Perl

Larry Wall began work on Perl in 1987, while employed as a programmer at Unisys;[16] he released version 1.0 on December 18, 1987.[1]

[11] The early language of Perl that Wall based it off of was the use of existing languages that would help with text manipulation.[11]

Perl 2, released in June 1988,[35][36] featured a better regular expression engine. Perl 3, released in October 1989,[35] added support for binary data streams.[37]

Originally, the only documentation for Perl was a single lengthy man page. In 1991, Programming Perl, known to many Perl programmers as the "Camel Book" because of its cover, was published and became the de facto reference for the language.[38] At the same time, the Perl version number was bumped to 4, not to mark a major change in the language but to identify the version that was well documented by the book.[39] Perl 4 was released in March 1991.[35]

Perl 4 went through a series of maintenance releases, culminating in Perl 4.036 in 1993, whereupon Wall abandoned Perl 4 to begin work on Perl 5. Initial design of Perl 5 continued into 1994. The perl5-porters mailing list was established in May 1994 to coordinate work on porting Perl 5 to different platforms. It remains the primary forum for development, maintenance, and porting of Perl 5.[40]

Perl 5.000 was released on October 17, 1994.[41] It was a nearly complete rewrite of the interpreter, and it added many new features to the language, including objects, references, lexical (my) variables, and modules. Importantly, modules provided a mechanism for extending the language without modifying the interpreter. This allowed the core interpreter to stabilize, even as it enabled ordinary Perl programmers to add new language features. Perl 5 has been in active development since then.

Perl 5.001 was released on March 13, 1995. Perl 5.002 was released on February 29, 1996 with the new prototypes feature. This allowed module authors to make subroutines that behaved like Perl builtins. Perl 5.003 was released June 25, 1996, as a security release.[42]

One of the most important events in Perl 5 history took place outside of the language proper and was a consequence of its module support. On October 26, 1995, the Comprehensive Perl Archive Network (CPAN) was established as a repository for the Perl language and Perl modules; as of December 2022, it carries over 211,850 modules in 43,865 distributions, written by more than 14,324 authors, and is mirrored worldwide at more than 245 locations.[43]

Perl 5.004 was released on May 15, 1997, and included, among other things, the UNIVERSAL package, giving Perl a base object from which all classes were automatically derived and the ability to require versions of modules. Another significant development was the inclusion of the CGI.pm module,[44] which contributed to Perl's popularity as a CGI scripting language.[45]

Perl 5.004 added support for Microsoft Windows, Plan 9, QNX, and AmigaOS.[44]

Perl 5.005 was released on July 22, 1998. This release included several enhancements to the regex engine, new hooks into the backend through the B::* modules, the qr// regex quote operator, a large selection of other new core modules, and added support for several more operating systems, including BeOS.[46]

Perl 5.6 was released on March 22, 2000. Major changes included 64-bit support, Unicode string representation, support for files over 2 GiB, and the "our" keyword.[48][49] When developing Perl 5.6, the decision was made to switch the versioning scheme to one more similar to other open source projects; after 5.005_63, the next version became 5.5.640, with plans for development versions to have odd numbers and stable versions to have even numbers.[50]

In 2000, Wall put forth a call for suggestions for a new version of Perl from the community. The process resulted in 361 RFC (request for comments) documents that were to be used in guiding development of Perl 6. In 2001,[51] work began on the "Apocalypses" for Perl 6, a series of documents meant to summarize the change requests and present the design of the next generation of Perl. They were presented as a digest of the RFCs, rather than a formal document. At this point, Perl 6 existed only as a description of a language.

Perl 5.8 was first released on July 18, 2002, and further 5.X versions have been released approximately yearly since then. Perl 5.8 improved Unicode support, added a new I/O implementation, added a new thread implementation, improved numeric accuracy, and added several new modules.[52] As of 2013 this version still remained the most popular version of Perl and was used by Red Hat 5, Suse 10, Solaris 10, HP-UX 11.31 and AIX 5.

In 2004, work began on the "Synopses" – documents that originally summarized the Apocalypses, but which became the specification for the Perl 6 language. In February 2005, Audrey Tang began work on Pugs, a Perl 6 interpreter written in Haskell.[53] This was the first concerted effort toward making Perl 6 a reality. This effort stalled in 2006.[54]

PONIE is an acronym for Perl On New Internal Engine. The PONIE Project existed from 2003 until 2006 and was to be a bridge between Perl 5 and Perl 6. It was an effort to rewrite the Perl 5 interpreter to run on Parrot, the Perl 6 virtual machine. The goal was to ensure the future of the millions of lines of Perl 5 code at thousands of companies around the world.[55] The PONIE project ended in 2006 and is no longer being actively developed. Some of the improvements made to the Perl 5 interpreter as part of PONIE were folded into that project.[56]

On December 18, 2007, the 20th anniversary of Perl 1.0, Perl 5.10.0 was released. Perl 5.10.0 included notable new features, which brought it closer to Perl 6. These included a switch statement (called "given"/"when"), regular expressions updates, and the smart match operator (~~).[57][58] Around this same time, development began in earnest on another implementation of Perl 6 known as Rakudo Perl, developed in tandem with the Parrot virtual machine. As of November 2009, Rakudo Perl has had regular monthly releases and now is the most complete implementation of Perl 6.

A major change in the development process of Perl 5 occurred with Perl 5.11; the development community has switched to a monthly release cycle of development releases, with a yearly schedule of stable releases. By that plan, bugfix point releases will follow the stable releases every three months.

On April 12, 2010, Perl 5.12.0 was released. Notable core enhancements include new package NAME VERSION syntax, the yada yada operator (intended to mark placeholder code that is not yet implemented), implicit strictures, full Y2038 compliance, regex conversion overloading, DTrace support, and Unicode 5.2.[59]

On May 14, 2011, Perl 5.14 was released with JSON support built-in.[60]

On May 20, 2012, Perl 5.16 was released. Notable new features include the ability to specify a given version of Perl that one wishes to emulate, allowing users to upgrade their version of Perl, but still run old scripts that would normally be incompatible.[61] Perl 5.16 also updates the core to support Unicode 6.1.[61]

On May 18, 2013, Perl 5.18 was released. Notable new features include the new dtrace hooks, lexical subs, more CORE:: subs, overhaul of the hash for security reasons, support for Unicode 6.2.[62]

On May 27, 2014, Perl 5.20 was released. Notable new features include subroutine signatures, hash slices/new slice syntax, postfix dereferencing (experimental), Unicode 6.3, and a rand() function using a consistent random number generator.[63]

Some observers credit the release of Perl 5.10 with the start of the Modern Perl movement.[64] In particular, this phrase describes a style of development that embraces the use of the CPAN, takes advantage of recent developments in the language, and is rigorous about creating high quality code.[65] While the book "Modern Perl"[66] may be the most visible standard-bearer of this idea, other groups such as the Enlightened Perl Organization[67] have taken up the cause.

In late 2012 and 2013, several projects for alternative implementations for Perl 5 started: Perl5 in Perl6 by the Rakudo Perl team,[68] moe by Stevan Little and friends,[69] p2[70] by the Perl11 team under Reini Urban, gperl by goccy,[71] and rperl, a Kickstarter project led by Will Braswell and affiliated with the Perll11 project.[72]

Camelia, the logo for the Perl 6 project[73]

At the 2000 Perl Conference, Jon Orwant made a case for a major new language-initiative.[74] This led to a decision to begin work on a redesign of the language, to be called Perl 6. Proposals for new language features were solicited from the Perl community at large, which submitted more than 300 RFCs.[75]

Wall spent the next few years digesting the RFCs and synthesizing them into a coherent framework for Perl 6. He presented his design for Perl 6 in a series of documents called "apocalypses" – numbered to correspond to chapters in Programming Perl. As of January 2011, the developing specification of Perl 6 was encapsulated in design documents called Synopses – numbered to correspond to Apocalypses.[76]

Thesis work by Bradley M. Kuhn, overseen by Wall, considered the possible use of the Java virtual machine as a runtime for Perl.[77] Kuhn's thesis showed this approach to be problematic. In 2001, it was decided that Perl 6 would run on a cross-language virtual machine called Parrot. This will mean that other languages targeting the Parrot will gain native access to CPAN, allowing some level of cross-language development.

In 2005, Audrey Tang created the Pugs project, an implementation of Perl 6 in Haskell. This acted as, and continues to act as, a test platform for the Perl 6 language (separate from the development of the actual implementation) – allowing the language designers to explore. The Pugs project spawned an active Perl/Haskell cross-language community centered around the Libera Chat #raku IRC channel. Many functional programming influences were absorbed by the Perl 6 design team.[78]

In 2012, Perl 6 development was centered primarily on two compilers:[79]

1. Rakudo, an implementation running on the Parrot virtual machine and the Java virtual machine.[80]
2. Niecza, which targets the Common Language Runtime.

In 2013, MoarVM ("Metamodel On A Runtime"), a C language-based virtual machine designed primarily for Rakudo was announced.[81]

In October 2019, Perl 6 was renamed to Raku.[82]

As of 2017 only the Rakudo implementation and MoarVM are under active development, and other virtual machines, such as the Java Virtual Machine and JavaScript, are supported.[83]

In June 2020, Perl 7 was announced as the successor to Perl 5.[84] Perl 7 was to initially be based on Perl 5.32 with a release expected in first half of 2021, and release candidates sooner.[85]

This plan was revised in May 2021, without any release timeframe or version of Perl 5 for use as a baseline specified.[86] When Perl 7 is released, Perl 5 will go into long term maintenance. Supported Perl 5 versions however will continue to get important security and bug fixes.[87]

Perl 7 was announced on 24 June 2020 at "The Perl Conference in the Cloud" as the successor to Perl 5.[85][84] Based on Perl 5.32, Perl 7 was planned to be backward compatible with modern Perl 5 code; Perl 5 code, without boilerplate (pragma) header needs adding use compat::perl5; to stay compatible, but modern code can drop some of the boilerplate.

The plan to go to Perl 7 brought up more discussion, however, and the Perl Steering Committee canceled it to avoid issues with backward compatibility for scripts that were not written to the pragmas and modules that would become the default in Perl 7. Perl 7 will only come out when the developers add enough features to warrant a major release upgrade.[88]

According to Wall, Perl has two slogans. The first is "There's more than one way to do it," commonly known as TMTOWTDI, (pronounced Tim Toady). As proponents of this motto argue, this philosophy makes it easy to write concise statements.[89][90][91]

The second slogan is "Easy things should be easy and hard things should be possible".[16]

The design of Perl can be understood as a response to three broad trends in the computer industry: falling hardware costs, rising labor costs, and improvements in compiler technology. Many earlier computer languages, such as Fortran and C, aimed to make efficient use of expensive computer hardware. In contrast, Perl was designed so that computer programmers could write programs more quickly and easily.[92]

Perl has many features that ease the task of the programmer at the expense of greater CPU and memory requirements. These include automatic memory management; dynamic typing; strings, lists, and hashes; regular expressions; introspection; and an eval() function. Perl follows the theory of "no built-in limits",[29] an idea similar to the Zero One Infinity rule.

Wall was trained as a linguist, and the design of Perl is very much informed by linguistic principles. Examples include Huffman coding (common constructions should be short), good end-weighting (the important information should come first), and a large collection of language primitives. Perl favors language constructs that are concise and natural for humans to write, even where they complicate the Perl interpreter.[93]

Perl's syntax reflects the idea that "things that are different should look different."[94] For example, scalars, arrays, and hashes have different leading sigils. Array indices and hash keys use different kinds of braces. Strings and regular expressions have different standard delimiters. This approach can be contrasted with a language such as Lisp, where the same basic syntax, composed of simple and universal symbolic expressions, is used for all purposes.

There is a broad practical bent to both the Perl language and the community and culture that surround it. The preface to Programming Perl begins: "Perl is a language for getting your job done."[16] One consequence of this is that Perl is not a tidy language. It includes many features, tolerates exceptions to its rules, and employs heuristics to resolve syntactical ambiguities. Because of the forgiving nature of the compiler, bugs can sometimes be hard to find. Perl's function documentation remarks on the variant behavior of built-in functions in list and scalar contexts by saying, "In general, they do what you want, unless you want consistency."[95]

The overall structure of Perl derives broadly from C. Perl is procedural in nature, with variables, expressions, assignment statements, brace-delimited blocks, control structures, and subroutines.[96]

Perl also takes features from shell programming. All variables are marked with leading sigils, which allow variables to be interpolated directly into strings. However, unlike the shell, Perl uses sigils on all accesses to variables, and unlike most other programming languages that use sigils, the sigil doesn't denote the type of the variable but the type of the expression. So for example, while an array is denoted by the sigil "@" (for example @arrayname), an individual member of the array is denoted by the scalar sigil "$" (for example $arrayname[3]). Perl also has many built-in functions that provide tools often used in shell programming (although many of these tools are implemented by programs external to the shell) such as sorting, and calling operating system facilities.

Perl takes hashes ("associative arrays") from AWK and regular expressions from sed. These simplify many parsing, text-handling, and data-management tasks. Shared with Lisp is the implicit return of the last value in a block, and all statements are also expressions which can be used in larger expressions themselves.

Perl 5 added features that support complex data structures, first-class functions (that is, closures as values), and an object-oriented programming model. These include references, packages, class-based method dispatch, and lexically scoped variables, along with compiler directives (for example, the strict pragma). A major additional feature introduced with Perl 5 was the ability to package code as reusable modules. Wall later stated that "The whole intent of Perl 5's module system was to encourage the growth of Perl culture rather than the Perl core."[97]

All versions of Perl do automatic data-typing and automatic memory management. The interpreter knows the type and storage requirements of every data object in the program; it allocates and frees storage for them as necessary using reference counting (so it cannot deallocate circular data structures without manual intervention). Legal type conversions — for example, conversions from number to string — are done automatically at run time; illegal type conversions are fatal errors.

Perl has been referred to as "line noise" and a write-only language by its critics. The earliest such mention was in the first edition of the book Learning Perl, a Perl 4 tutorial book written by Randal L. Schwartz,[98] in the first chapter of which he states: "Yes, sometimes Perl looks like line noise to the uninitiated, but to the seasoned Perl programmer, it looks like checksummed line noise with a mission in life."[99] He also stated that the accusation that Perl is a write-only language could be avoided by coding with "proper care".[99] The Perl overview document perlintro states that the names of built-in "magic" scalar variables "look like punctuation or line noise".[100] However, the English module provides both long and short English alternatives. perlstyle document states that line noise in regular expressions could be mitigated using the /x modifier to add whitespace.[101]

According to the Perl 6 FAQ, Perl 6 was designed to mitigate "the usual suspects" that elicit the "line noise" claim from Perl 5 critics, including the removal of "the majority of the punctuation variables" and the sanitization of the regex syntax.[102] The Perl 6 FAQ also states that what is sometimes referred to as Perl's line noise is "the actual syntax of the language" just as gerunds and prepositions are a part of the English language.[102] In a December 2012 blog posting, despite claiming that "Rakudo Perl 6 has failed and will continue to fail unless it gets some adult supervision", chromatic stated that the design of Perl 6 has a "well-defined grammar" as well as an "improved type system, a unified object system with an intelligent metamodel, metaoperators, and a clearer system of context that provides for such niceties as pervasive laziness".[103] He also stated that "Perl 6 has a coherence and a consistency that Perl 5 lacks."[103]

In older versions of Perl, one would write the Hello World program as:

print "Hello, World!\n";

Here is a more complex Perl program, that counts down seconds from a given starting value:

#!/usr/bin/perl
use strict;
use warnings;

my ( $remaining, $total );

$remaining = $total = shift(@ARGV);

STDOUT->autoflush(1);

while ( $remaining ) {
    printf ( "Remaining %s/%s \r", $remaining--, $total );
    sleep 1;
}

print "\n";

The Perl interpreter can also be used for one-off scripts on the command line. The following example (as invoked from an sh-compatible shell, such as Bash) translates the string "Bob" in all files ending with .txt in the current directory to "Robert":

$ perl -i.bak -lp -e 's/Bob/Robert/g' *.txt

No written specification or standard for the Perl language exists for Perl versions through Perl 5, and there are no plans to create one for the current version of Perl. There has been only one implementation of the interpreter, and the language has evolved along with it. That interpreter, together with its functional tests, stands as a de facto specification of the language. Perl 6, however, started with a specification,[104] and several projects[105] aim to implement some or all of the specification.

Perl is implemented as a core interpreter, written in C, together with a large collection of modules, written in Perl and C. As of 2010, the interpreter is 150,000 lines of C code and compiles to a 1 MB executable on typical machine architectures. Alternatively, the interpreter can be compiled to a link library and embedded in other programs. There are nearly 500 modules in the distribution, comprising 200,000 lines of Perl and an additional 350,000 lines of C code (much of the C code in the modules consists of character encoding tables).

The interpreter has an object-oriented architecture. All of the elements of the Perl language—scalars, arrays, hashes, coderefs, file handles—are represented in the interpreter by C structs. Operations on these structs are defined by a large collection of macros, typedefs, and functions; these constitute the Perl C API. The Perl API can be bewildering to the uninitiated, but its entry points follow a consistent naming scheme, which provides guidance to those who use it.

The life of a Perl interpreter divides broadly into a compile phase and a run phase.[106] In Perl, the phases are the major stages in the interpreter's life-cycle. Each interpreter goes through each phase only once, and the phases follow in a fixed sequence.

Most of what happens in Perl's compile phase is compilation, and most of what happens in Perl's run phase is execution, but there are significant exceptions. Perl makes important use of its capability to execute Perl code during the compile phase. Perl will also delay compilation into the run phase. The terms that indicate the kind of processing that is actually occurring at any moment are compile time and run time. Perl is in compile time at most points during the compile phase, but compile time may also be entered during the run phase. The compile time for code in a string argument passed to the eval built-in occurs during the run phase. Perl is often in run time during the compile phase and spends most of the run phase in run time. Code in BEGIN blocks executes at run time but in the compile phase.

At compile time, the interpreter parses Perl code into a syntax tree. At run time, it executes the program by walking the tree. Text is parsed only once, and the syntax tree is subject to optimization before it is executed, so that execution is relatively efficient. Compile-time optimizations on the syntax tree include constant folding and context propagation, but peephole optimization is also performed.[107]

Perl has a Turing-complete grammar because parsing can be affected by run-time code executed during the compile phase.[108] Therefore, Perl cannot be parsed by a straight Lex/Yacc lexer/parser combination. Instead, the interpreter implements its own lexer, which coordinates with a modified GNU bison parser to resolve ambiguities in the language.

It is often said that "Only perl can parse Perl",[109] meaning that only the Perl interpreter (perl) can parse the Perl language (Perl), but even this is not, in general, true. Because the Perl interpreter can simulate a Turing machine during its compile phase, it would need to decide the halting problem in order to complete parsing in every case. It is a longstanding result that the halting problem is undecidable, and therefore not even Perl can always parse Perl. Perl makes the unusual choice of giving the user access to its full programming power in its own compile phase. The cost in terms of theoretical purity is high, but practical inconvenience seems to be rare.[110]

Other programs that undertake to parse Perl, such as source-code analyzers and auto-indenters, have to contend not only with ambiguous syntactic constructs but also with the undecidability of Perl parsing in the general case. Adam Kennedy's PPI project focused on parsing Perl code as a document (retaining its integrity as a document), instead of parsing Perl as executable code (that not even Perl itself can always do). It was Kennedy who first conjectured that "parsing Perl suffers from the 'halting problem',"[111] which was later proved.[112]

Perl is distributed with over 250,000 functional tests for core Perl language and over 250,000 functional tests for core modules. These run as part of the normal build process and extensively exercise the interpreter and its core modules. Perl developers rely on the functional tests to ensure that changes to the interpreter do not introduce software bugs; additionally, Perl users who see that the interpreter passes its functional tests on their system can have a high degree of confidence that it is working properly.