Партнёры:
Хостинг:
Archive-name: perl-faq/part3
Version: $Id: part3,v 2.7 1995/05/15 15:44:43 spp Exp spp $
Posting-Frequency: bi-weekly
Last Edited: Thu Jan 11 00:55:42 1996 by spp (Stephen P Potter) on syrinx.psa.com
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-b8.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.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 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 Beaching It
spp@psa.pencom.com Pager: 1-800-759-8888, 547-9561 Work: 703-860-2222
Cthulhu for President in '96: When You're Tired of the Lesser of Two Evils
--
Stephen P Potter Pencom Systems Administration Beaching It
spp@psa.pencom.com Pager: 1-800-759-8888, 547-9561 Work: 703-860-2222
"I don't care whether people actually like Perl, just so long as they *think*
they like it... ;-)" -Larry Wall
|
Закладки на сайте Проследить за страницей |
Created 1996-2024 by Maxim Chirkov Добавить, Поддержать, Вебмастеру |