David Faure 46ec170d2b Speed up QTimeZone::isTimeZoneIdAvailable by a factor 43 ... Creating and sorting a list of bytearrays just to check if one entry is present, is really overkill. By adding a new virtual method isTimeZoneIdAvailable() in the backend classes, we can do this much more efficiently. Implemented for Utc and Tz backends, the others fall back to the slow way. The new benchmark shows, in release mode: Before: 43 msecs per iteration (total: 86, iterations: 2) After: 1.1 msecs per iteration (total: 73, iterations: 64) Change-Id: Ic0d79a41d74e2ce6aa088fa7986c41d33902c36b Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>		2018-04-23 06:48:12 +00:00
..
codecs	Updated license headers	2016-01-21 18:55:18 +00:00
io	Use dependencies instead of CONFIG+=ordered	2017-12-02 11:09:15 +00:00
json	Updated license headers	2016-01-21 18:55:18 +00:00
kernel	Make a benchmark out of tst_QObjectPerformance::emitToManyReceivers	2018-03-07 19:43:52 +00:00
mimetypes	Updated license headers	2016-01-21 18:55:18 +00:00
plugin	Updated license headers	2016-01-21 18:55:18 +00:00
thread	Replace Q_DECL_OVERRIDE with override where possible	2017-09-19 11:53:42 +00:00
tools	Speed up QTimeZone::isTimeZoneIdAvailable by a factor 43	2018-04-23 06:48:12 +00:00
corelib.pro	test: Moving tst_QMimeDatabase::inheritsPerformance() test	2012-10-31 19:56:42 +01:00