perldebug - Perl debugging |
perldebug - Perl debugging
First of all, have you tried using the -w switch?
If you're new to the Perl debugger, you may prefer to read the perldebtut manpage, which is a tutorial introduction to the debugger .
If you invoke Perl with the -d switch, your script runs under the Perl source debugger. This works like an interactive Perl environment, prompting for debugger commands that let you examine source code, set breakpoints, get stack backtraces, change the values of variables, etc. This is so convenient that you often fire up the debugger all by itself just to test out Perl constructs interactively to see what they do. For example:
$ perl -d -e 42
In Perl, the debugger is not a separate program the way it usually is in the typical compiled environment. Instead, the -d flag tells the compiler to insert source information into the parse trees it's about to hand off to the interpreter. That means your code must first compile correctly for the debugger to work on it. Then when the interpreter starts up, it preloads a special Perl library file containing the debugger.
The program will halt right before the first run-time executable statement (but see below regarding compile-time statements) and ask you to enter a debugger command. Contrary to popular expectations, whenever the debugger halts and shows you a line of code, it always displays the line it's about to execute, rather than the one it has just executed.
Any command not recognized by the debugger is directly executed
(eval
'd) as Perl code in the current package. (The debugger
uses the DB package for keeping its own state information.)
For any text entered at the debugger prompt, leading and trailing whitespace
is first stripped before further processing. If a debugger command
coincides with some function in your own program, merely precede the
function with something that doesn't look like a debugger command, such
as a leading ;
or perhaps a +
, or by wrapping it with parentheses
or braces.
The debugger understands the following commands:
h h
produces the entire help page, which is quite long.
If the output of the h h
command (or any command, for that matter) scrolls
past your screen, precede the command with a leading pipe symbol so
that it's run through your pager, as in
DB> |h h
You may change the pager which is used via o pager=...
command.
print {$DB::OUT} expr
in the current package. In particular,
because this is just Perl's own print
function, this means that nested
data structures and objects are not dumped, unlike with the x
command.
The DB::OUT
filehandle is opened to /dev/tty, regardless of
where STDOUT may be redirected to.
print
function in Perl. When dumping
hashes, you'll probably prefer 'x \%h' rather than 'x %h'.
See the Dumpvalue manpage if you'd like to do this yourself.
The output format is governed by multiple options described under Configurable Options.
If the maxdepth
is included, it must be a numeral N; the value is
dumped only N levels deep, as if the dumpDepth
option had been
temporarily set to N.
main
)
using a data pretty-printer (hashes show their keys and values so
you see what's what, control characters are made printable, etc.).
Make sure you don't put the type specifier (like $
) there, just
the symbol names, like this:
V DB filename line
Use ~pattern
and !pattern
for positive and negative regexes.
This is similar to calling the x
command on each applicable var.
V currentpackage [vars]
.
mY
variables)
in the current scope or level scopes higher. You can limit the
variables that you see with vars which works exactly as it does
for the V
and X
commands. Requires the PadWalker
module
version 0.08 or higher; will warn if this isn't installed. Output
is pretty-printed in the same style as for V
and the format is
controlled by the same options.
PrintRet
option is set (default).
n
or s
command.
incr+1
lines starting at min
.
min
through max
. l -
is synonymous to -
.
eval
statement. If filename
is not a full pathname found in the values of %INC, it is considered
a regex.
eval
ed strings (when accessible) are considered to be filenames:
f (eval 7)
and f eval 7\b
access the body of the 7th eval
ed string
(in the order of execution). The bodies of the currently executed eval
and of eval
ed strings that define subroutines are saved and thus
accessible.
AutoTrace
option).
expr
.
See Frame Listing Output Examples in the perldebguts manpage for examples.
if
:
b 237 $x > 30 b 237 ++$count237 < 11 b 33 /pattern/i
1. check for a breakpoint at this line 2. print the line if necessary (tracing) 3. do any actions associated with that line 4. prompt user if at a breakpoint or in single-step 5. evaluate line
For example, this will print out $foo every time line 53 is passed:
a 53 print "DB FOUND $foo\n"
1
.
o
pager="less -MQeicsNfr"
to call less with those specific options.
You may use either single or double quotes, but if you do, you must
escape any embedded instances of same sort of quote you began with,
as well as any escaping any escapes that immediately precede that
quote but which are not meant to escape the quote itself. In other
words, you follow single-quoting rules irrespective of the quote;
eg: o option='this isn\'t bad'
or o option="She said, \"Isn't
it?\""
.
For historical reasons, the =value
is optional, but defaults to
1 only where it is safe to do so--that is, mostly for Boolean
options. It is always better to assign a specific value using =
.
The option
can be abbreviated, but for clarity probably should
not be. Several options can be set together. See Configurable Options
for a list of these.
command
is missing, all actions are wiped out!
command
is
missing, all actions are wiped out!
command
is missing, all actions are wiped out!
Because this command is in some senses new, a warning is issued if
you appear to have accidentally entered a block instead. If that's
what you mean to do, write it as with ;{ ... }
or even
do { ... }
.
o recallCommand
, too.
o shellBang
, also. Note that the user's current shell (well,
their $ENV{SHELL}
variable) will be used, which can interfere
with proper interpretation of exit status or signal and coredump
information.
source
commands.
exit
twice might work.
Set the inhibit_exit
option to 0 if you want to be able to step
off the end the script. You may also need to set $finished to 0
if you want to step through global destruction.
exec()
ing a new session. We try to maintain
your history across this, but internal settings and command-line options
may be lost.
The following setting are currently preserved: history, breakpoints, actions, debugger options, and the Perl command-line options -w, -I, and -e.
|dbcmd
but DB::OUT is temporarily select
ed as well.
= quit q
or list current aliases.
Config
information
is used to invoke man using the proper MANPATH or -M
manpath option. Failed lookups of the form XXX
that match
known manpages of the form perlXXX will be retried. This lets
you type man debug
or man op
from the debugger.
On systems traditionally bereft of a usable man command, the debugger invokes perldoc. Occasionally this determination is incorrect due to recalcitrant vendors or rather more felicitously, to enterprising users. If you fall into either category, just manually set the $DB::doccmd variable to whatever viewer to view the Perl documentation on your system. This may be set in an rc file, or through direct assignment. We're still waiting for a working example of something along the lines of:
$DB::doccmd = 'netscape -remote http://something.here/';
The debugger has numerous options settable using the o
command,
either interactively or from the environment or an rc file.
(./.perldb or ~/.perldb under Unix.)
recallCommand
, ShellBang
!
, which is unfortunate.
pager
|
character.) By default, $ENV{PAGER}
will be used.
Because the debugger uses your current terminal characteristics
for bold and underlining, if the chosen pager does not pass escape
sequences through unchanged, the output of some debugger commands
will not be readable when sent through the pager.
tkRunning
signalLevel
, warnLevel
, dieLevel
To disable this default safe mode, set these values to something higher
than 0. At a level of 1, you get backtraces upon receiving any kind
of warning (this is often annoying) or exception (this is
often valuable). Unfortunately, the debugger cannot discern fatal
exceptions from non-fatal ones. If dieLevel
is even 1, then your
non-fatal exceptions are also traced and unceremoniously altered if they
came from eval'd
strings or from any kind of eval
within modules
you're attempting to load. If dieLevel
is 2, the debugger doesn't
care where they came from: It usurps your exception handler and prints
out a trace, then modifies all exceptions with its own embellishments.
This may perhaps be useful for some tracing purposes, but tends to hopelessly
destroy any program that takes its exception handling seriously.
AutoTrace
t
command, but can be put into
PERLDB_OPTS
).
LineInfo
|visual_perl_db
), then a short message is used. This is the
mechanism used to interact with a slave editor or visual debugger,
such as the special vi
or emacs
hooks, or the ddd
graphical
debugger.
inhibit_exit
PrintRet
r
command if set (default).
ornaments
frame
frame & 2
is false, messages are printed on entry only. (Printing
on exit might be useful if interspersed with other messages.)
If frame & 4
, arguments to functions are printed, plus context
and caller info. If frame & 8
, overloaded stringify
and
tie
d FETCH
is enabled on the printed arguments. If frame
& 16
, the return value from the subroutine is printed.
The length at which the argument list is truncated is governed by the next option:
maxTraceLen
frame
option's
bit 4 is set.
windowSize
The following options affect what happens with V
, X
, and x
commands:
arrayDepth
, hashDepth
dumpDepth
compactDump
, veryCompact
compactDump
, short array
may be printed on one line.
globPrint
DumpDBFiles
DumpPackages
DumpReused
quote
, HighBit
, undefPrint
quote
is auto
; one can enable double-quotish or single-quotish format
by setting it to "
or '
, respectively. By default, characters
with their high bit set are printed verbatim.
UsageOnly
After the rc file is read, the debugger reads the $ENV{PERLDB_OPTS}
environment variable and parses this as the remainder of a `O ...'
line as one might enter at the debugger prompt. You may place the
initialization options TTY
, noTTY
, ReadLine
, and NonStop
there.
If your rc file contains:
parse_options("NonStop=1 LineInfo=db.out AutoTrace");
then your script will run without human intervention, putting trace
information into the file db.out. (If you interrupt it, you'd
better reset LineInfo
to /dev/tty if you expect to see anything.)
TTY
noTTY
NonStop
mode and will not connect to a TTY. If
interrupted (or if control goes to the debugger via explicit setting of
$DB::signal or $DB::single from the Perl script), it connects to a TTY
specified in the TTY
option at startup, or to a tty found at
runtime using the Term::Rendezvous
module of your choice.
This module should implement a method named new
that returns an object
with two methods: IN
and OUT
. These should return filehandles to use
for debugging input and output correspondingly. The new
method should
inspect an argument containing the value of $ENV{PERLDB_NOTTY}
at
startup, or "/tmp/perldbtty$$"
otherwise. This file is not
inspected for proper ownership, so security hazards are theoretically
possible.
ReadLine
NonStop
Here's an example of using the $ENV{PERLDB_OPTS}
variable:
$ PERLDB_OPTS="NonStop frame=2" perl -d myprogram
That will run the script myprogram without human intervention,
printing out the call tree with entry and exit points. Note that
NonStop=1 frame=2
is equivalent to N f=2
, and that originally,
options could be uniquely abbreviated by the first letter (modulo
the Dump*
options). It is nevertheless recommended that you
always spell them out in full for legibility and future compatibility.
Other examples include
$ PERLDB_OPTS="NonStop LineInfo=listing frame=2" perl -d myprogram
which runs script non-interactively, printing info on each entry
into a subroutine and each executed line into the file named listing.
(If you interrupt it, you would better reset LineInfo
to something
``interactive''!)
Other examples include (using standard shell syntax to show environment variable settings):
$ ( PERLDB_OPTS="NonStop frame=1 AutoTrace LineInfo=tperl.out" perl -d myprogram )
which may be useful for debugging a program that uses Term::ReadLine
itself. Do not forget to detach your shell from the TTY in the window that
corresponds to /dev/ttyXX, say, by issuing a command like
$ sleep 1000000
See Debugger Internals in the perldebguts manpage for details.
DB<8>
or even
DB<<17>>
where that number is the command number, and which you'd use to
access with the built-in csh-like history mechanism. For example,
!17
would repeat command number 17. The depth of the angle
brackets indicates the nesting depth of the debugger. You could
get more than one set of brackets, for example, if you'd already
at a breakpoint and then printed the result of a function call that
itself has a breakpoint, or you step into an expression via s/n/t
expression
command.
DB<1> for (1..4) { \ cont: print "ok\n"; \ cont: } ok ok ok ok
Note that this business of escaping a newline is specific to interactive commands typed into the debugger.
T
command might
look like:
$ = main::infested called from file `Ambulation.pm' line 10 @ = Ambulation::legs(1, 2, 3, 4) called from file `camel_flea' line 7 $ = main::pests('bactrian', 4) called from file `camel_flea' line 4
The left-hand character up there indicates the context in which the
function was called, with $
and @
meaning scalar or list
contexts respectively, and .
meaning void context (which is
actually a sort of scalar context). The display above says
that you were in the function main::infested
when you ran the
stack dump, and that it was called in scalar context from line
10 of the file Ambulation.pm, but without any arguments at all,
meaning it was called as &infested
. The next stack frame shows
that the function Ambulation::legs
was called in list context
from the camel_flea file with four arguments. The last stack
frame shows that main::pests
was called in scalar context,
also from camel_flea, but from line 4.
If you execute the T
command from inside an active use
statement, the backtrace will contain both a require
frame and
an eval
) frame.
l
command can produce:
DB<<13>> l 101: @i{@i} = (); 102:b @isa{@i,$pack} = () 103 if(exists $i{$prevpack} || exists $isa{$pack}); 104 } 105 106 next 107==> if(exists $isa{$pack}); 108 109:a if ($extra-- > 0) { 110: %isa = ($pack,1);
Breakable lines are marked with :
. Lines with breakpoints are
marked by b
and those with actions by a
. The line that's
about to be executed is marked by ==>
.
Please be aware that code in debugger listings may not look the same as your original source code. Line directives and external source filters can alter the code before Perl sees it, causing code to move from its original positions or take on entirely different forms.
frame
option is set, the debugger would print entered (and
optionally exited) subroutines in different styles. See the perldebguts manpage
for incredibly long examples of these.
If you have compile-time executable statements (such as code within
BEGIN and CHECK blocks or use
statements), these will not be
stopped by debugger, although require
s and INIT blocks will, and
compile-time statements can be traced with AutoTrace
option set
in PERLDB_OPTS
). From your own Perl code, however, you can
transfer control back to the debugger using the following statement,
which is harmless if the debugger is not running:
$DB::single = 1;
If you set $DB::single
to 2, it's equivalent to having
just typed the n
command, whereas a value of 1 means the s
command. The $DB::trace
variable should be set to 1 to simulate
having typed the t
command.
Another way to debug compile-time code is to start the debugger, set a breakpoint on the load of some module:
DB<7> b load f:/perllib/lib/Carp.pm Will stop on load of `f:/perllib/lib/Carp.pm'.
and then restart the debugger using the R
command (if possible). One can use b
compile subname
for the same purpose.
The debugger probably contains enough configuration hooks that you
won't ever have to modify it yourself. You may change the behaviour
of debugger from within the debugger using its o
command, from
the command line via the PERLDB_OPTS
environment variable, and
from customization files.
You can do some customization by setting up a .perldb file, which contains initialization code. For instance, you could make aliases like these (the last one is one people expect to be there):
$DB::alias{'len'} = 's/^len(.*)/p length($1)/'; $DB::alias{'stop'} = 's/^stop (at|in)/b/'; $DB::alias{'ps'} = 's/^ps\b/p scalar /'; $DB::alias{'quit'} = 's/^quit(\s*)/exit/';
You can change options from .perldb by using calls like this one;
parse_options("NonStop=1 LineInfo=db.out AutoTrace=1 frame=2");
The code is executed in the package DB
. Note that .perldb is
processed before processing PERLDB_OPTS
. If .perldb defines the
subroutine afterinit
, that function is called after debugger
initialization ends. .perldb may be contained in the current
directory, or in the home directory. Because this file is sourced
in by Perl and may contain arbitrary commands, for security reasons,
it must be owned by the superuser or the current user, and writable
by no one but its owner.
If you want to modify the debugger, copy perl5db.pl from the
Perl library to another name and hack it to your heart's content.
You'll then want to set your PERL5DB
environment variable to say
something like this:
BEGIN { require "myperl5db.pl" }
As a last resort, you could also use PERL5DB
to customize the debugger
by directly setting internal variables or calling debugger functions.
Note that any variables and functions that are not documented in this document (or in the perldebguts manpage) are considered for internal use only, and as such are subject to change without notice.
As shipped, the only command-line history supplied is a simplistic one that checks for leading exclamation points. However, if you install the Term::ReadKey and Term::ReadLine modules from CPAN, you will have full editing capabilities much like GNU readline(3) provides. Look for these in the modules/by-module/Term directory on CPAN. These do not support normal vi command-line editing, however.
A rudimentary command-line completion is also available. Unfortunately, the names of lexical variables are not available for completion.
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 reminiscent of its interactions with C debuggers.
Perl comes with a start file for making emacs act like a syntax-directed editor that understands (some of) Perl's syntax. Look in the emacs directory of the Perl source distribution.
A similar setup by Tom Christiansen for interacting with any vendor-shipped vi and the X11 window system is also available. This works similarly to the integrated multiwindow support that emacs provides, where the debugger drives the editor. At the time of this writing, however, that tool's eventual location in the Perl distribution was uncertain.
Users of vi should also look into vim and gvim, the mousey and windy version, for coloring of Perl keywords.
Note that only perl can truly parse Perl, so all such CASE tools fall somewhat short of the mark, especially if you don't program your Perl as a C programmer might.
If you wish to supply an alternative debugger for Perl to run, just invoke your script with a colon and a package argument given to the -d flag. The most popular alternative debuggers for Perl is the Perl profiler. Devel::DProf is now included with the standard Perl distribution. To profile your Perl program in the file mycode.pl, just type:
$ perl -d:DProf mycode.pl
When the script terminates the profiler will dump the profile information to a file called tmon.out. A tool like dprofpp, also supplied with the standard Perl distribution, can be used to interpret the information in that profile.
use re 'debug'
enables you to see the gory details of how the Perl
regular expression engine works. In order to understand this typically
voluminous output, one must not only have some idea about how regular
expression matching works in general, but also know how Perl's regular
expressions are internally compiled into an automaton. These matters
are explored in some detail in
Debugging regular expressions in the perldebguts manpage.
Perl contains internal support for reporting its own memory usage, but this is a fairly advanced concept that requires some understanding of how memory allocation works. See Debugging Perl memory usage in the perldebguts manpage for the details.
You did try the -w switch, didn't you?
the perldebtut manpage, the perldebguts manpage, the re manpage, the DB manpage, the Devel::DProf manpage, dprofpp, the Dumpvalue manpage, and the perlrun manpage.
You cannot get stack frame information or in any fashion debug functions that were not compiled by Perl, such as those from C or C++ extensions.
If you alter your @_ arguments in a subroutine (such as with shift
or pop
), the stack backtrace will not show the original values.
The debugger does not currently work in conjunction with the -W command-line switch, because it itself is not free of warnings.
If you're in a slow syscall (like wait
ing, accept
ing, or read
ing
from your keyboard or a socket) and haven't set up your own $SIG{INT}
handler, then you won't be able to CTRL-C your way back to the debugger,
because the debugger's own $SIG{INT}
handler doesn't understand that
it needs to raise an exception to longjmp(3)
out of slow syscalls.
perldebug - Perl debugging |