The address of $this passed to drectly called internal constructor in execute_data->return_value.
Internal constructors should use ZEND_CTOR_MAKE_NULL() macro (insted of previous ZEND_NULL(EG(This))) to do the work.
This patch doesn't fix the problem for indirectly called constructors. e.g. parant::__construct().
* PHP-5.4:
Fixed defective cloning in ext/intl classes
NEWS for commit 72c807a
Allow Spoofchecker to be registered on ICU 49.1
Announce on NEWS change in 1ce572c
To better support IntlCalendar, added this function:
string IntlDateFormatter::formatObject(IntlCalendar|DateTime $obj [,
array|int|string $format = null [, string $locale = null).
$format is either of the constants IntlDateFormatter::FULL, etc., in
which case this format applies to both the date and the time, an array
in the form array($dateFormat, $timeFormat), or a string with the
SimpleDateFormat pattern.
This uses both the Calendar type and the timezone of the passed object
to configure the formatter (a GregorianCalendar is forced for
DateTime).
Some stuff was moved around and slighlt modified to allow for more code
reuse.
By Gustavo André dos Santos Lopes (4) and others
via Felipe Pena (2) and Xinchen Hui (2)
* PHP-5.4:
Remove unused codes
based on microsoft's description,the direct convert from FILETIME struct to __int64 is unsafe.
merge 5.3 entries
restore NEWS
Fix ext/intl build on ICU < 4.8
Optimization in ext/intl/msgformat
Fixed tests in ext/intl
Changed XFAILed collator_get_sort_key.phpt
By Gustavo André dos Santos Lopes (4) and others
via Felipe Pena (1) and Xinchen Hui (1)
* PHP-5.3:
Remove unused codes
based on microsoft's description,the direct convert from FILETIME struct to __int64 is unsafe.
Fix ext/intl build on ICU < 4.8
Optimization in ext/intl/msgformat
Fixed tests in ext/intl
Changed XFAILed collator_get_sort_key.phpt
The following changes were made:
* The IntlDateFormatter constructor now accepts the usual values
for its $timezone argument. This includes timezone identifiers,
IntlTimeZone objects, DateTimeZone objects and NULL. An empty
string is not accepted. An invalid time zone is no longer accepted
(it used to use UTC in this case).
* When NULL is passed to IntlDateFormatter, the time zone specified in
date.timezone is used instead of the ICU default.
* The IntlDateFormatter $calendar argument now accepts also an
IntlCalendar. In this case, IntlDateFormatter::getCalendar() will
return false.
* The time zone passed to the IntlDateFormatter is ignored if it is
NULL and if the calendar passed is an IntlCalendar object -- in this
case, the IntlCalendar time zone will be used instead. Otherwise,
the time zone specified in the $timezone argument is used instead.
* Added IntlDateFormatter::getCalendarObject(), which always returns
the IntlCalendar object that backs the DateFormat, even if a
constant was passed to the constructor, i.e., if an IntlCalendar
was not passed to the constructor.
* Added IntlDateFormatter::setTimeZone(). It accepts the usual values
for time zone arguments. If NULL is passed, the time zone of the
IntlDateFormatter WILL be overridden with the default time zone,
even if an IntlCalendar object was passed to the constructor.
* Added IntlDateFormatter::getTimeZone(), which returns the time zone
that's associated with the DateFormat.
* Depreacated IntlDateFormatter::setTimeZoneId() and made it an alias
for IntlDateFormatter::setTimeZone(), as the new ::setTimeZone()
also accepts plain identifiers, besides other types.
IntlDateFormatter::getTimeZoneId() is not deprecated however.
* IntlDateFormatter::setCalendar() with a constant passed should now
work correctly. This requires saving the requested locale to the
constructor.
* Centralized the hacks required to avoid compilation disasters on
Windows due to some headers being included inside and outside of
extern "C" blocks.
I don't think the current ICU API allows this bug to be completely fixed.
Right now, the code cannot control the time zone used in date/time formats
that appear inside complex subformats. See the comment inside
umsg_set_timezone().
MessageFormatter::parse and MessageFormat::format (and their static
equivalents) now don't throw away better than second precision in the
arguments.
It's already bad enough that in MessageFormatter and IntlDateFormatter we
use seconds since epoch instead of milliseconds since epoch, deviating
from the ICU date representations. But we don't need to throw away extra
precision when parsing dates; we can keep the seconds since epoch
convention and return non integer doubles with only a small BC impact.
Note that we already could return doubles from MessageFormatter::parse if
the date was sufficiently in the past or in the future.
The check does not work reliably across ICU versions when named arguments
are added to the mix. For instance, for recent versions of ICU like 49,
a pattern like "{foo,number} {foo}", has 0 returned from
umsg_format_arg_count(), but for ICU 4.0, this returns 2.
