Distribution · time

Date::Calendar::Strftime

Works

The C library's strftime as a role — %Y-%m-%d, %j, %V, widths and flags — mixed into a core Date or into any of the author's Date::Calendar classes, so one format string works across calendars.

Version: 0.1.1 zef:jforget
Depends: Date::Names
License: Artistic-2.0
Its own test suite: 4 files, green
Checked: 2026-09-14 against Raku++ 3.28.0 and Rakudo 2026.08
Where it lives: raku.land · source

Install it #

$ rakupp install Date::Calendar::Strftime

zef install Date::Calendar::Strftime writes the same store; either installer leaves the module usable by both engines.

What it is for #

Raku's Date formats itself one way, 2026-09-14, and every other layout — 14/09/2026, day-of-year, ISO week — is a sprintf you write by hand. The C world settled this long ago with strftime and its % codes, which a generation of programmers already know. This distribution is that vocabulary as a Raku role: mix it into a Date and it gains a strftime method. It is the formatting layer under the author's family of calendar modules (Hebrew, Coptic, French Revolutionary and a dozen more), which is why it is a role rather than a function — every calendar composes it and supplies its own names.

On a core Date #

File

use Date::Calendar::Strftime;

my $d = Date.new('2026-09-14') does Date::Calendar::Strftime;
say $d.strftime('%Y-%m-%d is day %j, ISO week %V of %G, weekday %u');
say $d.strftime('%F | %e | %-d | %05j | %%');
say $d.strftime('%A %d %B');

Output

2026-09-14 is day 257, ISO week 38 of 2026, weekday 1
2026-09-14 | 14 | 14 | 00257 | %
%A 14 %B

The codes are the C ones: %F for the ISO date, %j the day of the year, %V and %G the ISO week and its year, %u Monday-is-1. The flags work too — %-d strips the padding, %05j sets a width and a fill. Anything the object cannot answer is left in the output as it was written, which is what happened to %A and %B on the last line.

The one thing to know #

A core Date has no names. %A and %B need day-name and month-name methods, which the Date::Calendar::* classes provide and Date does not — so on a plain Date those two codes come out as themselves, silently. If you want September spelled out on a core Date, Date::Names (which this distribution already depends on) answers mon(9), and the %A and %B belong to the calendar classes.

The codes that do work on a core Date are the ones it can answer through .can — and that is the mechanism this page tripped over. Under the Raku++ 3.28.0 release, .can('day-of-week') on a Date with a role mixed in came back empty, so %u and %V printed as literal %u and %V while everything else was right. The engine has been fixed since, and the first example now prints the same line under both.