| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Change-Id: I3e7c5fc0ad9fba12d791930fab0f1e115a2a379a
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
|
|
|
|
|
| |
Merge-request: 1853
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
|
|
|
|
|
|
|
| |
of code that already uses exceptions.
Merge-request: 1853
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
|
|
|
|
|
| |
Merge-request: 1854
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
|
|
|
|
| |
No need to use our file extension; it's JavaScript code.
|
|
|
|
| |
Trolltech ASA is no more, long live Qt by Nokia!
|
|
|
|
|
|
| |
Since Date and RegExp are actually ECMA-compliant in 4.6, we
should now generate bindings for the corresponding Qt classes,
so that those may be used when you want Qt semantics.
|
| |
|
| |
|
|
|
|
|
| |
Merge-request: 1714
Reviewed-by: Kent Hansen <khansen@trolltech.com>
|
|
|
|
|
| |
Merge-request: 1531
Reviewed-by: Kent Hansen <khansen@trolltech.com>
|
|
|
|
| |
rootState is no more, the machine is the root.
|
| |
|
|
|
|
|
| |
*/* is slightly ambigious for the compiler, and depends on the
preceding code. MSVC warns about it.
|
|
|
|
|
| |
the qscript connect() function throws an error when attempting
to connect to an overloaded signal, so we need to disambiguate.
|
|\ |
|
| |
| |
| |
| |
| | |
The shellimplgenerator was not updated to pass in the correct path.
This caused linker errors.
|
|\| |
|
| |
| |
| |
| |
| |
| | |
Prefix by the package name, e.g. com_trolltech_qt_core_init.cpp.
This way qmake doesn't get confused if you include multiple .pri
files in the same project.
|
| | |
|
|\| |
|
| |
| |
| |
| | |
Hey, seems to compile now.
|
| | |
|
| |
| |
| |
| | |
QEasingCurve is a value type.
|
|\| |
|
| |
| |
| |
| | |
Avoid heaps of warnings.
|
| |
| |
| |
| | |
The cheap way for now.
|
| | |
|
| | |
|
|/
|
|
| |
Follow-up to http://code.google.com/p/qtscriptgenerator/issues/detail?id=8
|
| |
|
|
|
|
|
| |
Details: makes it easy to statically compile the bindings into an application without requiring a plugin;
just include the generated .pri file in your .pro file, then call e.g. qtscript_initialize_com_trolltech_qt_gui_bindings()
|
| |
|
|
|
|
| |
Details: it's a bug in Jambi's typesystem, but we can still fix it
|
|
|
|
|
|
|
| |
Task: http://code.google.com/p/qtscriptgenerator/issues/detail?id=29
Details: the generic conversion will check if the meta-type ID is that of short
or float, which it's not going to be because all numbers are doubles
when stored in a QScriptValue.
|
| |
|
| |
|
| |
|
|
|
|
| |
Task: http://code.google.com/p/qtscriptgenerator/issues/detail?id=21
|
|
|
|
| |
Task: http://code.google.com/p/qtscriptgenerator/issues/detail?id=18
|
|
|
|
| |
Task: http://code.google.com/p/qtscriptgenerator/issues/detail?id=16
|
|
|
|
|
|
|
| |
Details: this one is provided as a convenience in C++, but in QtScript you'll
just end up forgetting to call end() on the painter (very typical
error, I still make it myself!). So now you have to call begin(), which
makes it balanced.
|
| |
|
|
|
|
|
| |
Details: the debugger will pop up if there is an uncaught exception,
or if the debugger keyword is encountered
|
| |
|
|
|
|
|
|
| |
Details: no more noise when running the generator. Many of the
warnings were Jambi-specific, many were just plain
annoying. Yet others were valid and have now been fixed.
|
|
|
|
|
|
| |
Task: none, reported on qtscript-bindings mailing list
Details: the classes were excluded because they didn't compile at some point;
well, they seem to compile just fine now. :-)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Task: http://code.google.com/p/qtscriptgenerator/issues/detail?id=8
Details: The situation occurs when the virtual function QProgressBar::text() is called from C++.
As usual, the binding will try to look up a property with the corresponding name ("text")
in the script object. "text" is also the name of a Qt (C++) property, so it will be
resolved through the dynamic QObject binding, which will call the C++ getter again, and so
on and so on...
This submit fixes the problem by having the binding add a "_qs_" prefix when doing the
script property lookup; so if a script wanted to reimplement the text() function, it
would have to store it in a property named _qs_text. Probably not optimal, but at least
it fixes the recursion while still allowing you to reimplement the function. It might
or might not become the official (i.e. documented) way of doing it, at some point.
|
|
|
|
| |
Task: http://code.google.com/p/qtscriptgenerator/issues/detail?id=15
|
|
|
|
|
| |
Details: checking for isVariant() is useless; what we want to do is just skip the test
for the QVariant argument, since it effectively means "any value is accepted"
|