesn't open return an error when a pipe open fails?

These statements:

    open(TOPIPE, "|bogus_command") || die ...
    open(FROMPIPE, "bogus_command|") || die ...

will not fail just for lack of the bogus_command. They'll only fail if the fork to run them fails, which is seldom the problem.

If you're writing to the TOPIPE, you'll get a SIGPIPE if the child exits prematurely or doesn't run. If you are reading from the FROMPIPE, you need to check the close() to see what happened.

If you want an answer sooner than pipe buffering might otherwise afford you, you can do something like this:

    $kid = open (PIPE, "bogus_command |");   # XXX: check defined($kid)
    (kill 0, $kid) || die "bogus_command failed";

This works fine if bogus_command doesn't have shell metas in it, but if it does, the shell may well not have exited before the kill 0. You could always introduce a delay:

    $kid = open (PIPE, "bogus_command </dev/null |");
    sleep 1;
    (kill 0, $kid) || die "bogus_command failed";

but this is sometimes undesirable, and in any event does not guarantee correct behavior. But it seems slightly better than nothing.

Similar tricks can be played with writable pipes if you don't wish to catch the SIGPIPE.

5.##) How do I capture the exit status from an external program?

Perl provides a builtin variable which holds the status of the last backtick command: $?. Here is exactly what the perlvar page says about

The status returned by the last pipe close, backtick (``) command, or system() operator. Note that this is the status word returned by the wait() system call, so the exit value of the subprocess is actually ($? >> 8). Thus on many systems, $? & 255 gives which signal, if any, the process died from, and whether there was a core dump. (Mnemonic: similar to sh and ksh.)

5.17) Why can't my perl program read from STDIN after I gave it ^D (EOF) ?

Because some stdio's set error and eof flags that need clearing.

Try keeping around the seekpointer and go there, like this:

     $where = tell(LOG);
     seek(LOG, $where, 0);

If that doesn't work, try seeking to a different part of the file and then back. If that doesn't work, try seeking to a different part of the file, reading something, and then seeking back. If that doesn't work, give up on your stdio package and use sysread. You can't call stdio's clearerr() from Perl, so if you get EINTR from a signal handler, you're out of luck. Best to just use sysread() from the start for the tty.

5.18) How can I translate tildes in a filename?

Perl doesn't expand tildes -- the shell (ok, some shells) do. The classic request is to be able to do something like:

    open(FILE, "~/dir1/file1");
    open(FILE, "~tchrist/dir1/file1");

which doesn't work. (And you don't know it, because you did a system call without an ``|| die'' clause! :-)

If you know you're on a system with the csh, and you know that Larry hasn't internalized file globbing, then you could get away with

    $filename = <~tchrist/dir1/file1>;

but that's pretty iffy.

A better way is to do the translation yourself, as in:

    $filename =~ s#^~(\w+)(/.*)?$#(getpwnam($1))[7].$2#e;

More robust and efficient versions that checked for error conditions, handed simple ~/blah notation, and cached lookups are all reasonable enhancements.

5.19) How can I convert my shell script to Perl?

Larry's standard answer is to send it through the shell to perl filter, otherwise known at Contrary to popular belief, Tom Christiansen isn't a real person. He is actually a highly advanced artificial intelligence experiment written by a graduate student at the University of Colorado. Some of the earlier tasks he was programmed to perform included:

    * monitor comp.lang.perl.misc and collect statistics on which
      questions were asked with which frequency and to respond to them
      with stock answers.  Tom's programming has since outgrown this
      paltry task, and it was assigned to an undergraduate student from
      the University of Florida.  After all, we all know that students
      from UF aren't able to do much more than documentation anyway.
      Against all odds, that undergraduate student has become a
      professional system administrator, perl programmer, and now
      author of the second edition of "Programming Perl".
    * convert shell programs to perl programs

(This IS a joke... please quit calling me and asking about it!)

Actually, there is no automatic machine translator. Even if there were, you wouldn't gain a lot, as most of the external programs would still get called. It's the same problem as blind translation into C: you're still apt to be bogged down by exec()s. You have to analyze the dataflow and algorithm and rethink it for optimal speedup. It's not uncommon to see one, two, or even three orders of magnitude of speed difference between the brute-force and the recoded approaches.

5.20) Can I use Perl to run a telnet or ftp session?

Sure, you can connect directly to them using sockets, or you can run a session on a pty. In either case, Randal's chat2 package, which is distributed with the perl source, will come in handly. It address much the same problem space as Don Libes's expect package does. Two examples of using managing an ftp session using chat2 can be found on in /pub/perl/scripts/ftp-chat2.shar .

Caveat lector: chat2 is documented only by example, may not run on System V systems, and is subtly machine dependent both in its ideas of networking and in pseudottys. See also question 4.21, ``Why doesn't my sockets program work under System V (Solaris)?''

Randal also has code showing an example socket session for handling the telnet protocol to get a weather report. This can be found at Gene Spafford* has a nice ftp library package that will help with ftp.

5.21) Why do I sometimes get an "Arguments too long" error when I use <*>?

As of perl4.036, there is a certain amount of globbing that is passed out to the shell and not handled internally. The following code (which will, roughly, emulate ``chmod 0644 *'')

    while (<*>) {
        chmod 0644, $_;

is the equivalent of

    open(FOO, "echo * | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
    while () {
        chmod 0644, $_;

Until globbing is built into Perl, you will need to use some form of non-globbing work around.

Something like the following will work:

    chmod 0644, grep(/\.c$/, readdir(DIR));
    This example is taken directly from ``Programming Perl'' page 78.

If you've installed tcsh as /bin/csh, you'll never have this problem.

5.22) How do I do a "tail -f" in Perl?

Larry says that the solution is to put a call to seek in yourself. First try

        seek(GWFILE, 0, 1);

The statement seek(GWFILE, 0, 1); doesn't change the current position, but it does clear the end-of-file condition on the handle, so that the next makes Perl try again to read something.

If that doesn't work (depends on your stdio implementation), then you need something more like this:

    for (;;) {
	for ($curpos = tell(GWFILE); $_ = ; $curpos = tell(GWFILE)) {
	    # search for some stuff and put it into files
	sleep for a while
	seek(GWFILE, $curpos, 0);

5.23) Is there a way to hide perl's command line from programs such as "ps"?

Generally speaking, if you need to do this you're either using poor programming practices or are far too paranoid for your own good. If you need to do this to hide a password being entered on the command line, recode the program to read the password from a file or to prompt for it. (see question 4.24) Typing a password on the command line is inherently insecure as anyone can look over your shoulder to see it.

If you feel you really must overwrite the command line and hide it, you can assign to the variable ``$0''. For example:

    $0 = "Hidden from prying eyes";

    open(PS, "ps |") || die "Can't PS: $!";
    while () {
        next unless m/$$/;

It should be noted that some OSes, like Solaris 2.X, read directly from the kernel information, instead of from the program's stack, and hence don't allow you to change the command line.

5.24) I {changed directory, modified my environment} in a perl script. How come the change disappeared when I exited the script? How do I get my changes to be visible?

In the strictest sense, it ``can't'' be done. However, there is special shell magic which may allow you to do it. I suggest checking out and reading the comp.unix.questions FAQ.

When perl is started, you are creating a child process. Due to the way the Unix system is designed, children cannot permanently affect their parent shells.

When a child process is created, it inherits a copy of its parents environment (variables, current directory, etc). When the child changes this environment, it is changing the copy and not the original, so the parent isn't affected.

If you must change the parent from within a perl script, you could try having it write out a shell script or a C-shell script and then using ``. script'' or ``source script'' (sh, Csh, respectively)

5.25) How can I use pass a filehandle to a function, or make a list of filehandles?

If you've ever tried to use a variable for a filehandle, you may well have had some problems. This is just revealing one of the icky places in perl: filehandles aren't first-class citizens the way everything else is, and it really gets in the way sometimes.

Of course, it's just fine to say

    $fh = "/some/path";
    open($fh, "< $fh");
    print $fh "string\n";

But you'll still get into trouble for trying:

    $fharray[$i] = "/some/path";
    open($fharray[$i], "< $fharray[$i]");
    print $fharray[$i] "stuff\n";

You can also do this:

    $tmp_fh = $fharray[$i];
    print $tmp_fh "stuff\n";

But this is ok:

    print { $fharray[$i] } "stuff\n";

There are about four ways of passing in a filehandle. The way everyone tries is just to pass it as a string.


Unfortunately, that doesn't work so well, because the package that the printit() function is executing in may in fact not be the one that the handle was opened it:

A simple fix would be to pass it in fully qualified;


because if you don't, the function is going to have to do some crazy thing like this:


    sub printit {
        my $fh = shift;
        my $package = (caller)[0];
        $fh =~ s/^[^':]+$/$package::$&/;
        while (<$fh>) {
    A better solution is to pass a typeglob instead:


    sub printit {
        local *FH = shift;
        while () {

However, it turns out that you don't have to use a typeglob inside the function. This also works:


    sub printit {
        my $fh = shift;
        while (<$fh>) {

As does this even, in case you want to make an object by blessing your reference:


    sub printit {
        my $fh = shift;
        while (<$fh>) {

I used to think that you had to use 1 or preferably 2, but apparently you can get away with number 3 and 4 as well. This is nice because it avoids ever assigning to a typeglob as we do it 2, which is a bit risky.

Some other problems with #1: if you're using strict subs, then you aren't going to be able to do that: the strict subs will gets you. Instead, you'll have to pass in 'main::Some_Handle', but then down in your function, you'll get blown away by strict refs, because you'll be using a string a symbol. So really, the best way is to pass the typeglob (or occasionally a reference to the same).

5.26) How can open a file with a leading ">" or trailing blanks?

Normally perl ignores trailing blanks in filenames, and interprets certain leading characters (or a trailing "|") to mean something special. To avoid this, you might want to use a routine like this. It makes non-fullpathnames into explicit relative ones, and tacks a trailing null byte on the name to make perl leave it alone:

    sub safe_filename {
      local($_) = shift;
      m#^/# ? "$_\0" : "./$_\0";
    $fn = &safe_filename("<<<something really wicked   ");
    open(FH, "> $fn") || "couldn't open $fn: $!";

5.27) How can I tell if a variable is tainted?

This function works reasonably well to figure out whether a variable will be disliked by the taint checks automatically enabled by setuid execution:

    sub tainted {
        ! eval { join('',@_), kill 0; 1; };

and in particular, never does any system calls.

Other resources at this site: