Changes file for Date-Holidays
$Id: Changes 1780 2007-03-05 19:56:47Z jonasbn $
0.12 (Mon Mar 5 20:52:05 CET 2007)
- Fixed a problem with pod.t
- Increased the Perl::Critic severity to 5
- Exchanged Readonly for good old constants, one less dependecy, should make
the failing tests go away
0.11 (Sun Feb 25 12:35:34 CET 2007)
- Added missing requirement to Build.PL Readonly
Based on tests reports from CPAN testers
Thanks david at cantrell.org.uk
I thought Readonly was core, but according to Module::CoreList this is not the
case, so this has been corrected now
- Updated to more contemporary versions of:
t/pod-coverage.t
t/pod.t
These should be there for the quality, but will not be run as a part of the
normal test run, unless the TEST_POD environment variable is set.
This seem to be the defacto way of doing things. So they have been removed
from MANIFEST.SKIP so they are now a part of the distribution. At the same
time I removed some lines from another distribution
0.10 (Thu Feb 22 20:42:01 CET 2007)
- Updated and corrected POD in Date::Holidays
- Updated and corrected POD in Date::Holidays::Adapter
- Added $VERSION to all exception classes (Date::Holidays::Exception::*), I and
who just file an RT ticket on the same issue for another distribution
oh the sweet nemesis :)
- Updated POD in all exception classes (Date::Holidays::Exception::*)
- Updated POD in Date::Holidays::Adapter::AU
- Updated POD in Date::Holidays::Adapter::GB
- Updated POD in Date::Holidays::Adapter::PT
- Updated POD in Date::Holidays::Adapter::NO
- Updated POD in Date::Holidays::Adapter::DK
- Updated POD in Date::Holidays::Adapter::FR
- Updated POD in Date::Holidays::Adapter::DE
- Updated POD in Date::Holidays::Adapter::JP
- Added new adapter for Date::Holidays::Adapter::ES
lib/Date/Holidays/Adapter/ES.pm
- Added use of exceptions instead of carp in Date::Holidays::Adapter, this
however produces issues with adhering to Perl::Critics recommendation on
explicit returns. Be aware that return with in Error's try-catch block returns
from a sub, so you have to have a return outside the block - and Perl::Critic
is happy and I am happy
Updated version to 0.02 for Date::Holidays::Adapter
0.09 (Wed Feb 21 23:22:04 CET 2007)
- Added no_indexing of t/ directory to Build.PL
- Updated README with pod2text appending of Date::Holidays POD
- Code cleaned a bit, much work still to be done
- Wrote some better DIAGNOSTICS and added 3 more Exceptions
lib/Date/Holidays/Exception/InvalidCountryCodep.pm
lib/Date/Holidays/Exception/NoCountrySpecified.pm
lib/Date/Holidays/Exception/UnsupportedMethod.pm
- Added a few tests adapters for some of the tests dating before the
refactoring (all tests now pass):
t/lib/Date/Holidays/Adapter/NOPOLY.t
t/lib/Date/Holidays/Adapter/OOP.t
- Renamed _loader in Date::Holidays to _fetch and _load, see also the similar
methods in Date::Holidays::Adapter
_loader.t obsolete and removed
- Introduced use of Error (Exceptions) for better diagnostics:
Date/Holidays/Exception/AdapterInitialization.pm
Date/Holidays/Exception/AdapterLoad.pm
Date/Holidays/Exception/SuperAdapterLoad.pm
- Refactored the whole thing to a variation of an object adapter pattern,
introduced:
lib/Date/Holidays/Adapter.pm
lib/Date/Holidays/Adapter/
lib/Date/Holidays/Adapter/AU.pm
lib/Date/Holidays/Adapter/DE.pm
lib/Date/Holidays/Adapter/DK.pm
lib/Date/Holidays/Adapter/FR.pm
lib/Date/Holidays/Adapter/GB.pm
lib/Date/Holidays/Adapter/JP.pm
lib/Date/Holidays/Adapter/NO.pm
lib/Date/Holidays/Adapter/PT.pm
tests are located in t/Adapter:
_fetch.t
_load.t
new.t
- Set severity to 3 for Test::Perl::Critic test
ran code through perltidy to remove hard tabs
- Set severity to 4 for Test::Perl::Critic test
added use warnings statement
- Changed the constructor to no longer be able to initialize a shallow object.
If you want to make use of Date::Holidays ability to check all countries for
a given holiday please, use it using full namespace. POD updated
correspondingly.
- Addressed issue with method calling problems, OOP vs. Procedural vs. the weird
mix I personally had boiled up, thanks to Florian Merges for reporting this.
This has resulted in a minor cleanup and again I can see that the current
architecture is not optimal, I am planning a major rewrite, to eliminate
the problems of handling new classes introduced in the Date::Holidays::*
namespace.
I am not an authority of any kind, I just attempt to get things to play
along nicely.
New files introduced:
t/NOPOLY.t
t/OOP.t
t/SUPERED.t
t/ABSTRACTED.t
t/PRODUCERAL.t
t/new.t
and the test classes (in t/lib/Date/Holidays):
ABSTRACTED.pm
NOPOLY.pm
OOP.pm
PROCUDERAL.pm
SUPERED.pm
- Updated MANIFEST.SKIP with more contemporary version
- Added t/critic.t, Test::Perl::Critic test, currently at severity 5.
In addition to this I changed the 'return undef's scattered all over the code
to simple return statements.
- Added t/kwalitee.t, Test::Kwalitee test, diabled check for symlinks since
it reacted on symlinked .releaserc in directory, which however is NOT in the
distribution even though it thinks so
Added the following as a result of the Kwalitee test:
t/pod.t
t/pod-coverage.t
- Added use of Module::Load to Build.PL and to own _loader routine in
Date::Holidays. I also added:
t/_loader.t to isolate the actual test
0.08 (Wed Sep 6 16:55:04 CEST 2006)
- Added RT request #21089, helper scripts not part of distribution, but
mentioned in the autogenerated Makefile.PL
Renamed bin/ directory to scripts/
0.07 (Released Wed Aug 2 22:14:45 CEST 2006)
- Removed bin/ directory from distribution (it is development purpose only
anyway)
- Removed t/pod.t (and prerequisites in Build.PL) this test can be
performed by Module::Build
- Removed t/pod-coverage.t (and prerequisites in Build.PL) this test can be
performed by Module::Build
- Updated TODO with new point
- Changed way Makefile.PL is generated from 'passthrough' to 'traditional', I
kind of like the way it is done in Workflow, so I adapted this
- Fixed a bug in program flow concerning Date::Holidays::DE
- Fixed a test holding a wrong number of tests to be skipped
- Updated POD, added Date::Holidays::CN to SEE ALSO section (no work done to
implement this at this time
- Updated POD with TEST COVERAGE section, planning next release to be a
maintenance release focused on test coverage
Also in an attempt to address:
http://www.nntp.perl.org/group/perl.cpan.testers/276637
0.06 (Released Sun Dec 17 2005)
- Moved build requirements to special section in Build.PL
- Added AU and NZ to bin/test_date-holidays.t
- Updated MANIFEST.SKIP with:
- Komodo project file
- Module::Build parameters mentioned in Module::Signature
- Signed module using Module::Signature, added:
- t/0-signature.t
- SIGNATURE
- Parameter 'state' also added for holidays method, have sent patch to David
Dick author of Date::Holidays::AU requesting a more flexible API.
- Added some more POD on developing in the Date::Holidays::* namespace
- Date::Holidays::AU have been updated to 0.03, meaning that the exception
implemented in 0.05 could be removed, it was however changed to accomodate
the state parameter implemented in Date::Holidays::AU.
- Exchanged manually built Makefile.PL for a Makefile.PL maintained by
Module::Build
0.05 (Released Fri Dec 9 2005)
- Addressing failing test
http://www.nntp.perl.org/group/perl.cpan.testers/262252
So we have added yet another exception, this time for Date::Holidays::AU.
The problem is that its method is_holiday, needs an additional parameter
indicating the state, the holidays method in the same module defaults
to VIC, so for know we are using this as our default state, this will need
readdressing.
0.04 (Released 22. October 2005)
- Changes to unit-tests, it was assumed that some of the Date::Holidays
modules where installed, this is not always the case
This was pointed out to me by shild on use.perl.org:
http://use.perl.org/comments.pl?sid=28993&cid=43889
0.03 (Released 13. October 2005)
- Small refactoring added new sub _check_countries
- Moved portuguese exception, Date::Holiday::PT has changed name to
Date::Holidays::PT, but it has turned OOP.
- Implemented new parameter in is_holiday (countries). This method returns a
list of country codes having the holiday specified as a holiday for the
respective country (suggested by cog).
So it has to be used in conjunction with the data parameters: year, month and
day
if given a subset of countries only these countries are tested and a hashref
is returned pointing where the countries codes are the keys and the values
indicate true or false for the date specified.
The values are:
- undef if the country has no module or the data could not be obtained
- a name of the holiday if a holiday is present
- an empty string if the a module was located but the day is not a holiday
- Updated INSTALL
- Removed Exporter from prerequisites
- Replaced use of ExtUtils::Makemake with Module::Build
- Added realnames to Changes file (this file)
- Added suggestion on improvements from cog (Jose Castro) to TODO
0.02 Released 31st. of May 2004
- Indented code in POD after tip from RJBS (Ricardo Signes)
- 'jp' left out of the test suite for now
- Date::Holidays::NO added to test suite, courtesy of MRAMBERG (Marcus Ramberg)
- Added experimental subs using DateTime objects as suggested by BORUP
(Christian Borup) (SEE: TODO)
0.01 Released 22nd. of May 2004
- Intial release