aboutsummaryrefslogtreecommitdiffstats
path: root/sources/shiboken2/ApiExtractor/abstractmetabuilder.cpp
Commit message (Collapse)AuthorAgeFilesLines
...
* Remove attributes Final and FinalInCppFriedemann Kleint2017-12-211-12/+9
| | | | | | | | | | Apparently, the meaning of FinalInCpp for functions was "virtual and not private". The previous code cleanup removed most uses of them. For the remaining cases, checking for virtual is sufficient. Change-Id: I6f794e36b99920af36fd0eba895be673dc35d9dd Reviewed-by: Christian Tismer <tismer@stackless.com> Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
* shiboken/AbstractMetaBuilder: Remove create* functionsFriedemann Kleint2017-12-091-50/+15
| | | | | | | | There are no classes deriving from AbstractMetaBuilder, so there is no need to have virtual creation functions. Change-Id: I2d4c494ac6b0a17ceea71939a1f07a49a369df73 Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
* shiboken: Stop using the FinalInCpp attribute for classesFriedemann Kleint2017-12-091-3/+6
| | | | | | | | | | | | | | | Besides its use for functions, the FinalInCpp attribute was used for classes meaning something like neither the class nor its derived classes can be constructed due to presence of private destructors and absence of non-private constructors. Since 'final' is now a reserved word in C++ 11, this becomes confusing. Remove the usage of the attribute for classes, add a bit field indicating presence of a private constructors and a new function isConstructible() instead. Change-Id: Ieb2f9943c21511e4f59d228885361e3ad66fb266 Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io> Reviewed-by: Christian Tismer <tismer@stackless.com>
* shiboken/MetaLang: Add information about override/finalFriedemann Kleint2017-12-021-1/+11
| | | | | | | Take over information from code model and test. Change-Id: Ibed2973e09b117966ed241798125f60ba337b0e9 Reviewed-by: Christian Tismer <tismer@stackless.com>
* shiboken/AbstractMetaAttributes: Remove some unused attributesFriedemann Kleint2017-12-011-3/+0
| | | | | | | Remove Native, FinalOverload, InterfaceFunction, Fake. Change-Id: I1232751169fafb144387343857a962bafee714e9 Reviewed-by: Christian Tismer <tismer@stackless.com>
* CodeModel: Handle access of inherited classesFriedemann Kleint2017-09-251-10/+18
| | | | | | | | | | | | | | Store the access of the base class in the code model. Remove protected/private base classes in the abstract meta builder as otherwise invalid code will be generated for classes like QAbstract3DGraph : public QWindow, protected QOpenGLFunctions when the protected hack is not in use (on Windows). Task-number: PYSIDE-487 Change-Id: I5bc2dad16f903da86e6e41450c2f9a76726ac028 Reviewed-by: Christian Tismer <tismer@stackless.com>
* shiboken2: Make message about unmatched functions more verboseFriedemann Kleint2017-09-041-2/+23
| | | | | | | | List all member functions if a candidate cannot be found (spelling error or similar). Change-Id: I86ca1556078051d2fc1f78f2091f5176f2a65423 Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
* FunctionModification: Make it possible to specify a regular expressionFriedemann Kleint2017-08-291-3/+4
| | | | | | | | | It should make it easier to specify the <array> modifications for GL functions. Task-number: PYSIDE-516 Change-Id: Ieb2e540f61785d13ee46a196a18d03b311d308e1 Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
* Fix the chain of issues regarding QSslConfigurationAlexandru Croitor2017-08-241-1/+22
| | | | | | | | | | | | | | | | | | | | | | | | | The failure was seen on Windows CI build, but was actually present on all platforms if the QSsl checks were properly fixed, or if certain CMake versions randomly decided to force C++11 usage. The first issue was that QSsl wrappers were not generated on all platforms because the C++11 standard was not forced for the compilation tests done in the check_qt_class macro. This is a bug in CMake, and the official way of fixing this is introduced in the yet unreleased CMake 3.9 version. The current fix is a workaround to explicitly pass the C++11 standard switch to the try_compile invocation. The next issue concerns handling of const char[] C++ types. There are three such members in QSslConfiguration, which caused build failures due to incorrectly generated code. The solution is to treat "const char[]" types as "const char*" types, which generates correct conversion code. Tests were also added to check for such cases. Change-Id: I874a3591dfc5f385338de7e3aff2a2c0dd2f5719 Reviewed-by: Christian Tismer <tismer@stackless.com> Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
* Shiboken2: Add an index check for the <array> modificationFriedemann Kleint2017-08-221-1/+6
| | | | | | | | | Amends change 1578d14d1693122d4bcacede0d77527d8e898629. Task-number: PYSIDE-354 Task-number: PYSIDE-516 Change-Id: I50e5d0762f1b049a511c62f8aa8d08da4cb563d9 Reviewed-by: Christian Tismer <tismer@stackless.com>
* Add new <array> argument modification indicating array usageFriedemann Kleint2017-08-031-6/+46
| | | | | | | | | | | | The modification can be used to indicate that for example int* is meant to be used as int[]. This is reflected in the NativePointerAsArrayPattern usage pattern of AbstractMetaType. Task-number: PYSIDE-354 Task-number: PYSIDE-516 Change-Id: Icaeb3cce4be9ce06caa2cab628d4e8fc1b684819 Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
* Shiboken: No longer change arrays[] to pointer typesFriedemann Kleint2017-08-031-33/+24
| | | | | | | | | | In AbstractMetaBuilder, no longer change array types like "int[]" to "int*". Task-number: PYSIDE-354 Task-number: PYSIDE-516 Change-Id: Ia9e15ae3fca895bf179275eb31a94323d91f4941 Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
* Remove AbstractMetaFunction::invalid()Friedemann Kleint2017-05-291-18/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Change AbstractMetaBuilderPrivate::traverseFunction() to consistently return 0 if something does not fit. Previously, functions with half-parsed arguments with the invalid flag set were returned. This caused strange side effects since the flag was not checked in all places. The only relevant information is whether some constructors were rejected. In that case, no default constructors or default copy constructors should be generated. This is now determined by checking the code model function item; the attribute HasRejectConstructor is introduced for this. This fixes: - Make it possible to reject the QTextStreamManipulator constructor taking a function pointer without having a default constructor generated: typedef void (QTextStream::*QTSMFI)(int); QTextStreamManipulator(QTSMFI m, int a) Q_DECL_NOTHROW - Implement QtCharts whose class QAbstractSeries has a rejected constructor. Change-Id: I6310574ba677dac20699f257340d2c2a55674353 Reviewed-by: Christian Tismer <tismer@stackless.com>
* Use information about Constructor/Destructor functions from ClangFriedemann Kleint2017-05-291-45/+48
| | | | | | | Expand CodeModel::FunctionType accordingly. Change-Id: I2759f45e4c60f36d4583b1c475c4c2df1e00ace6 Reviewed-by: Christian Tismer <tismer@stackless.com>
* move everying into sources/shiboken2 (5.9 edition)Oswald Buddenhagen2017-05-221-0/+3459
in preparation for a subtree merge. this should not be necessary to do in a separate commit, but git is a tad stupid about following history correctly without it.
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-05 12:09:08 +0000