← All modules

One sub, Dump, that renders any structure — nested data, objects, attributes and all — as indented, optionally colored text a human can actually read.

Version

0.0.18 zef:tony-o

Depends

nothing outside the core

License

Artistic-2.0

Its own test suite

9 files, green

Checked

2026-08-28 against Raku++ 3.20.1 and Rakudo 2026.08

Where it lives

raku.land · source

Install it #

$ rakupp install Data::Dump

zef install Data::Dump writes the same store; either installer leaves the module usable by both engines.

What it is for #

say gives you a gist, .raku gives you something round-trippable, and neither is what you want at 2 a.m. with a nested structure that is not the shape you thought it was. Dump is the third answer: everything laid out vertically, one element per line, indentation carrying the nesting, each value labelled with its type:

use Data::Dump;

class Server { has $.host; has $.port; has @.tags }
my $s = Server.new(:host<db1>, :port(5432), :tags<primary ssd>);
say Dump($s, :!color, :skip-methods);
say Dump([1, 'two', 3.5], :!color);

Server :: (
  $!host => "db1".Str,
  $!port => 5432.Int,
  @!tags => [
    "primary".Str,
    "ssd".Str,
  ],

)
[
  1.Int,
  "two".Str,
  3.5.Rat,
]

Objects come apart into their attributes — real sigils, declaration order — and every leaf carries its type as a postfix, so "5432" and 5432 can never masquerade as each other in a dump. That type labelling is the feature you will miss in a plain say.

The adverbs that matter #

Two of the adverbs above are close to mandatory knowledge. :!color turns off the ANSI colors that Dump emits whenever Terminal::ANSIColor is installed — with it, dumps are painted for terminals and wrong for log files; the examples here disable it so the page shows bytes you can compare. :skip-methods keeps Dump from listing every method of the object's classes alongside the data — without it, a dump of one three-attribute object runs to dozens of lines of method signatures, which is occasionally what you want and usually is not.

The rest: :indent(4) widens the steps, :max-recursion(50) is the guard that keeps a self-referencing structure from dumping forever (Dump cuts off and says so rather than hanging), and :gist swaps the attribute view for each object's own .gist, for classes that already know how to present themselves.

One honest caution that belongs to hashes rather than to this module: a multi-key Hash dumps its pairs in hash order, which Raku does not promise between runs. Comparing dumps of hashes across runs — or between engines — is a coin flip by design; dump arrays and objects when the output has to be stable, as the examples on this page do.

Where the two engines differ #

Nothing on this page. Attribute discovery through the metaobject protocol, the sigil rendering, the type postfixes and the recursion handling answer identically under Raku++ and Rakudo — Dump leans hard on introspection (.^attributes, .^methods, .^mro), so this page is quietly a MOP conformance statement too.

What was run to put this page here #

1. Parse — every file of the distribution is parsed by Raku++ itself.
2. Install — rakupp install Data::Dump; it depends on nothing — Terminal::ANSIColor is picked up optionally at run time, via try require, when it happens to be installed.
3. Test — the distribution's own suite: 9 files, green.
4. Run — every example on this page, twice under each engine, as the site is built.

That optional-dependency trick — (try require ::('Terminal::ANSIColor')) falling back to a no-op colorizer — is worth stealing for your own modules, and its working here means dynamic require of a maybe-installed module behaves on both engines. Fourteen distributions depend on this one; an earlier engine campaign ("Six general faults behind Data::Dump") got its suite from red to 9/9 green, so this page is the receipt.