2001-12-26 21:37 rbowen
* t/: 09greg.t (1.2), 10subtract.t (1.5): Spacing changes (courtesy
of perltidy) in the tests, so that I can better read what is going
on here. No functional change.
2001-12-26 21:27 rbowen
* Makefile.PL (1.19): Add a vim syntax highlighting line to the
Change log file so that it looks pretty.
2001-12-26 21:25 rbowen
* lib/Date/ICal.pm (1.66): No longer need to do strange workaround
to subtract dates resulting in negative duration. It Just Works
(tm).
2001-12-26 21:07 rbowen
* lib/Date/ICal/Duration.pm (1.10), t/11duration.t (1.5): Permit
the creation of Duration objects with negative attributes. Note
that all attributes have to have the same sign. Mixed-sign
attributes will result in the entire duration being negative. This
could stand to be documented more clearly. Also, all accessors
have been modified to explicitly return undef if that attribute has
zero value. I'm not sure if this is desired, but it's the way that
it is documented, and what the tests expect. Tests added to verify
all this stuff.
2001-12-26 20:21 rbowen
* Makefile.PL (1.18): Add the Duration documentation to the README
on make dist
2001-12-26 20:15 rbowen
* lib/Date/ICal/Duration.pm (1.9): Code folds. No functional
change.
2001-12-26 15:55 rbowen
* lib/Date/ICal.pm (1.65), t/10subtract.t (1.4): A small
work-around because D::I::Duration does not correctly handle
negative arguments yet. See diff 1.7 -> 1.8 of
Date::ICal::Duration. When D::I::Duration correctly handles
negative arguments, this patch should be mostly reversed. Added
tests for subtracting dates to get negative duration.
2001-12-26 15:50 rbowen
* lib/Date/ICal/Duration.pm (1.8): This is step one. Components now
return signed. And the 'sign' argument is being honored now, rather
than being discarded. Need to add tests for this stuff. Still does
not correctly handle negative arguments.
2001-12-20 16:04 rbowen
* lib/Date/ICal.pm (1.64), t/10subtract.t (1.3): Added the ability
to subtract durations from a date, either as a Date::ICal::Duration
object, or as a ICal duration string. Tests to match.
2001-12-20 10:20 rbowen
* lib/Date/ICal/Duration.pm (1.7): Documentation: The parameter is
'ical' not 'string' to create an object with an ical duration
string.
2001-12-19 07:52 rbowen
* Makefile.PL (1.17), README (1.4): As suggested, generatel README
at release time, rather than having it in CVS.
2001-12-18 21:24 rbowen
* README (1.3), lib/Date/ICal.pm (1.63): Bring README up to date
with the POD. Minor POD fix in ICal.pm
2001-12-18 21:21 rbowen
* lib/Date/ICal.pm (1.62): Patch from Yitzchak. Adds day_of_week
method. Other minor POD fixes.
2001-12-16 19:21 rbowen
* MANIFEST (1.9, Date-ICal-1_61): Added t/11duration.t to MANIFEST
2001-12-16 19:06 rbowen
* Makefile.PL (1.16, Date-ICal-1_61): Added warning back in about
possibly changing API.
2001-12-16 17:15 rbowen
* Makefile.PL (1.15), lib/Date/ICal.pm (1.61, Date-ICal-1_61),
t/10subtract.t (1.2, Date-ICal-1_61): sub subtract in Date::ICal
now uses Date::ICal::Duration objects, rather than doing its own
math. Added tests to test this functionality, and some TODO notes
for what I would like it to do in the near future.
2001-12-16 17:14 rbowen
* lib/Date/ICal/Duration.pm (1.6, Date-ICal-1_61): Added sub
as_days to return the number of days without weeks %'ed out. And
some code folding foo.
2001-12-16 16:54 coral
* lib/Date/ICal/Duration.pm (1.5): Extracting clarity from
Duration.pm's synopsis is HARD.
2001-12-16 16:52 rbowen
* t/11duration.t (1.4, Date-ICal-1_61): Removes some redundant 'my'
in tests.
2001-12-16 16:30 coral
* lib/Date/ICal/Duration.pm (1.4): TODO: Figure out what we're
going to do with sign data.
2001-12-16 16:12 srl
* lib/Date/ICal/Duration.pm (1.3): Added code for accesors and
as_elements.
2001-12-16 16:11 srl
* t/11duration.t (1.3): Added tests for as_elements and for
accessors. Added a new sign() read-only accessor.
2001-12-16 15:47 srl
* lib/Date/ICal/Duration.pm (1.2): Fleshed out some more of the
methods; this is largely drawn from Net::ICal::Duration, which I
expect to be replaced by this module.
2001-12-16 15:45 srl
* t/11duration.t (1.2): Added more tests for Duration
functionality.
2001-12-16 15:14 rbowen
* Changes (1.3), Makefile.PL (1.14): Take Changes out of cvs.
Autogenerate it on make dist.
2001-12-16 15:11 rbowen
* MANIFEST (1.8): Add Changes to the manifest so that it gets
shipped with the dist.
2001-12-14 21:24 rbowen
* Changes (1.2): Updated change log to include the entire project,
using cvs2cl
2001-12-14 00:29 rbowen
* INTERNALS (1.2, Date-ICal-1_61): Remove inaccurate and/or
outdates notes about the internals. Once the internals stabilize,
I'll write something more useful here.
2001-12-14 00:27 rbowen
* Changes (1.1): Starter Changes file. Jesse said something about
generating a Changes file from CVS, which is what this is. Perhaps
his is a little more pleasant looking
2001-12-14 00:01 rbowen
* lib/Date/ICal.pm (1.60): Primarily documentation updates.
2001-12-14 00:01 rbowen
* README (1.2, Date-ICal-1_61): Bring README up to date with the
module.
2001-12-13 22:20 rbowen
* MANIFEST (1.7), lib/Date/ICal.pm (1.59), t/06add.t (1.6,
Date-ICal-1_61): Patch submitted by Yitzchak to make add() DWIMier,
to permit more reliable subtraction of dates via the add() method,
and tests to make sure that it is all doing what it should. Also
provides the eom_mode parameter to make adding months do what you
expect ALL THE TIME. All tests still pass. Yitzchak++
2001-12-13 17:33 srl
* lib/Date/ICal/Duration.pm (1.1), t/11duration.t (1.1): Added
Date::ICal::Duration trivial skeleton and basic tests. API comments
welcome; code coming soon.
2001-12-12 21:58 rbowen
* lib/Date/ICal.pm (1.58), t/10subtract.t (1.1): Possible
implementation of subtraction of dates. Comments welcomed.
2001-12-12 19:25 rbowen
* MANIFEST (1.6), Makefile.PL (1.13): ICal.pm has (finally!) moved
into lib/Date where it should have been to begin with.
2001-12-11 10:12 rbowen
* lib/Date/ICal.pm (1.57): I've removed some warnings, because we
are basically warning when people use documented default behavior.
This is very irritating. I also need to update the documentation so
that it is more clear on this point, but for the moment, this
scratches my immediate itch. --DrBacchus
2001-12-08 22:30 rbowen
* Makefile.PL (1.12): Memoize no longer a prereq
2001-11-30 22:25 rbowen
* lib/Date/ICal.pm (1.56): This is, I believe, the intent of
Yitzchak's first two patches. There is no content in this diff,
just style things. Ran perltidy on it to make the whole file
conform to agreed-upon style standards. And standardized the usage
of code folding characters.
2001-11-30 12:02 rbowen
* Makefile.PL (1.11): Need Test::More to support skip().
2001-11-27 21:00 rbowen
* lib/Date/ICal.pm (1.55), t/06add.t (1.5): Able to add n years to
a date via the add method. Tests to match.
2001-11-26 09:38 rbowen
* MANIFEST (1.5): Most of the tests are not making it into the
distribution, because the MANIFEST never got updated.
2001-11-24 13:57 rbowen
* lib/Date/ICal.pm (1.54): Oops. I reversed the order of the
argument list when I added this function back in, thereby breaking
all code that was calling it.
2001-11-24 11:25 rbowen
* lib/Date/ICal.pm (1.53): Since _seconds_from_offset returns a
number, not a string, we only need the sign if it is negative.
Resolves some test failures that I was seeing in t/08offset.t for
negative offsets.
2001-11-23 22:42 rbowen
* lib/Date/ICal.pm (1.52): Resolves one of the test failures in
offset/add - when add crossed a day boundary by virtue of a
difference in seconds, it was not compenting in the day value, and
could end up with negative times.
2001-11-23 22:15 rbowen
* Makefile.PL (1.10): There is no longer a module called
Pod::Tests. It is now called Test::Inline.
2001-11-23 22:11 rbowen
* lib/Date/ICal.pm (1.51): Added back in days_this_year method
using new greg2jd method.
2001-11-23 22:03 rbowen
* Makefile.PL (1.9): There are no embedded tests at this time, so
this is generating bogus test failures. Uncomment if you add inline
tests back in.
2001-11-23 21:54 rbowen
* lib/Date/ICal.pm (1.50), t/09greg.t (1.1, Date-ICal-1_61): This
is Yitzchak's patch to give us much more efficient gregorian <->
julian conversions, and to remove strange anomolous problems in the
17th and 18 centuries. Note that we lose a few internal methods
here, at least one of which I'll be putting back in a minute.
2001-11-22 05:56 srl
* lib/Date/ICal.pm (1.49), t/08offset.t (1.8, Date-ICal-1_61): This
version incorporates a patch by Yitzchak Scott-Thoennes to adjust
the offset() API. It no longer takes integer seconds as a
parameter, because there's no programmatic way to tell the
difference between, say, +3600 (UTC+1 in seconds) and +3600
(UTC+36, if you interpret that as an HHMM value).
I also refactored things a bit, creating an _offset_from_seconds
method to match _offset_to_seconds; this should eliminate some
duplication. There's also new POD to clear up some confusion about
new(offset => foo) used together with offset().
2001-11-22 04:22 srl
* lib/Date/ICal.pm (1.48): API-consistency patch from Yitzchak
Scott-Thoennes <sthoenna@efn.org>; Makes the ical() method take a
hash of parameters, not a hashref, so that ical() is like the other
methods. This crept in around 1.44, and it shouldn't have. My
mistake.
2001-11-22 04:02 srl
* lib/Date/ICal.pm (1.47): Fixed some 5.6isms; patch contributed by
Yitzchak Scott-Thoennes <sthoenna@efn.org>.
2001-11-15 08:25 srl
* lib/Date/ICal.pm (1.46), t/08offset.t (1.7): Minor patches to
tests; another piece of optimization from Mike Heins.
2001-11-15 00:32 srl
* benchmark.pl (1.1, Date-ICal-1_61), lib/Date/ICal.pm (1.45),
t/08offset.t (1.6): Added benchmark.pl to help developers in
optimizing the module. Also modified new() to warn more clearly if
the $TZ environment variable isn't set, and not to utterly fail
tests if $TZ isn't there.
2001-11-15 00:11 srl
* lib/Date/ICal.pm (1.44), t/08offset.t (1.5): Further patches from
Mike Heins, plus some documentation from me: - added
localtime argument to ical() for output in localtime. Added
documentation about the localtime argument. Note that
$ENV{$TZ} is now relevant to some of the module's behavior. -
removed a memoize() that wasn't providing significant speed
improvements. - minor optimization of _calc_local_offset
2001-11-14 23:20 srl
* lib/Date/ICal.pm (1.43): Committed another small patch by Mike
Heins, which precalculates the values returned by the months()
function so that the module is faster at runtime.
2001-11-14 23:11 srl
* lib/Date/ICal.pm (1.42): Another patch from Mike Heins
(mheins@minivend.com); an optimizing cheat for leapyears; uses a
precalculated table of values instead of always calculating
leapyear values. I edited Mike's patch slightly so that @leapcheat
isn't a package global.
2001-11-14 22:58 srl
* lib/Date/ICal.pm (1.41): Incorporated part of a patch by Mike
Heins (mheins@minivend.com); an optimization. Internal storage of
julian times is now in $self->{julian} and $self->{julsec}, instead
of using an array. This gives us slightly better speed. Also, made
some of the UTC behaviors slightly more consistent.
2001-10-16 06:33 srl
* lib/Date/ICal.pm (1.40), t/01sanity.t (1.6, Date-ICal-1_61),
t/02normalize.t (1.3, Date-ICal-1_61), t/03components.t (1.2,
Date-ICal-1_61), t/04epoch.t (1.2, Date-ICal-1_61), t/06add.t
(1.4), t/08offset.t (1.4): Further fixes to the offset() method.
This code isn't as well-tested as I'd like it to be, but it seems
to do the right thing for all the tests that are there. I had to
revise many of the tests, because the API semantics have changed.
Times must now be explicitly specified with a Z in order to be
handled as UTC.
2001-10-14 21:58 rbowen
* MANIFEST (1.4): Rename Readme to README, per request of coral
2001-10-14 21:56 rbowen
* README (1.1), Readme (1.3): coral says README is prefered to
Readme, wrt CPAN. I had no opinions either way, so here it is.
2001-10-09 22:58 srl
* lib/Date/ICal.pm (1.39), t/08offset.t (1.3): Added some tests,
reorganized some code to prepare spaces for offset/timezone-aware
output. Added at least one test that's known to fail for purposes
of knowing when we succeed. :)
2001-10-09 00:28 srl
* lib/Date/ICal.pm (1.38), t/08offset.t (1.2): Started working on
code to properly handle times with offsets from GMT. added a new
_calc_local_offset method to figure out what the current machine's
UTC offset is. We need tests for this that will work in any
timezone; patches welcome.
2001-09-30 09:19 lotr
* lib/Date/ICal.pm (1.37): * Oops, forgot some bits when I added
month to add() * use overload for compare
2001-09-29 07:01 lotr
* lib/Date/ICal.pm (1.36): Add the ability to add months to a date.
Needed for Net::ICal::Recurrence
2001-09-26 11:26 lotr
* lib/Date/ICal.pm (1.35), t/06add.t (1.3): * fix off-by-one error
in months() and add tests for that
2001-09-11 23:26 rbowen
* lib/Date/ICal.pm (1.34): There's no particular reason to have
Date::ICal be 5.6 dependant.
2001-08-25 08:20 rbowen
* lib/Date/ICal.pm (1.33), t/06add.t (1.2): Fixed bug reported by
Chris Jones. In sub add, I was checking one attribute and using
another. Added tests for this bug, and for adding durations by
attribute.
2001-08-09 23:27 srl
* lib/Date/ICal.pm (1.32), t/08offset.t (1.1): Started adding
timezone support by making an offset() method and an offset
property. This still needs to be wired into the new() method and
the output methods, but we have to resolve some interface details
first.
2001-08-06 22:41 rbowen
* lib/Date/ICal.pm (1.31): Test::More gets angry if there are no
tests.
2001-08-06 22:30 rbowen
* lib/Date/ICal.pm (1.30): Moved the inline tests into t/ for the
sake of making the module more readable. Please don't let this
discorage you from writing inline tests.
2001-08-06 22:25 rbowen
* t/: 04epoch.t (1.1), 05ical.t (1.1, Date-ICal-1_61), 06add.t
(1.1), 07compare.t (1.1, Date-ICal-1_61): Moved a lot of tests out
of the module into .t files.
2001-08-06 15:32 rbowen
* Makefile.PL (1.8), lib/Date/ICal.pm (1.29): Creating an object
without args was calling gmtime( $args{epoch} ). Fixed and added
tests. Also added Time::HiRes to PREREQ list.
2001-08-06 14:45 rbowen
* lib/Date/ICal.pm (1.28): sub epoch was referencing another sub
that has gone away. Fixed, and added tests.
2001-08-02 00:38 srl
* lib/Date/ICal.pm (1.27): Adjusted the add() method to return a
copy of $self instead of the return value of $self->jd(). This was
important to making the Net::ICal tests pass, but it's also the
Right Way, I think.
2001-08-01 23:47 rbowen
* lib/Date/ICal.pm (1.26): Handle negative durations correctly.
2001-07-31 22:19 rbowen
* lib/Date/ICal.pm (1.25): Two main changes here. 1) Split the
internal date/time representation into date, time integers, so that
we don't have any more roundoff error. 2) Memoized the parsetime
and parsedate methods, so that we're not doing that three times
every time we want three components, which we were doing.
2001-07-31 22:17 rbowen
* Makefile.PL (1.7): Added Memoize, Storable to prereq list, so
that I can memoize some functions.
2001-07-31 07:43 rbowen
* lib/Date/ICal.pm (1.24): Create a new D::I object using another
D::I object (or subclass) as the initial value. This allows for
immediate conversion of a date from one calendar format to another,
which was one of my main goals to start with. :-)
2001-07-27 12:24 rbowen
* lib/Date/ICal.pm (1.23), t/03components.t (1.1): Creating objects
from components was not working. Fixed.
2001-07-26 14:36 rbowen
* MANIFEST (1.3), lib/Date/ICal.pm (1.22): Removed unused internal
methods. Updated ToDo list.
2001-07-26 14:14 rbowen
* INTERNALS (1.1): Discussion of the internals, and the limitations
thereof.
2001-07-26 11:49 rbowen
* lib/Date/ICal.pm (1.21): To compare two dates, just need to
compare their mjd representation.
2001-07-26 11:21 rbowen
* Readme (1.2): Generated from POD
2001-07-26 11:19 rbowen
* MANIFEST (1.2): Added Readme
2001-07-25 21:32 srl
* Makefile.PL (1.6): Added a requirement for Time::Local to the
Makefile.PL.
2001-07-25 21:04 srl
* Makefile.PL (1.5): Did a release tag on Date-ICal-1_20, since
DrBacchus forgot to; then added a line for Date::Leapyear in the
PREREQ section of Makefile.PL.
2001-07-25 19:31 rbowen
* t/: 01sanity.t (1.5), 02normalize.t (1.2) (utags:
Date-ICal-1_20): Updated tests for changes to ICal. Removed some
that are no longer relevant. "fixed" some due to round-off error.
This all works, for some slightly relaxed definition of "works".
After releasing, we need to be a little more rigorous.
2001-07-25 19:29 rbowen
* lib/Date/ICal.pm (1.20, Date-ICal-1_20): Complete rewrite.
Internals all changed. More discussion on the mailing list.
2001-07-23 23:21 srl
* lib/Date/ICal.pm (1.19): Added a minor fix to Date::ICal to allow
accessors to be set to 0.
2001-07-23 22:28 rbowen
* Readme (1.1, Date-ICal-1_20): Readme generated from POD
2001-07-23 21:54 srl
* Makefile.PL (1.4, Date-ICal-1_20), lib/Date/ICal.pm (1.18)
(utags: date-ical-1_18_01): Minor spelling corrections. Also, added
a note about the future Julian rework to TODO, and a note about
failing tests in Makefile.PL.
2001-07-08 01:17 srl
* lib/Date/ICal.pm (1.17): Added some arithmetic-verification tests
to Date::ICal--- they all fail currently. The problems are related
to addition when the parameter is a duration string--- unit
rollover seems to be breaking, and second values seem to be
particularly off.
2001-07-07 22:24 rbowen
* lib/Date/ICal.pm (1.16): Should not attempt to calculate an epoch
time for dates outside of the epoch. Added tests also.
2001-07-07 18:20 rbowen
* Makefile.PL (1.3): Pod::Tests and Test::More are prereq's for
this module.
2001-07-07 15:44 rbowen
* lib/Date/ICal.pm (1.15): ICal is now the only authoritative
internal format, as far as I can tell. There are still some parts
of this that confuse me, but epoch is being recalculated from ICal
when it is requested. We need to make sure that if epoch is called
for dates outside of the epoch, that we warn. That is not being
done yet. But all tests currently pass. At least for me. There are
probably some timezone dependencies still, and I'd appreciate some
tests by folks in other timezones. However, I did not comment out
any tests in order to make it pass the tests. In fact, I added
several.
2001-07-07 15:25 rbowen
* t/01sanity.t (1.4, date-ical-1_18_01): It helps if you are
actually testing the right object.
2001-07-07 00:10 srl
* lib/Date/ICal.pm (1.7), t/01sanity.t (1.3): Adjusted some tests
in Date::ICal to use == in tests instead of eq.
Tried to adjust some mistakes in Date::ICal's 01sanity.t so that
those tests will pass, but no luck. Dates before the epoch are
still failing.
2001-07-06 13:05 rbowen
* lib/Date/ICal.pm (1.6): Calling accessor methods rather than hash
keys. Certain attributes get calculated the first time you call the
method, rather than being calculated up front, in case you never
actually need it. All Pod::Tests now succeed.
2001-07-05 22:37 srl
* lib/Date/ICal.pm (1.5): Added a bunch of tests for some error
conditions in Date::ICal.
In particular, we need to check out the results of the accessors,
and we need to decide whether they're supposed to return numbers
with leading zeroes.
2001-07-03 23:33 rbowen
* t/00load.t (1.2, Date-ICal-1_61, Date-ICal-1_20,
date-ical-1_18_01): Moved to Test::More for added readability
2001-07-03 23:33 rbowen
* t/01sanity.t (1.2): Moved the test to Test::More. Uncommented
several failing tests, so that I have some motivation to fix the
problem. Ran perltidy. I know, it's a fetish ...
2001-07-02 22:55 srl
* lib/Date/ICal.pm (1.4): Committed a bunch of tests for Date::ICal
that are failing; I don't have a solution, but I wanted to share
this in hopes someone can help.
2001-06-29 14:07 lotr
* lib/Date/ICal.pm (1.3): _epoch_from_ical: fix to change month
and year to be in the format timegm expects in the DATE-TIME
case, just like the DATE case already had
2001-06-27 22:22 srl
* Makefile.PL (1.2), lib/Date/ICal.pm (1.2): Implemented the
compare() function (with tests) for Date::ICal.
Added a section in the Makefile.PL to compile the embedded tests
into a t/ file.
Corrected some of the podtests that were already in D::I, because
pod2test didn't like them.
2001-06-27 20:53 rbowen
* MANIFEST (1.1), Makefile.PL (1.1), lib/Date/ICal.pm (1.1),
t/00load.t (1.1), t/01sanity.t (1.1), t/02normalize.t (1.1):
Initial revision
2001-06-27 20:53 rbowen
* MANIFEST (1.1.1.1, Date-ICal-1_20, date-ical-1_18_01),
Makefile.PL (1.1.1.1), lib/Date/ICal.pm (1.1.1.1), t/00load.t
(1.1.1.1), t/01sanity.t (1.1.1.1), t/02normalize.t (1.1.1.1,
date-ical-1_18_01) (utags: Date-ICal-1_14): Moving Date::ICal to
sourceforge reefknot repository. I hope I don't regret this.
# vim: filetype=changelog