Debugger: I must be missing something

throop has asked for the wisdom of the Perl Monks concerning the following question:

Brethern,

I must be missing something about using the debugger

Setting breakpoints in other files — I'm debugging. While looking at function1 in file1, I realize the problem is likely happening on line 100 of file2, which is inside function2. I don't know how to set a breakpoint of a file different than the one I'm in. So I breakpoint package2::function2, continue till it breaks, clear the breakpoint, set a new breakpoint at line 100. Is there a better way?
Similarly, if I use L to show me the breakpoints, I can see breakpoints in other files, but I can't see how to use B to clear them.
Taming the printing of @_ — I've arrived at line 100. I want to have a look at @_. But there's a pointer to some big hairy datastructure being passed as the 3rd argument. 'x @_' prints way too much text on my screen, and scrolls the first two variables offscreen. Looking at the stack is better, unless the stack is tall (which it often is) or if there are more than four variables in the call. Is there a good way to see a terse representation of @_ ? Or a way to just print the last three layers of the stack?
Fumble fingers and stacktrace — At line 100, I'm trying to match $_ against a long regex. It ought to be matching, but it isn't. I copy and paste part of the regex into the debugger commandline, trying to figure out what's failing. But I fumble finger it and the regex isn't well formed. The debugger now prints out a stack trace, with six lines of DB and Eval, followed by the stack of function2. Scrolling my recent history out of sight. Darn. All I really needed was it to ring the terminal bell, or print out 'OOPS!' or something. Is there some way to tame this?
One key history — In the tcsh shell, I can move through the command history with the uparrow key, and previous commands with the other arrow keys. Any way to do that within the debugger?
Emacs — perldebug says "If you have the FSF's version of emacs installed on your system, it can interact with the Perl debugger to provide an integrated software development environment. " Sounds good. Where do I go next?

throop

Comment on Debugger: I must be missing something

Replies are listed 'Best First'.

Re: Debugger: I must be missing something
by grep (Monsignor) on Dec 08, 2006 at 17:18 UTC

Setting breakpoints in other files

Taming the printing of @_ — You're in a perl interpreter so, p $_[3]->{foo} will work. I also will use Data::Dumper; p Dumper \@_ in the debugger. I set my terminal to scoll 5000 lines to scroll over the result.

Fumble fingers and stacktrace — I would setup another terminal window with a small script that runs the part of the regex you're interested in. You can w $data to get the current value being evaluated. Then test it in you other window.

One key history — The debugger will repeat the last command if you just hit return.

grep

XP matters not. Look at me. Judge me by my XP, do you?

[reply]
[d/l]
[select]

Re: Debugger: I must be missing something
by zentara (Cardinal) on Dec 08, 2006 at 17:43 UTC

ptkdb

I'm not really a human, but I play one on earth. Cogito ergo sum a bum

[reply]

Re: Debugger: I must be missing something
by grinder (Bishop) on Dec 08, 2006 at 21:54 UTC

Setting breakpoints in other files

f file2 # switch to viewing other file2
b 100 # set a breakpoint at file2:100
r # go!
[download]

To delete a breakpoint in another file, use f to switch to viewing the source, and then run B 100 to remove it. Yes, it would be nice to be able to say B file2:100, but I have it on good authority that patching the debugger will drive you insane.

One key history

You must be missing a module. Term::ReadLine::Perl or something. Otherwise you would already have it.

• another intruder with the mooring in the heart of the Perl

[reply]
[d/l]

Re^2: Debugger: I must be missing something

by pemungkah (Priest) on Dec 09, 2006 at 23:38 UTC

%INC

$ perl -de0

Loading DB routines from perl5db.pl version 1.28
Editor support available.

Enter h or `h h' for help, or `man perldebug' for more help.

main::(-e:1):   0
DB<1> x %INC
0  'warnings/register.pm'
1  '/usr/local/lib/perl5/5.8.8/warnings/register.pm'
2  'attributes.pm'
3  '/usr/local/lib/perl5/5.8.8/attributes.pm'
4  'XSLoader.pm'
5  '/usr/local/lib/perl5/5.8.8/darwin-2level/XSLoader.pm'
6  'IO/Handle.pm'
7  '/usr/local/lib/perl5/5.8.8/darwin-2level/IO/Handle.pm'
8  'Term/Cap.pm'
9  '/usr/local/lib/perl5/5.8.8/Term/Cap.pm'
10  'SelectSaver.pm'
11  '/usr/local/lib/perl5/5.8.8/SelectSaver.pm'
12  'Term/ReadLine/Perl.pm'
13  '/usr/local/lib/perl5/site_perl/5.8.8/Term/ReadLine/Perl.pm'
14  'warnings.pm'
15  '/usr/local/lib/perl5/5.8.8/warnings.pm'
16  'Carp/Heavy.pm'
17  '/usr/local/lib/perl5/5.8.8/Carp/Heavy.pm'
18  'Config.pm'
19  '/usr/local/lib/perl5/5.8.8/darwin-2level/Config.pm'
20  'Symbol.pm'
21  '/usr/local/lib/perl5/5.8.8/Symbol.pm'
22  'IO.pm'
23  '/usr/local/lib/perl5/5.8.8/darwin-2level/IO.pm'
24  'Carp.pm'
25  '/usr/local/lib/perl5/5.8.8/Carp.pm'
26  'Term/ReadLine.pm'
27  '/usr/local/lib/perl5/5.8.8/Term/ReadLine.pm'
28  'vars.pm'
29  '/usr/local/lib/perl5/5.8.8/vars.pm'
30  'strict.pm'
31  '/usr/local/lib/perl5/5.8.8/strict.pm'
32  'Exporter.pm'
33  '/usr/local/lib/perl5/5.8.8/Exporter.pm'
34  'Config_heavy.pl'
35  '/usr/local/lib/perl5/5.8.8/darwin-2level/Config_heavy.pl'
36  'Term/ReadLine/readline.pm'
37  '/usr/local/lib/perl5/site_perl/5.8.8/Term/ReadLine/readline.pm'
38  'Term/ReadKey.pm'
39  '/usr/local/lib/perl5/site_perl/5.8.8/darwin-2level/Term/ReadKey.p
+m'
40  'AutoLoader.pm'
41  '/usr/local/lib/perl5/5.8.8/AutoLoader.pm'
42  'perl5db.pl'
43  '/usr/local/lib/perl5/5.8.8/perl5db.pl'
44  'DynaLoader.pm'
45  '/usr/local/lib/perl5/5.8.8/darwin-2level/DynaLoader.pm'
46  'SelfLoader.pm'
47  '/usr/local/lib/perl5/5.8.8/SelfLoader.pm'

DB<2>
[download]

f

Edit: Devel::Command::NewF extends the f command to automatically look up the module in %INC, translating the name for you from Whatever::It::Is to Whatever/It/Is.pm, and loading the module if it isn't found.

$ perl -de0
Patching with Devel::Command::DBSub::DB_5_8_5

Loading DB routines from perl5db.pl version 1.27
Editor support available.

Enter h or `h h' for help, or `man perldebug' for more help.

main::(-e:1):    0
  DB<1> fx Test::More
Loaded Test::More
auto(-1)  DB<2> f /home/y/lib/perl5/5.8/i386-freebsd-thread-multi/Test
+/More.pm
1     package Test::More;
2     
3:    use 5.004;
4     
5:    use strict;
6     
7     
8     # Can't use Carp because it might cause use_ok() to accidentally
+ succeed9     # even though the module being used forgot to use Carp.
+  Yes, this
10     # actually happened.
  DB<3>
[download]

[reply]
[d/l]
[select]

Re: Debugger: I must be missing something (few)
by tye (Sage) on Dec 09, 2006 at 07:28 UTC

To set a breakpoint in another file, view some source in that file and then set the breakpoint. In your specific example, an easy way to do that would be:

l package2::function2
b 100
[download]

I often use x "@_" to get a summary of arguments without having to worry about some huge data structure being dumped.

You can turn down the default "die reporting level" so that dieing, for whatever reason, doens't spew so much text: "O die=0". You might also want to turn down the warning reporting level: "O warn=0".

- tye

[reply]
[d/l]
[select]

Re: Debugger: I must be missing something
by jbert (Priest) on Dec 08, 2006 at 18:16 UTC

Meta-x perldb<RET>

Does anyone know of anything similar for vim? (looks like vim.org is having MySQL problems at the moment).

[reply]
[d/l]

Re: Debugger: I must be missing something
by Popcorn Dave (Abbot) on Dec 08, 2006 at 18:06 UTC

Zentara's

ptkdb

Besides that small niggle, I'd be lost without it in my toolkit.

Revolution. Today, 3 O'Clock. Meet behind the monkey bars.

[reply]

Re: Debugger: I must be missing something
by aufflick (Deacon) on Dec 11, 2006 at 12:06 UTC

grep

p Dumper ...

$Data::Dumper::Maxdepth

Data::Dumper

[reply]

Re: Debugger: I must be missing something
by sfink (Deacon) on Dec 11, 2006 at 18:19 UTC

Taming the printing of @_

Do it in stages. Avoid 'x' as long as you need to if there's a possibility of an enormous data structure:

  <DB> p "@_"
  ARRAY(0x38234) Big::Nested::Mess=HASH(0x382437)
  <DB> x keys %{ $_[1] }
  ...just the keys, so you can see what to look for...
[download]

sub p {
  my $x = shift;
  if (! ref $x) {
    return $x;
  } elsif (ref($x) eq 'My::Class') {
    return bless { name => $x->{name},
                   manager => $x->{manager}->getId(),
                   children => p($x->{children}),
                   # add any other fields you want to see
                 }, ref($x);
  } elsif (UNIVERSAL::isa($x, 'ARRAY')) {
    return [ map { p($_) } @$x ];
  } elsif (UNIVERSAL::isa($x, 'HASH')) {
    return { map { $_ => p($x->{$_}) } keys %$x };
  } else {
    return $x;
  }
}
[download]

Then you can do 'x p($whatever)' and get your customized dump. (You can take this farther, and pass in a "brevity level" parameter that starts out zero to display everything, but increases as you get into further nested things so that they will display as one-line summaries rather than structural dumps.)

[reply]
[d/l]
[select]