Perl was invented by Larry Wall. He called it Practical Extraction and Reporting
Language (he also calls it Pathologically Eclectic Rubbish Lister). What
started as an exercise in unifying multiple tools used to write scripts to make
routine tasks of a system adminstrator evolved into a powerful scripting
language with lots of followers.
In all fairness, Perl (written always as Perl and not as PERL), is now treated
as a generic programming language, though its early beginnings as a melting pot
of multiple computing paradigms still make it possible to write undecipherable
programs! We will try to get an introduction to Perl and its prowess as a text
manipulation language without trying to write cryptic programs.
According to Larry Wall, the parents of Perl are
So, Perl is a computer language that helps to implement some
common sense with help from the principles of computer science in an artistic
way using common linguistic constructs.
LINGUISTICS Perl COMMON SENSE
Like mentioned before, Perl was born as a tool to aid system administrators. In
1986, Larry Wall was asked to build a bi-coastal CM system in a very short
time. That, he did and then his manager asked to produce reports from the
system. Awk in those days did not have the capability to manage multiple input
files and hence the new language was born.
Primarily, the new language was aimed at getting things done quickly with data
from and to multiple files.
From a quick hack by one system administrator, Perl has grown into a
full-fledged language. It is being developed and enhanced continuously by
hundreds of programmers around the world. One big step in earning recognition
was the addition of regular expression engine. Now, the regular expression
capabilities of Perl are so well known (especially since version 5.0), that it
is being used in other languages like Python as Perl5 regex'es.
The growth of Internet also complemented Perl. The initial attempt at providing
dynamic content was through CGI (even now CGI is used extensively), and Perl's
remarkable text handling features made it a quick fit. CGI programming is now
synonymous with Perl programming.
CPAN - Comprehensive Perl Archive Network, was set up to share Perl code. Perl
supports modules and chances are that for 99% of the programming requirements,
there is already a tested module in CPAN (for the remaining 1%, write modules
and contribute to CPAN!). Using modules really mask the complexities of
adhering to pre-defined standards and frees you to concentrate on your tasks -
no point in re-inventing the wheel. Now, you have modules which handles
graphics, CGI etc...
You can also embed Perl code in your C/C++ programs. A very popular embedded
Perl architecture is mod_perl for Apache web server.
JAPH is a project to get Java and Perl working together.
Perl can handle strings, dates, binary data, database connectivity, streams,
sockets and many more. This ability to manipulate multiple data types help
immensely in data conversion (and by the way, it is much faster than PL/SQL!).
Perl also has provision for lists (or arrays) and for hashes (associative
arrays). Perl also supports references, which are similar to the pointers in C.
Lists, hashes and references together make it possible to define and manipulate
powerful custom-defined data-types.
Perl does not differentiate between files and pipes. So, it makes it very easy
to use Perl as a glue language. Suppose you have a sed script, the output of
which is to be given to a Perl script. You can do this the UNIX way,
or the perl way
sedscript | perlscript
This really helps when people want to migrate from traditional
UNIX tools like Awk, sed, grep etc... You can use these tools straightaway
instead of worrying on how to do the same thing entirely in Perl.
open(FH,"sedscript|") or die "could not open sedscript\n"l
In this aspect, Perl is just like shell. However, we must consider other
features of Perl, which shell simply cannot provide easily.
CGI.pm. Period. Almost all CGI programs written today are using the CGI.pm
module from CPAN. Even before this was written, people used to use Perl
extensively for CGI programming. CGI.pm made the process streamlined and easy,
even for beginners. The graphics library GD is used extensively in producing
dynamic web charts.
The ease with which Perl can be employed to write programs quickly cannot be
overstressed. A disturbing fact about this is that such quick code can tend to
be dirty and quickly get out of hand if you keep extending it! Most of
the time, you must control your urges to over-extend short programs! But, as a
prototyping tool, or as a fast reporting/text-processing tool, Perl is
Two very good tools worth mentioning in this context are s2p and a2p
tools which come with the Perl distribution. s2p converts a sed script to Perl
script and a2p converts from Awk scripts. These two help a lot in extending sed
and awk scripts.
Most of the Perl code will run without any change in Unix or Windows or
Macintosh. Typical changes you might have to make include specifying file paths
and use of low-level OS specific functions.
Just go to http://www.perl.com and download the source or pre-compiled
binaries. Installation typically includes extracting the binary and then
changing your PATH variable to reflect where Perl executable resides. Even when
you want to compile Perl from scratch, it is a simple job.
Perl has a remarkable resemblance to the syntax of C, AWK, SED and SHELL.
99% of code looks like C code. So, it is very easy for C programmers to switch
to Perl. And believe me, the code as you go philosophy of Perl really
makes C programmers happy - especially for small programs. All C functions that
are available through standard libraries are available with little or no change
at all in Perl.
AWK & SED
The string processing strategy of Perl is very similar to that of Awk and sed,
making it easy to migrate.
Again, the commenting scheme, variable naming scheme etc of Perl look similar to
that of Shell. Many shell utilities like grep, tr etc are available as
functions within Perl.