CONTENTS

NAME

Data::Printer - colored & full-featured pretty print of Perl data structures and objects

SYNOPSIS

Want to see what's inside a variable in a complete, colored and human-friendly way?

use DDP;  # same as 'use Data::Printer'

p $some_var;
p $some_var, as => "This label will be printed too!";

# no need to use '\' before arrays or hashes!
p @array;
p %hash;

# printing anonymous array references:
p [ $one, $two, $three ]->@*;    # perl 5.24 or later!
p @{[ $one, $two, $three ]};     # same, older perls
&p( [ $one, $two, $three ] );    # same, older perls

# printing anonymous hash references:
p { foo => $foo, bar => $bar }->%*;   # perl 5.24 or later!
p %{{ foo => $foo, bar => $bar }};    # same, older perls
&p( { foo => $foo, bar => $bar } );   # same, older perls

The snippets above will print the contents of the chosen variables to STDERR on your terminal, with colors and a few extra features to help you debug your code.

If you wish to grab the output and handle it yourself, call np():

my $dump = np $var;

die "this is what happened: " . np %data;

The np() function is the same as p() but will return the string containing the dump. By default it has no colors, but you can change that easily too.

That's pretty much it :)

samples of Data::Printer output for several kinds of data and objects

Data::Printer is fully customizable, even on a per-module basis! Once you figure out your own preferences, create a .dataprinter configuration file for yourself (or one for each project) and Data::Printer will automatically use it!

FEATURES

Here's what Data::Printer offers Perl developers, out of the box:

DESCRIPTION

The ever-popular Data::Dumper is a fantastic tool, meant to stringify data structures in a way they are suitable for being "eval"'ed back in. The thing is, a lot of people keep using it (and similar ones, like Data::Dump) to print data structures and objects on screen for inspection and debugging, and while you can use those modules for that, it doesn't mean you should.

This is where Data::Printer comes in. It is meant to do one thing and one thing only:

format Perl variables and objects to be inspected by a human

If you want to serialize/store/restore Perl data structures, this module will NOT help you. Try Storable, Data::Dumper, JSON, or whatever. CPAN is full of such solutions!

Whenever you type use Data::Printer or use DDP, we export two functions to your namespace:

p()

This function pretty-prints the contents of whatever variable to STDERR (by default), and will use colors by default if your terminal supports it.

p @some_array;
p %some_hash;
p $scalar_or_ref;

Note that anonymous structures will only work if you postderef them:

p [$foo, $bar, $baz]->@*;

you may also deref it manually:

p %{{ foo => $foo }};

or prefix p() with &:

&p( [$foo, $bar, $baz] );    # & (note mandatory parenthesis)

You can pass custom options that will work only on that particular call:

p @var, as => "some label", colored => 0;
p %var, show_memsize => 1;

By default, p() prints to STDERR and returns the same variable being dumped. This lets you quickly wrap variables with p() without worrying about changing return values. It means that if you change this:

sub foo { my $x = shift + 13; $x }

to this:

sub foo { my $x = shift + 13; p($x) }

The function will still return $x after printing the contents. This form of handling data even allows method chaining, so if you want to inspect what's going on in the middle of this:

$object->foo->bar->baz;

You can just add DDP::p anywhere:

$object->foo->DDP::p->bar->baz; # what happens to $object after ->foo?

Check out the customization quick reference section below for all available options, including changing the return type, output target and a lot more.

np()

The np() function behaves exactly like p() except it always returns the string containing the dump (thus ignoring any setting regarding dump mode or destination), and contains no colors by default. In fact, the only way to force a colored np() is to pass colored => 1 as an argument to each call. It is meant to provide an easy way to fetch the dump and send it to some unsupported target, or appended to some other text (like part of a log message).

CUSTOMIZATION

There are 3 possible ways to customize Data::Printer:

1. [RECOMMENDED] Creating a .dataprinter file either on your home directory or your project's base directory, or both, or wherever you set the DATAPRINTERRC environment variable to.

2. Setting custom properties on module load. This will override any setting from your config file on the namespace (package/module) it was called:

use DDP max_depth => 2, deparse => 1;

3. Setting custom properties on the actual call to p() or np(). This overrides all other settings:

p $var, show_tainted => 1, indent => 2;

The .dataprinter configuration file

The most powerful way to customize Data::Printer is to have a .dataprinter file in your home directory or your project's root directory. The format is super simple and can be understood in the example below:

# global settings (note that only full line comments are accepted)
max_depth       = 1
theme           = Monokai
class.stringify = 0

# use quotes if you want spaces to be significant:
hash_separator  = " => "

# You can set rules that apply only to a specific
# caller module (in this case, MyApp::Some::Module):
[MyApp::Some::Module]
max_depth    = 2
class.expand = 0
escape_chars = nonlatin1

[MyApp::Other::Module]
multiline = 0
output    = /var/log/myapp/debug.data

# use 'quiet' to silence all output from p() and np()
# called from the specified package.
[MyApp::Yet::Another]
quiet = 1

Note that if you set custom properties as arguments to p() or np(), you should group suboptions as a hashref. So while the .dataprinter file has "class.expand = 0" and "class.inherited = none", the equivalent code is "class => { expand => 0, inherited => 'none' }".

live updating your .dataprinter without restarts

Data::Printer 1.1 introduces a new 'live_update' flag that can be set to a positive integer to enable live updates. When this mode is on, Data::Printer will check if the .dataprinter file has been updated and, if so, it will reload it. This way you can toggle features on and off and control output verbosity directly from your .dataprinter file without needing to change or restart running code.

Properties Quick Reference

Below are (almost) all available properties and their (hopefully sane) default values. See Data::Printer::Object for further information on each of them:

# scalar options
show_tainted      = 1
show_unicode      = 1
show_lvalue       = 1
print_escapes     = 0
scalar_quotes     = "
escape_chars      = none
string_max        = 4096
string_preserve   = begin
string_overflow   = '(...skipping __SKIPPED__ chars...)'
unicode_charnames = 0

# array options
array_max      = 100
array_preserve = begin
array_overflow = '(...skipping __SKIPPED__ items...)'
index          = 1

# hash options
hash_max       = 100
hash_preserve  = begin
hash_overflow  = '(...skipping __SKIPPED__ keys...)'
hash_separator = '   '
align_hash     = 1
sort_keys      = 1
quote_keys     = auto

# general options
name           = var
return_value   = pass
output         = stderr
use_prototypes = 1
indent         = 4
show_readonly  = 1
show_tied      = 1
show_dualvar   = lax
show_weak      = 1
show_refcount  = 0
show_memsize   = 0
memsize_unit   = auto
separator      = ,
end_separator  = 0
caller_info    = 0
caller_message = 'Printing in line __LINE__ of __FILENAME__'
max_depth      = 0
deparse        = 0
alias          = p
warnings       = 1

# colorization (see Colors & Themes below)
colored = auto
theme   = Material

# object output
class_method             = _data_printer
class.parents            = 1
class.linear_isa         = auto
class.universal          = 1
class.expand             = 1
class.stringify          = 1
class.show_reftype       = 0
class.show_overloads     = 1
class.show_methods       = all
class.sort_methods       = 1
class.inherited          = none
class.format_inheritance = string
class.parent_filters     = 1
class.internals          = 1

Settings' shortcuts

Colors & Themes

Data::Printer lets you set custom colors for pretty much every part of the content being printed. For example, if you want numbers to be shown in bright green, just put colors.number = #00ff00 on your configuration file.

See Data::Printer::Theme for the full list of labels, ways to represent and customize colors, and even how to group them in your own custom theme.

The colorization is set by the colored property. It can be set to 0 (never colorize), 1 (always colorize) or 'auto' (the default), which will colorize p() only when there is no ANSI_COLORS_DISABLED environment variable, the output is going to the terminal (STDOUT or STDERR) and your terminal actually supports colors.

Profiles

You may bundle your settings and filters into a profile module. It works like a configuration file but gives you the power and flexibility to use Perl code to find out what to print and how to print. It also lets you use CPAN to store your preferred settings and install them into your projects just like a regular dependency.

use DDP profile => 'ProfileName';

See Data::Printer::Profile for all the ways to load a profile, a list of available profiles and how to make one yourself.

Filters

Data::Printer works by passing your variable to a different set of filters, depending on whether it's a scalar, a hash, an array, an object, etc. It comes bundled with filters for all native data types (always enabled, but overwritable), including a generic object filter that pretty-prints regular and Moo(se) objects and is even aware of Role::Tiny.

Data::Printer also comes with filter bundles that can be quickly activated to make it easier to debug binary data and many popular CPAN modules that handle date and time, databases (yes, even DBIx::Class), message digests like MD5 and SHA1, and JSON and Web content like HTTP requests and responses.

So much so we recommend everyone to activate all bundled filters by putting the following line on your .dataprinter file:

filters = ContentType, DateTime, DB, Digest, Web

Creating your custom filters is very easy, and you're encouraged to upload them to CPAN. There are many options available under the Data::Printer::Filter::* namespace. Check Data::Printer::Filter for more information!

Making your classes DDP-aware (without adding any dependencies!)

The default object filter will first check if the class implements a sub called '_data_printer()' (or whatever you set the "class_method" option to in your settings). If so, Data::Printer will use it to get the string to print instead of making a regular class dump.

This means you could have the following in one of your classes:

sub _data_printer {
    my ($self, $ddp) = @_;
    return 'Hey, no peeking! But foo contains ' . $self->foo;
}

Notice that you can do this without adding Data::Printer as a dependency to your project! Just write your sub and it will be called with the object to be printed and a $ddp object ready for you. See Data::Printer::Object for how to use it to pretty-print your data.

Finally, if your object implements string overload or provides a method called "to_string", "as_string" or "stringify", Data::Printer will use it. To disable this behaviour, set class.stringify = 0 on your .dataprinter file, or call p() with class => { stringify => 0 }.

Loading a filter for that particular class will of course override these settings.

CAVEATS

You can't pass more than one variable at a time.

p $foo, $bar;       # wrong
p $foo; p $bar;     # right

You can't use it in variable declarations (it will most likely not do what you want):

p my @array = qw(a b c d);          # wrong
my @array = qw(a b c d); p @array;  # right

If you pass a nonexistent key/index to DDP using prototypes, they will trigger autovivification:

use DDP;
my %foo;
p $foo{bar}; # undef, but will create the 'bar' key (with undef)

my @x;
p $x[5]; # undef, but will initialize the array with 5 elements (all undef)

Slices (both array and hash) must be coerced into actual arrays (or hashes) to properly shown. So if you want to print a slice, instead of doing something like this:

p @somevar[1..10]; # WRONG! DON'T DO THIS!

try one of those:

my @x = @somevar[1..10]; p @x;   # works!
p [ @somevar[1..0] ]->@*;        # also works!
p @{[@somevar[1..0]]};           # this works too!!

Finally, as mentioned before, you cannot pass anonymous references on the default mode of use_prototypes = 1:

p { foo => 1 };       # wrong!
p %{{ foo => 1 }};    # right
p { foo => 1 }->%*;   # right on perl 5.24+
&p( { foo => 1 } );   # right, but requires the parenthesis
sub pp { p @_ };      # wrapping it also lets you use anonymous data.

use DDP use_prototypes => 0;
p { foo => 1 };   # works, but now p(@foo) will fail, you must always pass a ref,
                  # e.g. p(\@foo)

BACKWARDS INCOMPATIBLE CHANGES

While we make a genuine effort not to break anything on new releases, sometimes we do. To make things easier for people migrating their code, we have aggregated here a list of all incompatible changes since ever:

Any undocumented change was probably unintended. If you bump into one, please file an issue on Github!

TIPS & TRICKS

Using p() in some/all of your loaded modules

(contributed by Matt S. Trout (mst))

While debugging your software, you may want to use Data::Printer in some or all loaded modules and not bother having to load it in each and every one of them. To do this, in any module loaded by myapp.pl, simply write:

::p @myvar;  # note the '::' in front of p()

Then call your program like:

perl -MDDP myapp.pl

This also has the advantage that if you leave one p() call in by accident, it will trigger a compile-time failure without the -M, making it easier to spot :)

If you really want to have p() imported into your loaded modules, use the next tip instead.

Adding p() to all your loaded modules

(contributed by Árpád Szász)

If you wish to automatically add Data::Printer's p() function to every loaded module in you app, you can do something like this to your main program:

BEGIN {
    {
        no strict 'refs';
        require Data::Printer;
        my $alias = 'p';
        foreach my $package ( keys %main:: ) {
            if ( $package =~ m/::$/ ) {
                *{ $package . $alias } = \&Data::Printer::p;
            }
        }
    }
}

WARNING This will override all locally defined subroutines/methods that are named p, if they exist, in every loaded module. If you already have a subroutine named 'p()', be sure to change $alias to something custom.

If you rather avoid namespace manipulation altogether, use the previous tip instead.

Using Data::Printer from the Perl debugger

(contributed by Árpád Szász and Marcel Grünauer (hanekomu))

With DB::Pluggable, you can easily set the perl debugger to use Data::Printer to print variable information, replacing the debugger's standard p() function. All you have to do is add these lines to your .perldb file:

use DB::Pluggable;
DB::Pluggable->run_with_config( \'[DataPrinter]' );  # note the '\'

Then call the perl debugger as you normally would:

perl -d myapp.pl

Now Data::Printer's p() command will be used instead of the debugger's!

See perldebug for more information on how to use the perl debugger, and DB::Pluggable for extra functionality and other plugins.

If you can't or don't want to use DB::Pluggable, or simply want to keep the debugger's p() function and add an extended version using Data::Printer (let's call it px() for instance), you can add these lines to your .perldb file instead:

$DB::alias{px} = 's/px/DB::px/';
sub px {
    my $expr = shift;
    require Data::Printer;
    print Data::Printer::p($expr);
}

Now, inside the Perl debugger, you can pass as reference to px expressions to be dumped using Data::Printer.

Using Data::Printer in a perl shell (REPL)

Some people really enjoy using a REPL shell to quickly try Perl code. One of the most popular ones out there are Reply and Devel::REPL. If you use them, now you can also see its output with Data::Printer!

Just install Reply::Plugin::DataPrinter and add a line with [DataPrinter] to your .replyrc file. That's it! Next time you run the 'reply' REPL, Data::Printer will be used to dump variables!

Just install Devel::REPL::Plugin::DataPrinter and add the following line to your re.pl configuration file (usually ".re.pl/repl.rc" in your home dir):

load_plugin('DataPrinter');

The next time you run re.pl, it should dump all your REPL using Data::Printer!

Easily rendering Data::Printer's output as HTML

To turn Data::Printer's output into HTML, you can do something like:

use HTML::FromANSI;
use Data::Printer;

my $html_output = ansi2html( np($object, colored => 1) );

In the example above, the $html_output variable contains the HTML escaped output of p($object), so you can print it for later inspection or render it (if it's a web app).

Using Data::Printer with Template Toolkit

(contributed by Stephen Thirlwall (sdt))

If you use Template Toolkit and want to dump your variables using Data::Printer, install the Template::Plugin::DataPrinter module and load it in your template:

[% USE DataPrinter %]

The provided methods match those of Template::Plugin::Dumper:

ansi-colored dump of the data structure in "myvar":
[% DataPrinter.dump( myvar ) %]

html-formatted, colored dump of the same data structure:
[% DataPrinter.dump_html( myvar ) %]

The module allows several customization options, even letting you load it as a complete drop-in replacement for Template::Plugin::Dumper so you don't even have to change your previous templates!

Migrating from Data::Dumper to Data::Printer

If you are porting your code to use Data::Printer instead of Data::Dumper, you could replace:

use Data::Dumper;

with something like:

use Data::Printer;
sub Dumper { np @_, colored => 1 }

this sub will accept multiple variables just like Data::Dumper.

Unified interface for Data::Printer and other debug formatters

(contributed by Kevin McGrath (catlgrep))

If you want a really unified approach to easily flip between debugging outputs, use Any::Renderer and its plugins, like Any::Renderer::Data::Printer.

Printing stack traces with arguments expanded using Data::Printer

(contributed by Sergey Aleynikov (randir))

There are times where viewing the current state of a variable is not enough, and you want/need to see a full stack trace of a function call.

The Devel::PrettyTrace module uses Data::Printer to provide you just that. It exports a bt() function that pretty-prints detailed information on each function in your stack, making it easier to spot any issues!

Troubleshooting apps in real time without changing a single line of your code

(contributed by Marcel Grünauer (hanekomu))

dip is a dynamic instrumentation framework for troubleshooting Perl programs, similar to DTrace. In a nutshell, dip lets you create probes for certain conditions in your application that, once met, will perform a specific action. Since it uses Aspect-oriented programming, it's very lightweight and you only pay for what you use.

dip can be very useful since it allows you to debug your software without changing a single line of your original code. And Data::Printer comes bundled with it, so you can use the p() function to view your data structures too!

# Print a stack trace every time the name is changed,
# except when reading from the database.
dip -e 'before { print longmess(np $_->{args}[1], colored => 1)
if $_->{args}[1] } call "MyObj::name" & !cflow("MyObj::read")' myapp.pl

You can check dip's own documentation for more information and options.

Sample output for color fine-tuning

(contributed by Yanick Champoux (yanick))

The "examples/try_me.pl" file included in this distribution has a sample dump with a complex data structure to let you quickly test color schemes.

VERSIONING AND UPDATES

As of 1.0.0 this module complies with Major.Minor.Revision versioning scheme (SemVer), meaning backwards incompatible changes will trigger a new major number, new features without any breaking changes trigger a new minor number, and simple patches trigger a revision number.

CONTRIBUTORS

Many thanks to everyone who helped design and develop this module with patches, bug reports, wishlists, comments and tests. They are (alphabetically):

Adam Rosenstein, Alexandr Ciornii (chorny), Alexander Hartmaier (abraxxa), Allan Whiteford, Anatoly (Snelius30), Andre Klärner, Andreas König (andk), Andy Bach, Anthony DeRobertis, Árpád Szász, Athanasios Douitsis (aduitsis), Baldur Kristinsson, Benct Philip Jonsson (bpj), brian d foy, Chad Granum (exodist), Chris Prather (perigrin), Curtis Poe (Ovid), David D Lowe (Flimm), David E. Condon (hhg7), David Golden (xdg), David Precious (bigpresh), David Raab, David E. Wheeler (theory), Damien Krotkine (dams), Denis Howe, dirk, Dotan Dimet, Eden Cardim (edenc), Elliot Shank (elliotjs), Elvin Aslanov, Eugen Konkov (KES777), Fernando Corrêa (SmokeMachine), Fitz Elliott, Florian Schlichting (fschlich), Frew Schmidt (frew), GianniGi, Graham Knop (haarg), Graham Todd, Gregory J. Oschwald, grr, Håkon Hægland, Iaroslav O. Kosmina (darviarush), Ivan Bessarabov (bessarabv), J Mash, James E. Keenan (jkeenan), Jarrod Funnell (Timbus), Jay Allen (jayallen), Jay Hannah (jhannah), jcop, Jesse Luehrs (doy), Joel Berger (jberger), John S. Anderson (genehack), Karen Etheridge (ether), Kartik Thakore (kthakore), Kevin Dawson (bowtie), Kevin McGrath (catlgrep), Kip Hampton (ubu), Londran, Marcel Grünauer (hanekomu), Marco Masetti (grubert65), Mark Fowler (Trelane), Martin J. Evans, Matthias Muth, Matt S. Trout (mst), Maxim Vuets, Michael Conrad, Mike Doherty (doherty), Nicolas R (atoomic), Nigel Metheringham (nigelm), Nuba Princigalli (nuba), Olaf Alders (oalders), Paul Evans (LeoNerd), Pedro Melo (melo), Philippe Bruhat (BooK), Przemysław Wesołek (jest), Rebecca Turner (iarna), Renato Cron (renatoCRON), Ricardo Signes (rjbs), Rob Hoelz (hoelzro), Salve J. Nilsen (sjn), sawyer, Sebastian Willing (Sewi), Sébastien Feugère (smonff), Sergey Aleynikov (randir), Slaven Rezić, Stanislaw Pusep (syp), Stephen Thirlwall (sdt), sugyan, Tai Paul, Tatsuhiko Miyagawa (miyagawa), Thomas Sibley (tsibley), Tim Heaney (oylenshpeegul), Toby Inkster (tobyink), Torsten Raudssus (Getty), Tokuhiro Matsuno (tokuhirom), trapd00r, Tsai Chung-Kuan, Veesh Goldman (rabbiveesh), vividsnow, Wesley Dal`Col (blabos), y, Yanick Champoux (yanick).

If I missed your name, please drop me a line!

LICENSE AND COPYRIGHT

Copyright (C) 2011-2024 Breno G. de Oliveira

This program is free software; you can redistribute it and/or modify it under the terms of either: the GNU General Public License as published by the Free Software Foundation; or the Artistic License.

See http://dev.perl.org/licenses/ for more information.

DISCLAIMER OF WARRANTY

BECAUSE THIS SOFTWARE IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE SOFTWARE, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE SOFTWARE "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE SOFTWARE IS WITH YOU. SHOULD THE SOFTWARE PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR, OR CORRECTION.

IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE SOFTWARE AS PERMITTED BY THE ABOVE LICENCE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE SOFTWARE (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE SOFTWARE TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.