comp.lang.perl.* FAQ 3/5 - Programming Aids



Archive-name: perl-faq/part3
Version: $Id: perl.faq.3.html,v 1.1 2006-04-22 14:12:23 rmoffitt Exp $
Posting-Frequency: bi-weekly
Last Edited: Wed May 24 14:36:30 1995 by spp (Stephen Potter) on squid

This posting contains answers to general information questions, mostly
about programming aids.  


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 		<lisch@sysserver1.mentor.com>
	Kresten Krab Thorup 	<krab@iesd.auc.dk>  

    The original articles by these folks containing their profilers are
    available at ftp://convex.com/pub/perl/info/profiling.shar.

    Recently, Dean Roerich* 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
    ftp://ftp.sterling.com/local/perl-byacc1.8.2.tar.Z, 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: 

	PERL|perl|Perl:\
	    :pb=^\d?(sub|package)\d\p\d:\
	    :bb={:be=}:cb=#:ce=$:sb=":se=\e":lb=':\
	    :le=\e':tl:\
	    :id=_:\
	    :kw=\
	    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 <levine@ics.uci.edu> 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 $# doesn't get
    # interpreted as one).
    perl4:\
	    :pb=^\d?(sub|package)\d\p\d:\
	    :id=$%@_:\
	    :bb=\e{:be=\e}:cb=\d\e#:ce=$:sb=\e":se=\e":\
	    :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:

	A) speed
	B) secrecy
	C) maintainability

    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.

	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.


    SOURCE-CODE SECRECY:

    2)  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 (webber@lnk.com) for more
    	information.
    	
    3)  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.

    4)  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.

    MAINTAINABILITY:

    5)  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.

    6)  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 <dls@best.com> 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 ftp.best.com 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):

	ftp://ftp.ncsu.edu/pub/math/wsetzer/cursperl5a6.tar.gz
        ftp://ftp.metronet.com/pub/perlinfo/perl5/cursperl5a6.tar.gz
        ftp://ftp.cs.ruu.nl/pub/PERL/perl5.0/cursperl5a6.tar.gz

    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 ("menu.pl") via
    anonymous FTP from "ftp.iastate.edu".  It's in the directory /pub/perl
    as:

        menu.pl.v3.1.tar.Z

    menu.pl 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.

    menu.pl 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
    ftp.cs.ruu.nl [131.211.80.17] 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 mail-server@cs.ruu.nl
    containing these lines:

	begin
	send PERL/cterm.shar.Z
	end

    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

    	ftp://ftp.wpi.edu/perl5/private/Tk-b3.tar.gz

    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 Frederic Chauveau*.  It's available from these sites:

        ftp://ftp.pasteur.fr/pub/Perl/Sx.tar.gz
        ftp://ftp.khoros.unm.edu/pub/perl/extensions/Sx.tar.gz
        ftp://ftp.metronet.com/pub/perlinfo/perl5/Sx.tar.gz
 
    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 
	ftp.wu-wien.ac.at[137.208.3.5]:pub/src/X11/wafe-0.9.tar.Z

    Alternatively, you could use wish from tcl. 

    #!/usr/local/bin/perl
    #####################################################################
    #  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 dov@menora.weizmann.ac.il
    #  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;
	    select(STDOUT);

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

	# Create the widgets
	print WISH <<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}

TCL
	;
	# Here is the main loop which receives and sends commands
	# to wish.
	while (<WISH>) {
	    chop;
	    print "Wish sais: <$_>\n";
	    if (/^quit/) { print WISH "destroy .\n"; last; }
	}
	    wait;
    } elsif (defined $pid) {
	open(STDOUT, ">&W0");
	open(STDIN, ">&W0");
	close(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 libperl.so library and
    then your /usr/local/bin/perl binary reduced to just 50k or so.
    See DynLoader(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.h 
	      dyl.c - code extracted from Oliver Sharp's article

	      hash.h
	      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 -          "
	     test.pl - 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
    dyl_find() 

    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, test.pl 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 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 ftp.ee.umanitoba.ca in the
    directory /pub/msdos/perl/hbp_30.zip.  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 $PATH | perl5 -nl -072 -e '
	s![^/+]*$!man!&&-d&&!$s{$_}++&&push@m,$_;END{print"@m"}'

    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...

	&$h("world");
	&$g("earthlings");

    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.





--
Stephen P Potter				Pencom Systems Administration
spp@epix.net					       steve_potter@psisa.com
Current Assignment: AMP Inc, Technology Division         steve.potter@amp.com