Talk:Perl

I just removed a link to http://xahlee.org/UnixResource_dir/perlr.html as the page it links to is a) very out of date, b) rather badly written and c) extremely critical of Perl. I've replaced it with a link to http://books.perl.org/

Davorg 14:30, 15 Mar 2005 (UTC)

The link to http://xahlee.org/UnixResource_dir/perlr.html has returned, this time under a ===criticism=== sub-heading. However, the page has many problems

The page is badly written. Partly, this is because english is not the author's native language. If this were a page on Wikipedia, we could bring the text up to standard, but it is an external page, so we can't.

The author clearly doesn't like Perl, but he doesn't say why, or support his position with argument, evidence, or references.

The author uses invective freely against people and things that he doesn't like. This is neither useful nor informative.

There are a half-dozen capsule reviews of Perl books, with pictures of the covers. This might be worth linking to; however, on personal knowledge I dispute some of the assertions in the reviews. Again, since this is an external page, we don't have any mechanism for vetting or correcting its content.

There are links to 4 "essays" on the same site. Like the main page, they are marked by strong opinion, insults, and an almost complete lack of argument or evidence.

There are links to a half-dozen outside pages. Some are on perl.com, which is already linked from the Wikipedia Perl page. Others are of only tangential relevance to Perl.

I'm all for giving the reader links to relevant material, but I don't think a link to this page should remain, even marked as "criticism".

Swmcd 20:51, 2005 Mar 26 (UTC)

I agree. There is some insight on the page, but it takes too much work on the part of the reader to extract it. Or to put it another way, the page sucks :-) --Yath 23:22, 26 Mar 2005 (UTC)

I vote for just removing it again. Davorg 09:44, 27 Mar 2005 (UTC)

I came to the discussion page here intending to make the very same point (that the link to Xah Lee's perl critique should be removed). Even as criticism, its information content is really low. I'm going ahead and removing it again. Oh, except someone already did. Yay. John Callender 17:46, 26 May 2005 (UTC)

A sample Perl program

A sample Perl program

       $message = "Hello, world! The magic number is 234542354.";
       print $message;
       $message =~ s/d+/-1/;
       print $message;
       exit 0

The above program can be written as:

       $_ = "The magic number is 150\n";
       print;
       s/\d+/-1/;
       print;

which demonstrates the idea of topic ($_) in Perl. Also, the exit 0 at the end is completely redundant. (See: pl:Perl) Rafał Pocztarski 07:01, 7 Aug 2004 (UTC)

A version showing slightly more powerful regex, err, power in Perl:

       $_ = "The magic number is 150\n";
       print;
       s/ (\d+) / $1 * 2 /ex;
       print;

Quite a nice example. Rafał Pocztarski 07:11, 7 Aug 2004 (UTC)

You're saying that it shows "high use of meaningful punctuation characters" ? You've never seen any REAL perl code :). What about using this (uppercase input) as an example:

 while(<>) {
   $_=~y/a-z/A-Z/;
   print;
 }

--Taw

Even better would be a line noise winner from one of the obfuscated perl contests, if that code is public domain. :-)

But that would miss the point - we want to show real Perl code, not one that was deliberately obfuscated. --Taw

I feel this entry should shy away from obfuscated code entirely! It just helps encourage the FUD that Perl is unreadable/write only/whatever, and that's a biased opinion. I can write incredibly obfuscated Java, but no-one feels the need to be posting that on the Java page. It tarnishes the neutral tone of the entry. I'd go as far as to remove the example (it's a bit pointless anyhow, what exactly is it demonstrating? I could come up with far more interesting examples of punctuation character usage, so that can't be it).

Anway, as an aside, you do realize you can get rid of the $_ in your code?:

 while(<>) {
   tr/a-z/A-Z/;
   print;
 }

Any books that tell you otherwise, burn 'em!

Finally there's the succinct:

 print uc while (<>);

... and I'll be surprised if there's anything more readable for the same task in any other popular language, assuming you've bothered to learn at least the rudiments of Perl, that is.

-- Derek

Or better yet:

   print uc while <>;

Even more clean and readable. One only has to say: uc is upper-case and the diamond is input. Great example. Rafał Pocztarski 07:04, 7 Aug 2004 (UTC)

Perl programming language?

Why is this at Perl programming language? Is there any reason for the disambiguation? Perl is just a redirect. --Eloquence 14:02 Jan 20, 2003 (UTC)

For consistency in naming. It was decided some time ago that human languages would be suffixed "language" and computer languages would be suffixed "programming language". It's got nothing to do with disambiguation. If you are going to move all these computer language pages, make sure that you check the What links here for double redirects which you may have introduced and fix them -- Derek Ross 16:30 Jan 24, 2003 (UTC)

Yeah, I found out by now. We have changed this policy, also for languages (e.g. Sanskrit). Don't worry about double redirs, I will take care of those. --Eloquence 16:33 Jan 24, 2003 (UTC)

I agree with the term "programming language". However, I've noticed lots of [older] references to this class of language which call them "computer languages". I don't really agree with the latter term, since they are really just human languages used to program computers. A minor point. As to appending "programming language" to the official titles, I think that's a good idea... the single example of Java vs. Java should suffice to convince most people. 2 cents. --Patrick Corcoran

Java is a case of disambiguation, see Wikipedia:Disambiguation. Perl is not. --Eloquence 16:46 Jan 24, 2003 (UTC)

Oh, great. The naming policy changed and I never even saw the discussion, never mind took part in it. For the record I don't agree with the change and I find it annoying that it took place without any of my input. -- Derek Ross 16:42 Jan 24, 2003 (UTC)

See the interlanguage links on Perl. Everyone else is doing the right thing already. --Eloquence 16:46 Jan 24, 2003 (UTC)

[Pp]erl

See http://www.perldoc.com/perl5.8.0/pod/perlfaq1.html, under "What's the difference between Perl and perl?". [Pp]erl is not and has never been an acronym. -- nknight 23:36 Mar 6, 2003 (UTC)

It would seem, though, that the Perl documentation is self-contradictory in this regard, since http://www.perldoc.com/perl5.8.0/pod/perl.html#NAME says "perl - Practical Extraction and Report Language," which is how Larry Wall is supposed to have gotten the name for the language in the first place. -- James Munroe 04:53 Aug 31, 2003 (UTC)

But if it was Practical Extraction and Report Language, the language name should be PEARL, not Perl. (Note: All caps, an `A'.) In fact, that was its original name until Larry found another language named PEARL (or Pearl). So it went to Perl, and the acronym no longer makes sense.

Like what anon user (above) says, pearl (now perl) started off as an acronym and Larry Wall had to change the name in order to avoid clash with another language with the same name.

The old expansion would still apply to perl as words like "and" are usually excluded from an acronym. However, Wall, the jovial person that he is, came up with the new acronym "Pathologically Eclectic Rubbish Lister" to match the new name. However I feel that these events and the link given by Nknight above should not be interpreted to mean that perl is no longer an acronym but a backronym. (refer interview with Larry Wall and the question ".. How'd you come up with that name?" http://www.linuxjournal.com/article.php?sid=3394 ) Jay 08:32, 3 Mar 2004 (UTC)

Hi, folks; perl never was an acronym. If it was, then Larry himself is lying. Anyone who's been in perl long enough to read the FAQ knows the story, and I've never heard anyone who's actually read the story claim that it's not true. Most everyone who claims perl was originally an acronym is much further removed than those who know the truth.

Incidentally, I corrected the famous quip, "Nothing but perl can parse Perl," from Tom Christiansen. The source on this is the FAQ (http://www.perldoc.com/perl5.8.0/pod/perlfaq1.html). Jdavidb 15:55, 8 Mar 2004 (UTC)

Full Ack regarding the Perl FAQ. According to it, Perl is only backronymed to Practical Extraction and Report Language.

But (and I noticed this just in this moment) if you look at http://history.perl.org/PerlTimeline.html under 1987, the (said to be) first man page shows exactly that acronym. Perl History also talks about 18th of December 1987, release date of Perl 1.000.

But in fact the first occurence of Perl at all was in May 1987, see http://use.perl.org/articles/03/12/15/0326207.shtml?tid=1. The therein mentioned Usenet posting can be found at http://groups.google.com/groups?selm=4628%40sdcrdcf.UUCP. The posting mentions that at that date, Perl is already 3 months old.

I'll put the Usenet posting and the Perl History Timeline into the article as external links, so the details of what is true and what is not can be discussed before any decisions regarding the article's content are made. :-) --XTaran 23:07, 11 Mar 2004 (UTC)

Blame me for the initial-cap spelling of Perl for the language. When we were writing the first Camel book, the lowercase version of "perl" kept disappearing into the text. So I discussed with Larry, and we decided together that "perl" would represent the interpreter (and be in constant-width font as a result), and "Perl" would represent the language. And no, "PERL" was never valid. -- Randal Schwartz

Name

""Perl" was originally named "pearl", after the "pearl of great price" of Matthew 13:46"

What is the reference to this information ? Jay 14:44, 6 Mar 2004 (UTC)

This (http://www.wired.com/wired/archive/8.10/cruise.html?pg=3) seems to be the source, with a little bit of this (http://groups.google.com/groups?q=%22pearl+of+great+price%22+perl+wall&hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=1995Nov3.231257.6290%40netlabs.com&rnum=1) thrown in for seasoning. chocolateboy 17:38, 6 Mar 2004 (UTC)

Regular Expressions with Perl Examples

"===Regular Expressions with Perl Examples===" is used in a novel way in this article. I'd have reserved that markup for section headings. - Bevo 03:20, 2 Mar 2004 (UTC)

Perl - Practical Extraction and Report Language

The manpage for Perl begins with:

NAME
       perl - Practical Extraction and Report Language

This article suggests that "Practical Extraction and Report Language" is deprecated and "was intended humorously". I see no evidence that this is the case. I intend to move "Practical Extraction and Report Language" into the intro and remove the presumably false statement if there are no objections.--Eloquence * 02:24, Apr 13, 2004 (UTC)

I have strong objections, because Perl is not an acronym. The etymology of the name is very clear and would be widely known if people weren't so desirous of ignorance. All one has to do is read chapter 27 of the 3rd edition of the camel book, or google around a bit. The name of the language was originally to be "Pearl", before Larry heard of another language known as "PEARL". He removed a letter and all was well in the world. "Practical Extraction and Report Language" is a post-facto rationalization. Perl is not an acronym. -- nknight 06:53, 13 Apr 2004 (UTC)

It doesn't matter if it is a "post-facto rationalization". We are not an etymological guide and the history of the name is only of marginal interest to most readers. If this is the way the word Perl is resolved in manuals today, then it should be prominently mentioned, not falsely claimed to be "intended humorously". I can agree to include the fact that Perl was technically not intended to be an acronym, but it is also a fact that it is today widely used as one, including by the makers of Perl.--Eloquence * 06:57, Apr 13, 2004 (UTC)

Feel free to remove any discussion about the name itself, but I will strongly oppose any attempt to define Perl as an acronym of "Practical Extraction and Report Language", as it is simply false, as anyone that has done substantive research on the subject will tell you. -- nknight 07:06, 13 Apr 2004 (UTC)

That's just plain silly in a case where the official documentation uses the term as an acronym, but I can agree to label it a backronym if that makes you happy.--Eloquence * 07:15, Apr 13, 2004 (UTC)

Etymological information is of encylopedic value, so don't remove anything of it. User:Eloquence can refer the discussions above under sections "[Pp]erl" and "Name" and take inputs from it. Jay 10:32, 13 Apr 2004 (UTC)

Explanation of for/foreach Syntax

I understand that this isn't supposed to be an exhaustive explanation of Perl syntax, but for/foreach are much more flexible and useful than this suggests. At the very least, I suggest replacing the "foreach ( array )" with "foreach ( LIST )", which is more accurate. (fish_at_uc_org)

I just noticed the same thing and fixed it. --Diberri | Talk 17:17, Aug 16, 2004 (UTC)

Technically, ARRAY is more accurate. If you give a literal list, it's first copied/aliased into an array, then the array is walked. -- Randal Schwartz

After I read your post, I immediately reverted my change (thus replacing list with array). But I'm not sure that was the right thing to do, and I don't believe you were even suggesting such a reversion. (Namely because foreach( array ) would disallow useful constructs like foreach( qw/foo bar baz/ ) and even foreach( foo() ), which Perl obviously allows.) So I've reverted my original reversion; the article now uses foreach( list ). --Diberri | Talk 06:40, Aug 23, 2004 (UTC)

Yeah, list is more appropriate and I believe Perl does copy the contents of the list given in a foreach context. -- Sundar 10:06, Aug 23, 2004 (UTC)

As I understand it, that part of the article is trying to explain the syntax of for, so the internals of the implementation, and the copying to a temporary array, are irrelevant. Syntactically, the target of the for is a list, not an array. Compare this with the target of pop or tie, which is syntactically an array. -- Dominus 13:44, 23 Aug 2004 (UTC)

I agree. The array copy is an implementation detail. PhilHibbs 09:03, 24 Aug 2004 (UTC)

It is undoubtedly an implementation detail, but nevertheless quite an important one affecting the code meaning, so the list-like syntax notwithstanding, maybe it should be stated that in fact for iterates over an array, not a list:

@a=(1..10); for (@a) {@a=(); print} # prints only 1

Rafał Pocztarski 18:07, 24 Aug 2004 (UTC)

Saying that perl expects a LIST at that point doesn't imply that an array is being flattened into a LIST, does it? An array is still a LIST. (fish_at_uc_org)

You make a good point, but I'm not sure this article (or any article on a programming language) should get into such minutiae. Since arrays are lists, but not vice versa, foreach( list ) should be used because 1) it's correct, even considering minute details, and 2) it's simpler than saying foreach( array ) and then having to qualify it. --Diberri | Talk 00:55, Sep 15, 2004 (UTC)

Sample code formatting

Regarding Ævar Arnfjörð Bjarmason's changes to the code formatting (using <tt> instead of <pre>), I think I prefer the look that the stylesheet imparts to <pre> blocks. --Yath 23:26, 15 Aug 2004 (UTC)

Logo sugegstion

perl.com logo (http://www.perl.com)

<div style="float:right;width:97px;text-align:center">
[[Image:Programming-republic-of-perl.gif|Programming Republic Of Perl]]
<br>
<small>
[http://www.perl.com perl.com logo]
</small>
</div>

Would the <div> code above be acceptable as an alternative to the bare image link, given the copyright restriction on the image (which I have expressed concerns about on the image talk page)?

No need for explicit DIV, just use the WikiImage code (as I have done). HTH HAND --Phil | Talk 13:24, Aug 20, 2004 (UTC)

OK, I was trying to avoid the frame, but I guess that's in keeping with the rest of the 'pedia. In fact, I think it's better to have the frame, as it makes it clear that we are not using it as an approval stamp.PhilHibbs 13:35, 20 Aug 2004 (UTC)

Merits and demerits of Perl as a programming language

Here's some text I removed, from a section named "Merits and demerits of Perl as a programming language":

Like all programming languages, perl does some things well, and other things, well, less than efficiently. Many programmers have come to terms with this and use a judicious mix of shell-script and perl to get around the more obvious deficiencies (date manipulation can be slow in terms of relative processing time, for example). On the upside, perl is very useful for low level file manipulation at which, in many respects, its performance is often comparable with e.g. sed.

I'm sure a section like ==Tasks for which Perl is unsuited== could be useful and informative, but the above paragraph has no redeeming qualities. Allow me to nitpick.

a judicious mix of shell-script and perl to get around the more obvious deficiencies -- Perl replaces and exceeds sh in nearly every capacity. When perl falls short, you either switch languages or call a utility, but the instances in which that utility would be sh are vanishingly rare.
date manipulation can be slow in terms of relative processing time, for example -- perhaps this is a rare example of when sh is better than Perl? If you can demonstrate this, I won't object to it being reinserted into the article.
perl is very useful for low level file manipulation -- This is too vague. By "low level file manipulation", do you mean manipulating links and permissions, or something even lower? Perl is ok for fiddling with links and permissions, I guess, but as a strength, it's no big deal. For lower-level operations... well, you'll have to clarify if that's what you meant.
at which, in many respects, its performance is often comparable with e.g. sed. sed? Perhaps you meant data manipulation. This is one of the foundations of Perl and is well covered elsewhere in the article.

--Yath 08:01, 14 Sep 2004 (UTC)

Yeah, I agree with Yath -- Sundar 08:12, Sep 14, 2004 (UTC)

I also agree. There's a lot of be said about Perl's failings and deficiencies (see Sins of Perl Revisited (http://www.perl.com/pub/a/1999/11/sins.html), for example) but I don't think this is it. -- Dominus 13:14, 14 Sep 2004 (UTC)

Web hosting advertisements

I am removing a link to a list of Web hosting advertisements ("Webmaster.org: Perl Web Hosting Plans (http://www.webmaster.org/webhosting/plans/perl.htm) - List of companies offering perl in their web hosting plans.") added to External links by 69.158.155.131 whose every single edit was adding links to similar lists in other articles. Rafał Pocztarski 09:41, 14 Nov 2004 (UTC)

Data types and Variables

With all the Perl hackers around, we have an article on Perl that does not explain, or even do more than mention in passing, Perl data types and variables. The word data only appears as part of "database", "type" doesn't appear at all, and "variable" only appears in a misdescription of function parameters. I've half a mind to submit this article for deletion, if this is all there is to show after three years. --Tony Sidaway 11:30, 26 Nov 2004 (UTC)

Are you kidding? If incompleteness was grounds for deletion, Wikipedia would dissolve. :-) Sure, the article needs work. So be bold! For now, I've added a brief bit on built-in data types ($scalars, @arrays, and %hashes), but there's lots of room for improvement. Why don't you take a stab at it? --David Iberri | Talk 09:18, Nov 27, 2004 (UTC)

I don't think adding yet more of the same is going to make this into an encyclopedia article. I think it would probably make more sense to wikify the perldoc files (see User:Tony_Sidaway/Projects/man_perl/man_perl for an example based on Perl 5.6.1, with an example of a link to man perlbook) and reserve the main Perl article for a brief language summary as is the case with languages like Fortran, lisp and so on. --[[User:Tony Sidaway|Tony Sidaway ^Talk]] 17:27, 27 Nov 2004 (UTC)

I'm not sure I understand what you mean bo "adding yet more of the same". If you mean that the existing article is a bit too low-level (e.g. the regular expression section needs more commentary and fewer examples, IMO) and that the built-in data types section I added is similarly too low-level, then I might agree. But I'm not sure that a brief summary would suffice. Wiki is not paper, so I don't think we need to be overly concise. Examples are good, as demonstrated by Lisp programming language. --David Iberri | Talk 19:44, Nov 27, 2004 (UTC)

Someone coming to Perl should be able to find out what he or she needs to know. The type of language it is, what people use it for, what it looks like, in general terms the kinds of facilities it provides to the user, the design considerations of the language, and its history.

What is there would probably be all right if it were more concise and (above all) accurate. The section on subroutines, for instance, erroneously refers to "Variables passed to a subroutine"--an error that could be forgiven in a first year Computer Science student, but is not to be expected in an encyclopedia.

I think I'll continue with my efforts to wikify man perl and, once I have completed that task to my satisfaction (I'll be using scripting to do the translation to wiki), I'll compose a rewrite of Perl that refers the interested reader to the appropriate sections of that copious and very readable electronic manual of the Perl language. This should provide the best of both worlds. --[[User:Tony Sidaway|Tony Sidaway ^Talk]] 21:21, 27 Nov 2004 (UTC)

Time to split the page?

The Perl page is at 30K. It's long to read and big to edit. Would it make sense to split some of the big sections, e.g.

off into separate pages? Swmcd 16:19, 2005 Mar 1 (UTC)

Yep :-)

chocolateboy 19:34, 1 Mar 2005 (UTC)

Obfuscation example

That's a really, really weak example of obfuscation given, especially after just mentioning the annual obfuscation contest. Either give a real example, or don't give an example at all.

I assume that this pathetic example (which actually serves to confuse rather than enlighten) was included because all the good ones are of course copyright their authors and hence incompatible with Wikipedia. On the other hand, I'm sure at least some of those authors would be willing to GFDL their Obfuscated Perl entries for Wikipedia. I have one in mind; I may go and ask the author. --Pete

[Time Passes] I've asked for and got permission to use the example I was thinking of. I'll include the text here so that in future people will know it was used with permission.

> for your program to be included in an article you would  
> need to license it under the Gnu Free Documentation License, a  
> suitable Creative Commons license, or similar.

I am happy to license it to you under either or both of those licenses.

[ --Mark Jason Dominus]

Your effort is appreciated. -- Sundar (talk · contributions) 03:43, May 13, 2005 (UTC)

propagandistic errors

i think there are many propagandistic errors in this article. Here's one example: quote:“...Any syntax errors are caught during the compile stage instead of later during execution.” That's kinda ridiculous or meaningless to claim without specifying exactly what is meant by “any“. Quote: “Subroutine calls can be placed in the file before the subroutines themselves are defined.” This is not true always. Xah Lee 22:37, 2005 Mar 30 (UTC)

I disagree. Any error that isn't picked up at the compile stage isn't a syntax error. Care to give a counter-example? Davorg 10:45, 31 Mar 2005 (UTC)

The problem isn't that it's false, but that it isn't notable. Syntax errors are generally caught during the compile stage of any language, so to note this as a feature of perl is strange. It's even stranger still when you consider that the "eval" function can make it false for perl. --Yath 19:56, 31 Mar 2005 (UTC)

It's notable because many of the languages Perl derives from (like shell scripts) interpret their programs line-by-line. Since they don't have a formal parse phase, a syntax error halfway through your script will leave the task half-finished. —Brent Dax 02:16, 26 Apr 2005 (UTC)

But those languages don't have a compile phase. I found that part of the article to be strange for the same reason that Yath did. -- Dominus 13:53, 26 Apr 2005 (UTC)

Quote: “Subroutine calls can be placed in the file before the subroutines themselves are defined.”

this is not true always.

Again, care to give an example? Davorg 10:45, 31 Mar 2005 (UTC)

 use File::Find qw(find);
 $mydir= '/Users/t/web/SpecialPlaneCurves_dir';
 find(\&wanted, $mydir);
 sub g($){print shift, "\n";}
 sub wanted {
     if ($_ =~/\.html$/ && -T $File::Find::name) { g $File::Find::name;}
     $File::Find::name;
 }

http://xahlee.org/perl-python/traverse_dir.html Xah Lee 10:00, 2005 Apr 1 (UTC)

Of course, all you need to do to fix that it to put parentheses around the arguments to 'g'. So it's hardly a compelling argument.

      #!/usr/bin/perl
      use File::Find;
      $mydir= '.';
      find(\&wanted, $mydir);
      sub wanted {
           g($File::Find::name) if /\.html$/ && -T $File::Find::name;
         $File::Find::name;
      }
      sub g($) {print shift, "\n";}

Davorg 16:59, 1 Apr 2005 (UTC)

Xah Lee 22:37, 2005 Mar 30 (UTC)

History

Perhaps there should be some information on the history of the language (when was it first introduced, for example) --24.21.243.107 20:43, 2 Apr 2005 (UTC)

Copyedits

1.2 Applications says

   Systems administrators use Perl as an all-purpose tool

Generic Player qualified this to

   Many systems administrators use Perl as an all-purpose tool

on the grounds that not all systems administrators use Perl. I took "Many" back out.

Unqualified, "systems administrators" doesn't mean "all systems administrators", it refers to systems administrators as a group, and asserts that use of Perl is observed among that group. Adding the qualifier "many" actually asserts knowledge of the prevalence of Perl use among systems administrators, and we shouldn't do that unless we have evidence to support the assertion.

Retrieved from "https://academickids.com:443/encyclopedia/index.php/Talk:Perl"