On ubuntu 14.04 building qtdeclarative from the dev branch is impossible
due to various errors in the openvg plugin.

Change-Id: I1f2d7d61c36428e64cf3248e550ac5c4f0baa4b0
Reviewed-by: Paolo Angelelli <paolo.angelelli@qt.io>

Task-number: QTBUG-58025
Change-Id: I0768ea834d5666f2831f24c3b2c71b7a4260d5d9
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

If you're an user it's a bit annoying since it may show a lot and you don't know what to do with it
If you're a developer it's a bit annoying since it's telling you to use a private class

If we had a "debug for developers only mode" we could improve the message to be a bit more fine tuned
and suggest that if they really want to be precise they can implement QQuickImageProviderWithOptions
but as far as i know we don't have it, so i think it's better to remove the message altogether.

Also mark the internal classes with the private export and add \internal to documentation of the internal class

Change-Id: I5d7fc8d8b333c9a12321ed8135433603251815fc
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Task-number: QTBUG-56881
Change-Id: I5e1614c2854b166f87b20804e13d7686e11f3a3a
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Let the destroy() method in QV4::String clean up the
unmanaged heap size instead of having a special hook in
the code that sweeps the GC heap.

Change-Id: I989ee99604f0cc67b896d3acc94e200dd5e56a60
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Change-Id: I49419e45ee2686e6e8646c49b839b4d77f2e05fd
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Fix the one place where we marked a Pointer without checking
whether it is 0 first.

Change-Id: I93ad0d9bc2d49594f3759ca429fd15615a16565f
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Since I'm working on this, it is nice to have some simple verification
that things are working. I don't expect this to provide full coverage,
but it should at least make sure that the basics are ok, and nicely
provides a place to put further tests if it ever somehow breaks.

Change-Id: If91154dee836cc514515bde122e48b271de22676
Reviewed-by: Lars Knoll <lars.knoll@qt.io>

They are moved from qtquickcontrols:tests/auto/controls in 3437fd56.

Task-number: QTBUG-58294
Change-Id: I281d4e505b2f699154dd4efb7a4601bdf3856295
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

Both QQuickPathLine and QQuickPathSvg inherit QQuickCurve class which
has “x” and “y” properties that return qreal type, but internally they
are stored as QQmlNullableValue<qreal>. At the same time, if any of them
is not specified explicitly, its getter returns 0.

QQuickPath processes QQuickPath%Type% objects and produces a
QPainterPath which later used by QQuickPathAnimation.
QQuickPathAnimation only created a QAbstractAnimationJob if
QQuickPath::hasEnd returned true, and hasEnd returned true only if both
“x” and “y” were specified explicitly.

All that in conjunction led to the situation when if you had either
- a PathLine with unspecified “x” or “y”; or
- a PathSvg
which was the last (or the only) path element in your Path,
PathAnimation would not start.
This patch removes hasEnd check, it should be safe to do because
QPainterPath is always valid anyway due to the fact QQuickCurve::x()
and QQuickCurve::y() return 0 if they have not been not explicitly set.

Task-number: QTBUG-57666
Change-Id: Id320aaeb5aff0964d6493b7b80d5d9a7d36acce8
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Useful for running a test tree against different qmljs or if qmljs is
outside PATH.

Change-Id: Ibaa24a15d32b21f9293db2c042fe3f1de3bb75eb
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Task-number: QTBUG-57288
Change-Id: Id106f09dc824fbd95780dc5b479e16802d183e57
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Change-Id: Ibedd0d0c23cf194ea02a229ab643450dbefd40aa
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

Causes a divide-by-zero. Low risk, but let's fix it anyway.

Coverity-Id: 161333
Change-Id: I7631abdca469a5dd9e2becf32cc9aa4f5cf515dc
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

If roleCount is 0, alloca(0) will give us a pointer. The loop will then not
initialize the returned pointer (due to the roleCount being 0), passing an
uninitialized changedRoles to emitDirectNotifies.

emitDirectNotifies doesn't access changedRoles unconditionally (via
another for loop), but this is probably better to check than not.

Coverity-Id: 172868
Done-with: John Brooks
Change-Id: I821c06221d2659d3310082b4e81442cc58b197f7
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

We don't exactly know when the animation speed is actually set and the
system clock may play tricks on us. Consider the test failed when the
wrong animation speed is witnessed 3 times in a row and successful if
the correct one is witnessed 3 times in a row. Also, make sure we don't
confuse lines from different hits of the timer.

Task-number: QTBUG-58186
Change-Id: Iaa2c35f723a92f32131e36084399b3d32accb7d0
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Change-Id: I7dfbd5f47893a3244c96989ad7eea7e93ea28603
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

Change-Id: Ief706dfce0c81351ac06b8791267107ad8165974
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Change-Id: If168e24e3e240182707602071125ac3538703fe7
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Change-Id: I0e60c54281cf6838845400fe1a7cf5f2b3a14d65
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Just src/imports/ to go...

Change-Id: Ib4484676e24655cb950b3c4bb6495ff0e7f9700f
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

While we're here, deduplicate the implementation (just forward to the
"right" setter), and change qWarning to qmlWarning to get context
information on the source of the setter.

Change-Id: I32acfee10dd34905bb2c472408e3abc6fa56a386
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

These are not performance-critical, so it's cleaner to do this.

Change-Id: I6a876636d50e2bfe25f9e6e882d9a98dc032101e
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

qmlInfo predated info-level messages in QtCore, and as such previously sent
warning-level messages despite the unfortunate naming. Now that we have
an actual qmlWarning function, and we have switched our code to use it,
we can change qmlInfo's behavior to better match the function naming.

This does have the impact that existing qmlInfo callers will
basically need a s/qmlInfo/qmlWarning/g to retain the same QDebug level in
user code, but I feel that this behavior change makes sense given the better
consistency with C++-side QDebug we attain.

[ChangeLog][QtQml][Important Behavior Changes] qmlInfo now reports messages with
a QtMsgType of QtInfoMsg instead of QtWarningMsg. To continue to send warnings,
callers should migrate to the newly-introduced qmlWarning function.

Change-Id: I16c88d94377b5956eb6921b64af7c84d1ca024f6
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Seems like I missed some hunks in 3057ec04

.

Change-Id: I009c838cae856f3fc2e08c7fbc2d3273fc40e221
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

The last remnants were removed from qtbase in 5.7 making this all dead code,
so match here too.

Change-Id: I10f3f1c614562f2a97ade7cdf5002065d6f79e07
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

We now have file & line number information pointing us to the location of the
faulty child. Also take the effort to split the message into two, so
it's clear which property is faulty (and which the value is
representing).

Change-Id: I8e515feeaf8fa1decf8aaf7adf0a02cce7aec0ea
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>

The attached isRowSet and isColumnSet check for values >= 0, and
row/column getters also enforce a minimum return value of 0, so it is
impossible for us to get a negative value for row/column if either of these is
set -- and we can simply call the getter and rely on it to return 0 in
the unset case.

Change-Id: Iec80e7d7cf3738cf0a81e90b027ffe41e0f57369
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>

Change-Id: Ie16508d22554c9b6dcd66adf6a59f32117edf1e3
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>

Change-Id: Iccc1ec0cf4408e25c04ca0be2562175e76d2d198
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Previously flicking was restricted to the orientation of the ListView.

[ChangeLog][QtQuick][ListView] Made it possible to enable horizontal
flicking in a vertical ListView, and vice versa. The only thing apps
must do is to specify the desired flick direction and the content
width (vertical ListView) or content height (horizontal ListView),
which is not calculated by ListView.

Change-Id: Ic370e57f5d18679940d48e7a2c20c200b2ef36d1
Task-number: QTBUG-52553
Task-number: QTBUG-56501
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

It's not part of this module, and promising something against a module
so distinct from qtdeclarative seems like a risky choice, even if the
class was a public and supported one -- checking qtwebkit's source shows
QQuickWebView is in a private header.

Change-Id: Ib250542e5a78083cbabe55217b9b9192d3a863fb
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Saves us from any future renaming or deprecation hassle.

Change-Id: Ied7bec12bb53d4e5a6661c36873a14fc7d68711b
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

qtdeclarative hasn't relied on qtwebkit for a long time (since
fbfb27a4 &
56d34a65

 as early as 2011!).

Change-Id: If02572617034bf2c3eecbf081b96b1ed0ad65b45
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

Picture the following application:
    QQmlApplication qAppEngine("main.qml");

With main.qml:
    ...
    AComponentInTheSameDirectory {
    }
    ...

This was failing, with the error:
    file:main.qml:13 AComponentInTheSameDirectory is not a type

Which is wrong, but also reveals the root cause in that the
original filename was not a fully resolved path.

Change-Id: Ifc5557cc43f4bb92fd121ea9f7a37f09b3b38a9b
Reviewed-by: David Faure <david.faure@kdab.com>

Longer term, we should QT_NO_FOREACH .qmake.conf, but a number of other
parts of the codebase still use foreach.

Change-Id: I105cdbe2a64ae40ffa53d3aaa8765b09195af841
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

These tests were blacklisted after QtSvg/4bd5d6ced07d2d0e643a13e7cebb228c521d2046.
in order to integrate qt5.git. Now, fix the tests to match the new
behavior, and remove the blacklisting.

Task-number: QTBUG-58082
Change-Id: I77abe995095ee52978c5957e49e1547b3af7708b
Reviewed-by: Lars Schmertmann <lars.schmertmann@governikus.de>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>

Conflicts:
	.qmake.conf

Change-Id: I9d87ed86e95b5901a86cc3aa65d7ac39b0b708c2

Change-Id: I178a9630353346f85dd67cfb552dde8ad4ca350e
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>

Item views and positioners all have now forceLayout(), so make the
Text element API consistent with them. The old doLayout(), which
sounds more like an internal helper method, is deprecated and marked
for removal in Qt 6.

[ChangeLog][QtQuick][Text] Deprecated doLayout() in favor of
forceLayout().

Change-Id: I051988fca13c4cd84904f7b268d51f6a96f28af3
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>