| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
QDialog for instance adds a setter for the modal property already
exposed by QWidget.
Object.defineProperty requires configurable set to true in order to
add that setter at a later stage.
Adds some reusable autotest logic with a soft dependency on QJSEngine to
test some of the C++/JS integration aspects.
Task-number: QTBUG-46548
Change-Id: Ibd49274f7d334c068c4006fb09417abf911c24e9
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When wrapping a registered object the code generated a new
ID for an already known object. That new ID wasn't stored
but returned to the client including the objects information.
That resulted in a new created object on client side but
the remote object (on the server) was not accessible.
This patch fixes the issue by just returning the known ID
of a known object. Because the client already has a local
representation of that object it does not have to unwrap
the object description.
Change-Id: I31964823c84c84fd7ebce4386865c18fb5518be7
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
|
| |
Qt copyrights are now in The Qt Company, so we could update the source
code headers accordingly. In the same go we should also fix the links to
point to qt.io.
Change-Id: Iebed549451f58a9fbdd86adf5d0340412d7766d7
Reviewed-by: Jani Heikkinen <jani.heikkinen@theqtcompany.com>
|
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This merge required extensive conflict handling because
the bug fix in 5.4 to properly wrap and forward QObjects
referenced by published objects' properties clashed with
some feature additions in dev, namely the client separation
logic.
All unit test pass for me locally now again.
Conflicts:
.qmake.conf
src/webchannel/qmetaobjectpublisher.cpp
tests/auto/qml/tst_webchannel.qml
Change-Id: If3d00e13b265c6ab9fb2c38023014f97f8e7779b
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Implemented a remove method in SignalHandler that allows
us to remove and disconnect an object from SignalHandler
w/o decrementing the connection counter until it hits
zero or deleting the object
That same functionality was used to remove an object from
internal lists when receiving a destroyed signal from an
object. In case of deregistering an object we haven't
received a destoryed signal but simulated reception of that
signal and so that code was not called in that case.
Change-Id: Ie20cf628a2de028375f5d29f913682e25ebf8d44
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Similar to the support for factory-methods, we must wrap objects in
properties to make them accessible to clients. This patch adds the
required code for that. Besides support for simple properties that
reference an object, this patch also adds support for list properties
that contain objects.
The client-side unwrap of properties is delayed until all objects
are initialized, as a property might reference another registered
object.
Change-Id: I9fb90a8eab4c66d2f4231fdb482e0d97d128df3e
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The tests have been flaky on the CI. And indeed, running the tests
locally under valgrind, which heavily influences the timings, I could
reproduce failures on my local machine as well.
The failure was due to multiple things, but most notably boils down
to intermittend property update noficiations. To reduce that, the
myFactory.lastObj property is propagated up. This way, any change to
that property won't be propagated to the clients and influence the
test.
Furthermore, the failing tests are hardened by also checking the
server messages, which is crucial to correct the flow of messages.
Finally, a bug was found, where an empty property change could be
broadcasted to all clients, when only a specific client's object got
its properties changed. This is also fixed by this patch now.
Change-Id: I2c7ba53253e2841db1a290872dacb097a7b984cf
Reviewed-by: Bernd Lamecker <bernd.lamecker@basyskom.com>
Reviewed-by: Risto Avila <risto.avila@theqtcompany.com>
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Signals and property changes caused by dynamically created
qobjects (qobjects returned from a method call at runtime)
should not be broadcasted to any client but only sent to clients
which know these qobjects.
Also added testcases for the changes.
Change-Id: I9aacfa9e7e9df9314b44c6ba8e7339a2069e3c37
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Currently, a new client gets a list of _all_ registered QObjects,
whether they were explicitly registered or not.
This leaks internal information which the clients cannot use right away
anyway.
Change-Id: I4b25a731e9bc2d646f903057b409aecd34dc7f11
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|/
|
|
|
|
|
|
|
|
|
|
|
| |
Instead, send the data as a response to the initialization request
message. This simplifies the code and makes it more predictable, as we
do not spam all clients with initialization broadcasts anymore.
Note that the pending initialization "feature" is removed, but I don't
see a need for it anymore. If you want to delay client initialization,
do it on the client side.
Change-Id: I1ab71fd6c9e809ccb6085f1a3fbac3eb9b2e910b
Reviewed-by: Jocelyn Turcotte <jocelyn.turcotte@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
The web channel always connects to that signal automatically for
internal book-keeping purposes. Thus we do not have to connect to the
signal for wrapped objects. This simplifies the tests and reduces the
IPC traffic for wrapped objects.
Change-Id: Iaf8d9ce0b87874917cdcdf9013e21a53ee36b53a
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
| |
Some testcases do not consider all messages. These changes make
the test wait at specific points until these messages are sent.
Change-Id: Ie8f4b9492fefade0be91e7f829824db39d61ddee
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Renamed LICENSE.LGPL to LICENSE.LGPLv21
- Added LICENSE.LGPLv3 & LICENSE.GPLv2
- Removed LICENSE.GPL
Furthermore we need to update the sync.profile to not
point to the dev branch in order to integrate
this patch in the CI.
Change-Id: I06b5496b5d865e2da4808532362616429c969658
Reviewed-by: Sergio Ahumada <sahumada@blackberry.com>
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
This was a historic artefact which should not be required anymore.
Back then it was added when trying to improve the perceived performance
of the webchannel on an embedded device. Nowadays, using the built-in
WebKit IPC mechanism instead of WebSockets, paired with the rest of
the code cleanup and other changes, it should not be required anymore.
Change-Id: I4b2d76d84aaf92e12592a0b2bcf94710b697846e
Reviewed-by: Lutz Schönemann <lutz.schoenemann@basyskom.com>
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
| |
Returned QObject pointer got wrapped correctly but
a wrong QJsonObject was given to initializePropertyUpdates()
method.
Change-Id: I157b862ba2a90e87c295beb3c02fff932aac83c6
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The JS lib will miss to add QObjects to the local
object descriptions when receiving a new description
and the user has not registered a callback function.
This patch makes the lib execute unwrapQObject for
all responses received from a function call.
Change-Id: I72d872d65a7e7bec1d7e6ffb1eea674d0f6e50d7
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
| |
Change-Id: I66b102f0076d8ab39f871ce2be300a1e33228eac
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
| |
The tests don't depend on QtWebKit anymore, so they should be run even
without QtWebKit installed.
Change-Id: Id60ebec861b29b73522fba670b1a0c19375cb826
Reviewed-by: Milian Wolff <milian.wolff@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes a bug, where messages from the transport would still be
send to the QtMetaObjectPublisher even though it was previously
disconnected.
I'll refactor this code eventually to get rid of QtMetaObjectPublisher
alltogether and merge its code into QWebChannelPrivate where
appropriate.
Change-Id: Ie0c35bd81a5e633bdcb6be55b64f947d4a545a59
Reviewed-by: Jocelyn Turcotte <jocelyn.turcotte@digia.com>
Reviewed-by: Lutz Schönemann <lutz.schoenemann@basyskom.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This assimilates the JavaScript side to the QML/C++ side. We get rid
of the automagic WebSocket code. Instead, users pass in the WebSocket
from the outside, if they want to use that for communication. In the
QtWebKit/QtWebEngine cases, we will pass in our custom IPC objects.
Change-Id: I15e15b5130f99dc8b39dfbfa8cd3d8b2d34dbbc0
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@digia.com>
Reviewed-by: Lutz Schönemann <lutz.schoenemann@basyskom.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch removes the obsolete API support to send raw messages using
a QWebChannel. Instead, it is encouraged to directly use WebSockets or
navigator.qt.
By doing so, we can cleanup the code considerably. While at it, the
transport API is adapted to work on QJsonObject messages, instead of
QStrings. This will allow us to use more efficient formats in e.g.
QtWebKit or QtWebEngine. One could also implement a JSONRPC interface
using a custom transport then.
Change-Id: Ia8c125a5558507b3cbecf128a46b19fdb013f47b
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@digia.com>
|
|
|
|
|
|
|
| |
It worked before, but this patch makes the code more explicit.
Change-Id: Id7cf90cb3e2488106e08077e9a8165112734dd88
Reviewed-by: Michael Bruning <michael.bruning@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
This is a minor cleanup commit. Renaming the .pro file is required
for qt.pro to find it. The rename of the resource path makes it more
explicit by using the module name.
Change-Id: Ib724882d02d2a8aea7468ef984ed573dca1d3e7b
Reviewed-by: Michael Bruning <michael.bruning@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
This is required for proper QtWebKit/QtWebEngine integration, as
otherwise these modules would have to redo a lot of the QtWebChannel
QML API. Furthermore, without this, we could not use the WebChannel.id
attached property everywhere, independent of the web browser technology.
Change-Id: I032a9326841d505c2f77959a240bbfc71e94b6e8
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is a quite big changeset, but necessary to get the roadmap
implemented that was discussed at QtCS.
With this patchset landed, the QWebChannel does not depend on
QtWebKit anymore, not even for the tests. Rather, we will introduce
the dependency in the other way (i.e. QtWebKit will optionally use
QtWebChannel if available).
For the pure Qt/C++ use-case, we ship a utility implementation of
a QWebChannelAbstractTransport that uses a QWebSocket for the
server-client communication. This way, we can get rid of the custom
WebSocket implementation.
The tests are refactored to run the qwebchannel.js code directly
inside QML. Integration tests for QtWebKit/QtWebEngine as well
as examples will be added to these repositories.
Change-Id: Icc1c1c5918ec46e31d5070937c14c4ca25a3e2d6
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
|
|
|
|
|
|
|
|
|
| |
This simplifies the usage and lets properties be used just like normal
JavaScript properties. This is possible since the properties are cached
on the HTML side.
Change-Id: Ic60076f4596cd8df063567dfbd630e5bd6403119
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The auto tests are now located in tests/auto instead of directly in
tests/. This is required to ensure the cmake test is found.
Furthermore, the sync.profile is updated to point to refs/heads/dev,
as we target Qt 5.4 with this new module.
Change-Id: I1e6e99968b7081b5774eaf30319cac1fbaed35c2
Reviewed-by: Sergio Ahumada <sahumada@blackberry.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before, the response was sent to all clients in a broad-cast and had
to be filtered on the client-side. This required additional client
identification data to be added to all requests and responses.
Now, we keep track of the transport and transport-internal client and
only send the response to that client. This is very benefitial for
multi-client setups but also reduces traffic for single-client setups
and thus their performance.
Change-Id: Ia1ef5e031b0058222083d352a8aa28a7d566a6ca
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This further reduces the network traffic and thus leads to a small
performance boost.
Personally, I also think this code is a bit nicer to read and grasp.
Change-Id: I943c621142e9982f0e52d24e3a0976428856541b
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
|
|
|
|
|
|
|
|
|
| |
This will allow us to refactor the handling of multiple clients without
breaking the functionality.
Change-Id: I277eb5d7337d398fdf0694ef539d3b6ab7e5d332
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This enables us to optionally use navigator.qt instead of a WebSocket,
which is nicer setup-wise and is also slightly faster:
navigator.qt:
284.0 msecs per iteration (total: 2,840, iterations: 10)
WebSocket:
295.8 msecs per iteration (total: 2,959, iterations: 10)
The baseline is ca. 203 msecs, which would mean a performance boost
of ca. 12.7%.
Furthermore, this sets the fundation to eventually add a WebEngine
transport mechanism. The WebViewTransport should also be removed and
instead the WebView itself should directly implement the
WebChannelTransportInterface.
Change-Id: I368bb27e38ffa2f17ffeb7f5ae695690f6f5ad21
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The current CI system fails to build the webchannel when
QtWebKit is included in sync.profile, so we ignore it for now
and adapt the QMake build system to only include the declarative
tests when webkit is found.
As discussed with tronical, sifalt and sahumada, the qtqa scripts
will need to be adapted to cope with optional dependencies of non-qt5
modules.
Change-Id: Id89b763ef2697e9e72eb4064c150971b13ebccc3
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some tests referenced Nokia in their license even though that was never
the case. The tests where written completely by me after Qt Nokia times.
What is missing are the examples which are still mostly original work
by Noam back then in Nokia times. The rest was (re-)written by me
completely since then anyways.
Change-Id: Ib423fb3459bcc1f7464a02de4fd82ddfd614d282
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This used to work but broke in one of the last few commits apparently.
Now we add a proper unit test to ensure it stays working.
The issue was that JSON only knows e.g. numeric types stored as double.
When we then try to call a method taking an int with the VariantArgument
that tries to convert the double to int, we failed and produced an
invalid QVariant which then converts to 0.
Now we use the appropriate API to convert the JSON data to the correct
target type before calling the method. Furthermore, it became clear
that we can greatly cleanup the VariantArgument thanks to that. It now
is reduced to just a QVariant wrapper class with an implicit cast
operator to QGenericArgument.
Change-Id: Ieaf60f548ea9584e7d760f9cd935da455787f376
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The new registeredObjects list property is now preferred over the old
imparative registerObject/registerObjects API.
Items that are added to the list need an attached WebChannel.id property
which holds the identifier under which the object is published to remote
clients.
Change-Id: I96a8047b9a85e27f3fd48c900180c22ebd20eb35
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This is achieved by hiding the MetaObjectPublisher completely as
private API. The QWebChannel is the only publisher API and now handles
both the socket as well as the publisher internally.
This now allows us to create a proper QML api in the new QmlWebChannel.
Change-Id: I3096364af8485353ca9bc19df4a81a8e4552c3d7
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
| |
These messages are somewhat unpredictable in order and as such we skip
them instead of relying on their order.
Change-Id: I24c082ea1afe6ede1a8058195f01ffba1e6ca8e3
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The code now resides in a single qwebchannel.js file and there is only
a single callback-nesting required to setup a MetaObjectPublisher
connection.
The server-side will be simplified in the next step.
Change-Id: Ib5fc77a03c2b281c61af91713411eed571ec6108
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
| |
In addition to these steps, the developer must symlink the qmldir from
the source directory to <his build-folder>/qml/QtWebChannel.
Change-Id: Ieb1131edc907c98d9ce9134760f14633b07eb080
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reduces the traffic as the indices are usually much smaller than
the property names.
As such, the benchPropertyUpdates gets a speed boost of about 9% (or
10ms vs. 11ms). As we need to transmit the index during initialization
that degrades its performance slightly by ca. 4% (13ms vs. 12.5ms).
Considering that the initialization only takes place once whereas
the property updates potentially often, this is a good tradeoff.
Change-Id: If7df3e360f1528b7d7aa26c63ce851363ae9fd6a
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
| |
Change-Id: Ibd9a696a86939530cd08d57a9ef9cbeb3125bbd8
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Before, we constructed QVariant maps and lists and then converted them
to JSON to send the data to the webchannel.
By obsoleting the conversion step, benchInitializeClients shows a good
performance boost of ca. 19% (11.81ms vs 14.58ms).
Change-Id: Ief8e8127207a046f481488a478cd6a18fa0ebffe
Reviewed-by: Zeno Albisser <zeno.albisser@digia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This allows us to remove the public API for the tests and allows for
more tests and benchmarks in the future.
To achieve this, we re-use the new qmetaobjectpublisher_p.h, which then
also must be exported.
Change-Id: I3c33b2f5be6cc674cd3092667151dd8da2263cf5
Reviewed-by: Zeno Albisser <zeno.albisser@digia.com>
|
|
|
|
|
|
| |
Change-Id: I6a852e1545e7aac0f9f8d83010a3e905524a8069
Reviewed-by: Zeno Albisser <zeno.albisser@digia.com>
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This module can hopefully be done in time for 5.3. This commit changes
the source structure and QMake files to adapt to typical Qt modules.
With this in place, we can now use QT += webchannel in qmake files to
link against the pure Qt/C++ QtWebChannel library.
The QML plugin is separated from it and can be loaded optionally, if
the quick module could be found. Also added is now a qmlplugindump
for tooling integration. Note that the Qt.labs namespace is removed.
The test file structure is also adapted to how its done in the
QtDeclarative module.
Note that this setup apparently does not support to run tests without
running make install first.
Change-Id: I1c15d72e7ab5f525d5a6f651f4e965ef86bc17bd
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This will allow us to create a stand-alone WebChannel C++ library,
without any QML dependencies, that can be used to publisher QObjects
to remote clients running in any browser engine supporting WebSockets.
The patch is large, as working with introspection through the
QMetaObject from C++ is more complicated compared to QML.
On the other hand, the move to C++ allows a much more performant
implementation of quite some parts of the publisher. One thing is
that signal and method invocations can be handled via numeric indices,
where before we needed to transmit the string identifier of the
signal or method.
Eventually this can now also be applied to properties, further
decreasing the size of messages between server and clients.
Note that this patch contains quite some TODOs and rough edges,
such as the invokable bench_* helper functions in the public API
of the MetaObjectPublisher. These are to be seen as temporary, and
will be cleaned up in followup commits later. This is done to prevent
a further blow-up of this already big patch.
Change-Id: I57e788d8a19edd410651611382d912f9ad6660c9
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
|
|
|
|
|
|
|
|
| |
To do so, we need to remember the connected functor and pass that to
disconnect, in both the client side as well as the QML server side.
Change-Id: Ic61fc5d2a203212278c23471c216683e309e2c9f
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
An HTML client can trigger the creation of new objects, or a published
object on the server side might return other objects.
These are then wrapped on the fly and can be used like the other
objects. Note though that the HTML client can call deleteLater on
these objects. Also, it does not yet work to wrap objects on the fly
in signal arguments or property values.
Change-Id: I92aa8a3e52f42d5325dd0771bbf9e2ae213e88f9
Reviewed-by: Arvid Nilsson <anilsson@blackberry.com>
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This runs benchmarks on various functions of the MOP with hundred
objects each containing ten properties, signals and methods.
Furthermore a function is included which benchmarks the whole roundtrip
of loading a simple client and waiting for it to be initialized.
Change-Id: If51c5b5f606151885611cabc0e1330bd2643ad67
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
| |
For performance reasons, we group property updates and send a single
batch update notification. This is now tested to work as intended.
This also uncovered a bug in webchannel.js when multiple functions
subscribe to the same id.
Change-Id: Ic8648d664dd1fe54df7e25fade6a6088386af992
Reviewed-by: Zeno Albisser <zeno.albisser@digia.com>
|