This reverts commit ab1e507574. That was
supposed to be a minor behavior change, but ends up having visible
effects such as QtXmlPatterns xs:dateTime type now reporting sub-second
fractions. So we're reverting in 5.10 and re-applying in 5.11.
Change-Id: I741e49459c9a688c1c329d6cbd521cd4a0b2aa84
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Actually check that there's a T where ISO 8601 wants it (instead of
just skipping over whatever's there), with something after it; move
some declarations later; add some comments; and use the QStringRef API
more cleanly (so that it's easier to see what's going on). Simplify a
loop condition to avoid the need for a post-loop fix-up.
This incidentally prevents an assertion failure (which brought the
mess to my attention) parsing a short string as an ISO date-time; if
there's a T with nothing after it, we won't try to read at index -1 in
the following text. (The actual fail seen had a Z where the T should
have been, with nothing after it.)
Add tests for invalid ISOdate cases that triggered the assertion.
Task-number: QTBUG-66076
Change-Id: Ided9adf62a56d98f144bdf91b40f918e22bd82cd
Reviewed-by: Israel Lins Albuquerque <israelins85@yahoo.com.br>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
(cherry picked from commit a9c111ed8c)
QHeaderView is doing a complete rebuild of the sections when the layout
changed because everything could have happened. But since layoutChanged
is also called during e.g. sorting, the old data must be restored when
possible.
Task-number: QTBUG-65478
Change-Id: I088d4d843cad362b97df6dc5e0dcb9819b13547f
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
The backwards iteration was done under the assumption that the only
valid modification of the winEventNotifierList in a slot connected to
activated() would be the removal of the notifier itself. This is wrong.
Instead, iterate forwards, like before 85403d0a, and check the index
against the current list size in every iteration. This ensures that we
do not run out of bounds while the list is modified.
Also, retry the activation loop if the list was modified by a slot
connected to activated(). This ensures that all notifiers with signaled
handles are activated.
Task-number: QTBUG-65940
Change-Id: I25f305463b9234f391abc51fe0628d02f49b6931
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
This way, it's lossless.
Change-Id: I5e421e32396d44e4b39efffd150b744e40fff3a1
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The idx member of EventWithNotifier is unused.
Change-Id: I0f5aacaaad4b4e82c57ff7bb020586944014f139
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Setting speed to 0 means the current frame will continue to be shown,
the finished signal is not emitted, and state remains QMovie::Running.
Task-number: QTBUG-65758
Change-Id: I681d902e3211c5899b21043e5177b7c73d5d3fb5
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Passing -1 to waitForReadyRead() may cause it to wait for some time
but the data retrieved may be enough for processing. So if 0 is passed
from read, indicating that there is potentially more to come, then
it will do a waitForReadyRead() then for more data to come.
Change-Id: I75f270d1f124ecc12b18512cc20fb11f7a88f02e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The Embedded Android build (Boot to Qt Android injection) is defined by
having both Q_OS_ANDROID and Q_OS_ANDROID_EMBEDDED flags defined,
as well as having Qt config android-embedded.
This commit enables the possibility to build embedded Android builds.
(i.e. Qt build for Android baselayer only, without JNI)
Change-Id: I8406e959fdf1c8d9efebbbe53f1a391fa25f336a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
For Q_OS_WIN, a path is only truly absolute if it includes a drive
letter; merely starting with a slash is not enough. (We can't support
UNC paths, so don't even try: qmake runs various commands in the
source directory using CMD.exe, which doesn't support UNC as PWD.)
This requires, when resolving a path relative to a root, transcribing
the root's drive to such not-quite-absolute paths.
Changed QMakeGlobals, $$absolute_path() and $$relative_path() to now
use IoUtils::resolvePath() rather than delegating to QDir's absolute
path method, since that doesn't correctly recognize the need for a
drive letter (and qmake did run into problems with some paths, from
splitPathList and a failing test, as a result).
Moved existing ioUtils tests for handling of relative / absolute paths
out into separate functions and expanded significantly. Fixed some
existing tests to use an absolute path where one is needed; added two
tests involving driveless (but rooted) paths; and fixed the test init
to set a value for QT_HOST_DATA/src property (the lack of which lead
to an assertion failure with this fix).
Task-number: QTBUG-50839
Change-Id: I2bfc13c1bfbe1ae09997274622ea55cb3de31b43
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The operator_eqeq(data13) test expected the local-time epoch and UTC
epoch to agree precisely if the localTimeType set by the test's
constructor says local time is UTC; however, when the local zone is
*sometimes* ahead of (or behind) UTC, due to DST, localTimeType is
duly set to indicate that, which doesn't preclude the zone agreeing
with UTC at the epoch. This indeed happens for Europe/London, which
agrees on the epoch but was ahead a few months later. So we can't
determine what outcome to expect based solely on localTimeType,
although we can be sure of a match when local time is UTC. So skip
this test when local time isn't UTC (and document what's going on a
bit better).
Task-number: QTBUG-65435
Change-Id: Id9b8aa0402f2a2b410e0234f6eca4ab0d1010bc4
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Remove usage of outdated header.LGPL21 and replace those with proper
one (header.LGPL in src, header.GPL-EXCEPT in tests)
Change-Id: Ia4d1c0d84b77f09787fe7c30670747a1fe2aff29
Reviewed-by: Liang Qi <liang.qi@qt.io>
We have the ability to blacklist tests for CI runs now.
Task-number: QTBUG-35109
Change-Id: I8590e83faba764dce2d52e8c62e2e2c63f7bf219
Reviewed-by: Liang Qi <liang.qi@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The code used to fall back to anonymous login independently for username
and password; however, it should only use a fall-back password if the
username is missing or (case-insensitive) "anonymous". When a
non-anonymous username is given without password, we should simply skip
he PASS message to FTP.
If the FTP server requests a password, in the latter case, QFtp will
signal authenticationRequired; in all cases, if the server rejects the
given credentials, QFtp signals authenticationFailed. Either way, the
client code can then query the user for credentials as usual.
Task-number: QTBUG-25033
Change-Id: I2a4a3b2725819ab19c8a7e4baa431af539edcd8d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
After we stopped sanitizing the fallback font list (with change
6ca48a847a), we now need to make
sure it is ordered so that the fonts that support the writing
system in question are always tested first, otherwise we can end up
loading a lot of fonts that will never be used.
Task-number: QTBUG-65605
Change-Id: Id2a65bbff3e64e6d6e6b4f72500778ee3e811e84
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Introduce a special hash modeled on the one used for QFileSystemWatcher
on Windows.
Task-number: QTBUG-31103
Task-number: QTBUG-64147
Change-Id: I69ebabe841716e4957ae3fb04fa5c43d233a3552
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Old header.LGPL21 header was used at some files. Replase those with
new header.LGPL one
Remove old header.LGPL21
Task-number: QTBUG-57147
Change-Id: I650e39024ed4876bba27e954c7d61fdb025b46ef
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
When we request fallback fonts, we cannot discriminate the fonts
based on the writing system support. This is especially important
since common script is now merged with other scripts, meaning that
a common script character will always go through the fallback
mechanism when not supported by the main font. When drawing
for instance a string of Devanagari characters on macOS, we would
get a list of 33 fallback fonts, but almost all of them would be
the default Devanagari font, since none of the other fallbacks
would support that script. Meaning that we would just check the
same font over and over, which makes no sense. The fallback list
has been retrieved specifically for the given script, so we do
not need to consider that when fetching the fonts.
For most of the common set, we will not have noticed the bug,
because at least one of the writing system-specific fallbacks will
have had support for latin characters as well. But when trying to
mix emojis and some non-common script, we would get a box in
place of the emoji, which had been adopted to the main script and
would only be looked for in the fonts supporting this.
Note that this exposed an issue with the QRawFont test on some
systems. When the sample text contained a space, it would
be possible to get a fallback font for this character, since
we now effectively support fallbacks. This is not the correct
behavior, but it is unrelated to this fix, and it was not what
the QRawFont::unsupportedWritingSystem() test was written to
check. I have therefore removed the space from the sample text
to make the test pass, and will make a separate task of fixing
the issue of merging fonts for whitespace characters.
[ChangeLog][QtGui][Text] Fixed a bug where mixing different
writing systems with emojis could lead to missing glyphs.
Task-number: QTBUG-61882
Change-Id: I00f6043bb01af1f2277723ccf643034aebf3e18f
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
sqlite will reuse the index for bound parameters when the named
placeholder is duplicated so we only need to call bind one time for each
placeholder. Therefore we need to have just one instance of each value
when doing the bind.
Task-number: QTBUG-65150
Change-Id: I75c4bcc4563e43c180a59a7a4cbb770dbe994642
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Remove black-listing of test; the problem was due to misconfigured
time-zone on the CI system - it was on some zone that presently
coincides with UTC, not actually on UTC as supposed.
This reverts commit 597b96b8fa.
Change-Id: I72ad1dfa38532362c05aef33cd874f7f79879a41
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
Setting the same brush on the same group and role should
not detach nor alter the result of QPalette::isCopyOf().
Task-number: QTBUG-56743
Change-Id: Ic2d0dd757d703b01e8c5d835a8c124b3317653f4
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Files to test binary compatibility for 5.10 added. It looks like
tst_bic (that runs the BC tests) is implemented so that it will start
testing new content when e.g in case of 5.10 branching to 5.10.1 has
happened.
Change-Id: Ic97db0339cc2942c088962dda495eb42b96fe153
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Unlike higher scopes (like scope 4, admin-local, which the last commit
used), scopes 1 and 2 require a scope in order to bind, even if some
operating systems are lenient. So test that we are able to bind to them
and do bind properly.
Change-Id: Ifb5969bf206e4cd7b14efffd14fba153eab965b9
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Binding without an interface and expecting the OS to select something is
not supported in all OSes. On FreeBSD, I keep getting EADDRNOTAVAIL. So
modify our test to only join, leave and send to multicast groups with an
interface selection.
With this, all tests either pass or are skipped for me on Linux,
FreeBSD, and macOS. On Windows, this revealed an inconsistency in
behavior, which this commit adds a workaround for.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb6815456494d2
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
tst_QUdpSocket::broadcasting and tst_QUdpSocket::pendingDatagramSize
fail on the new Ubuntu 16.04 clean template.
Task-number: QTBUG-65440
Change-Id: I0e973b9c90b7c5827406bac8138370b61992a115
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
This autotest fails on the new Ubuntu 16.04 template with UTC timezone
in the system settings.
Task-number: QTBUG-65435
Change-Id: I397f01ab3fed354a4eeec8b05415226a75fce5a1
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
The FreeBSD kernel treats them specially, just like link-local (that's
probably why it calls them "interface-local" instead of "node-local").
So instead let's use a random address, which will avoid multiple
tst_qudpsocket, when run on the same network at the same time,
receiving each other's datagrams. It could happen, considering this test
has an 800-second timeout limit.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb592a3166547e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
This is not an official feature of the networking stacks and does not
work portably across operating systems. So just stop trying to do that.
This was failing reliably (not flaky!) with IPv6 on FreeBSD and
Windows. For IPv4, Windows apparently accepts 239.255.0.0/16 but not
other addresses, so remove IPv4 too.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb682c2839e95d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
On Windows, QFileInfo.isWritable() was returning true in situations
where the file would only be writable with elevated privileges. Using
AccessCheck instead of GetEffectiveRightsFromAcl to get the correct
results.
Done-with: Friedemann Kleint <Friedemann.Kleint@qt.io>
Done-with: Edward Welbourne <edward.welbourne@qt.io>
Task-number: QTBUG-30148
Change-Id: I7a3468ac069bf782ca312078e3a84107b6cd468c
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
old style connects have protection against null sender and null receiver,
but new style only had against null sender.
Change-Id: Ie555ac078412918e60c3b60830fe1f3abfb7f5af
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
QHeaderViewPrivate::_q_layoutChanged() was not called when used in a
QTreeView because it was explicitly disconnected in setModel().
The disconnect was added sometime prio to Qt 4.3, but there the signal
was connected to the doItemsLayout() slot. This was correct since
QTreeView::doItemsLayout() is calling header->doItemsLayout().
In Qt 4.3.0 _q_layoutChanged() was introduced and the disconnect was
adjusted. But since _q_layoutChanged() is doing much more than
doItemsLayout() (e.g. restoring hidden sections), functionality was
lost. The problem was already observed for Qt 4.6 (QTBUG-18196) but
only partially fixed.
Task-number: QTBUG-41124
Task-number: QTBUG-54610
Change-Id: Id13a9930d0163812e12a0287016bab9c3aa02068
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
If a signal is emitted more than once in a multithreaded application the
QSignalEventGenerator::execute function asserts in the check for a valid
signal index. It happens after abandoning the state and all the
connections are disconnected. If we have pending signal to be
processed the QObject::sender() won't be able to resolve the sender
object.
Task-number: QTBUG-61463
Change-Id: I9d4b7266c6dddc9ff2e7453b05a6989876ccb332
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
