summaryrefslogtreecommitdiffstats
path: root/src/script/api/qscriptengine_p.h
Commit message (Collapse)AuthorAgeFilesLines
* Fix conversion between JavaScript Date and QDateTimeKent Hansen2010-06-251-4/+6
| | | | | | | | | | Use JavaScriptCore's conversion functions rather than our own (incomplete) implementation. Specifically, this means daylight saving time is finally handled correctly on Windows. Task-number: QTBUG-9770 Reviewed-by: Olivier Goffart
* Use custom static scopes to improve QML/JavaScript performanceKent Hansen2010-06-251-0/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This commit introduces a new internal JS object type, QScriptStaticScopeObject, that enables the JS compiler to make more aggressive optimizations of scoped property access. QScriptStaticScopeObject registers all its properties in a symbol table that the JS compiler has access to. If the compiler finds the property in the symbol table, it will generate the fast index-based op_{get,put}_scoped_var bytecodes, rather than the dynamic (slow) op_resolve and friends. If the compiler _doesn't_ find the property in the symbol table, it infers that it's safe to skip the scope object when later resolving the property, which will also improve performance (see op_resolve_skip bytecode). QScriptStaticScopeObject is only safe to use when all relevant properties are known at JS compile time; that is, when a function that has the static scope object in its scope chain is compiled. It's up to the user of the class (e.g. QtDeclarative) to ensure that this constraint is not violated. The API for constructing QScriptStaticScopeObject instances is not public; it lives in QScriptDeclarativeClass for now, an internal class exported for the purpose of QML. The instance is returned as a QScriptValue and can be manipulated like any other JS object (e.g. by QScriptValue::setProperty()). The other part of this commit utilizes QScriptStaticScopeObject in QtDeclarative in the two major places where it's currently possible: 1) QML disallows adding properties to the Global Object. Furthermore, it's not possible for QML IDs and properties to "shadow" global variables. Hence, a QScriptStaticScopeObject can be used to hold all the standard ECMA properties, and this scope object can come _before_ the QML component in the scope chain. This enables binding expressions and scripts to have optimized (direct) access to e.g. Math.sin. 2) Imported scripts can have their properties (resulting from variable declarations ("var" statements) and function declarations) added to a static scope object. This enables functions in the script to have optimized (direct) access to the script's own properties, as well as to global properties such as Math. With this change, it's no longer possible to delete properties of the Global Object, nor delete properties of an imported script. It's a compromise we make in order to make the optimization safe. Task-number: QTBUG-8576 Reviewed-by: Aaron Kennedy Reviewed-by: Olivier Goffart Reviewed-by: Jedrzej Nowacki
* Usefully convert from QtScript object/array to QVariantKent Hansen2010-06-171-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For arrays, the conversion would crash if the array was cyclic. Introduce a set that keeps track of which objects are being converted, and return an empty list when a cycle is detected. For other types of objects, the object was previously attempted to be converted to a primitive, which typically meant you would get the string representation ("[object Object]"), since most (practically all) objects can be converted to a string -- not useful at all. Change the conversion so it converts the object to a QVariantMap instead. This was already done for slots that took a QVariantMap as argument, but only one level deep. Make the conversion recursive, using the same mechanism as for arrays to detect cycles. This change also means that you get a meaningful JS object => QVariant => JS object roundtrip. It also aligns the behavior with the Qt WebKit bridge. Update the documentation to describe the new behavior. The 4.7 changelog will also be updated under "important behavioral changes". This change exposed an issue with one of the QML autotests: A JS object was assigned to a QVariant property, which caused it to be converted to a string (rather than a QVariantMap) -- just shows that the previous behavior was unintuitive). Later, this variant property is compared to another object, the intention being to compare the _properties_ of the two objects; but because the variant property contained a string, this would cause the other operand (object) to be converted to a string as well ("[object Object]"), causing a meaningless test pass. Change the test to deserialize both objects using JSON.stringify, and compare the resulting strings, so that actual JS object => QVariant(Map) => JS object roundtrip is tested (the intention). Task-number: QTBUG-3511 Reviewed-by: Olivier Goffart
* Speed up qsTr() by caching the translation contextKent Hansen2010-04-151-1/+6
| | | | | | | | | | | | | | | | | | | | | | qsTr() uses QFileInfo::baseName() to determine the translation context from a URL. The problem is that creating a QFileInfo object (and thus a file engine object), as well as processing the URL to determine the base name part, is very expensive. By caching the last translation URL and context, qsTr() becomes 5x faster. Only the most recent URL is cached, because a script's text will be translated all at once (as the script is evaluated), so for scripts with multiple translated strings only the first qsTr() call will cause a cache miss. The performance could be improved even further by getting rid of the QFileInfo dependency altogether; created QTBUG-9939 for that since it's a more risky change. Task-number: QTBUG-6908 Reviewed-by: Olivier Goffart
* QtScript: Make sure the old identifier table is restoredKent Hansen2010-03-261-1/+0
| | | | Use a shim to take care of that.
* QtScript: Add API for reporting additional memory costsKent Hansen2010-03-251-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | QScriptEngine::reportAdditionalMemoryCost(int). This function provides the ability to give a hint to the engine that it should perhaps trigger garbage collection sooner rather than later. For example, if you've implemented a JS ByteArray class that wraps a QByteArray, and a user constructs a few hundred temporary ByteArray objects of large sizes, failure to report the additional memory cost may cause the application's memory consumption to grow and grow (because the script engine thinks they are "cheap" objects, the GC won't kick in). Reporting the correct size can be difficult (or impossible) in some cases. For example, it's difficult to predict the total amount of system memory & resources consumed by a QImage. But even reporting a heuristic / approximate cost can be better than reporting no cost. Task-number: QTBUG-6238 Reviewed-by: Simon Hausmann
* Add default argument values for QScriptValuePrivate property functionsKent Hansen2010-03-101-3/+3
| | | | | | | | Make them mirror the defaults of the public API, so that the QScriptValuePrivate property functions can be used internally in the same manner as the public API. Reviewed-by: Jedrzej Nowacki
* Move method implementation to private classKent Hansen2010-03-101-0/+1
| | | | | | | | In preparation of being able to call toStringHandle() directly on a JSC::Identifier in QScriptValueIterator implementation. Reviewed-by: Jedrzej Nowacki
* Update src/3rdparty/javascriptcore and adapt src/script to the changesKent Hansen2010-03-101-0/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Update qscriptvalueiterator test to expect length property when iterating arrays and strings. - Use EvalExecutable::create() instead of EvalExecutable constructor. The constructor is private. - Reimplement getOwnPropertyDescriptor() in all custom script objects. - Remove all reimplementations of getPropertyAttributes(). It doesn't exist in trunk anymore (getOwnPropertyDescriptor() is used instead). - Remove checkDontDelete argument from deleteProperty() reimplementations. The purpose of this argument was to support deleting properties with attribute Undeletable from C++. But it was quite an invasive patch to JavaScriptCore, and it doesn't seem worth it. If this feature is really crucial it should be re-done upstream. One of the tests needed to be updated so it's not sensitive to the C++ undeletability. - Adapt getOwnPropertyNames() reimplementations to signature change. - Add missing QScriptObject structure flags, otherwise we don't get all virtual calls. - Remove our patch for reporting column numbers in the debugger callbacks. It was just too intrusive. As with the checkDontDelete issue, this should be redone upstream if it's really important. In 4.7, QScriptEngineAgent will always report a column number of 1. Other compilation fixes: - InternalFunction::name() takes an ExecState* argument, not GlobalData* - ScopeChain::globalObject is no longer a function but a member variable - ScopeChainNode constructor takes a GlobalObject argument - Heap::collect() is called collectAllGarbage() - JSValue::strictEqual() takes an ExecState* argument - Debugger::exception() takes a bool hasHandler argument - Debugger no longer reports column number (we decided to drop that patch from JSC) - UString doesn't have operator+=(char*) - Update the autotests to reflect the columnNumber=1 change. - Add helper class to avoid crashing inside JSC. Ever since r52856 in WebKit trunk, this is needed. There are probably a lot of other public API functions that need this guard as well, but I'll add them as they are discovered. - Update mkdist-javascriptcore tag, exclude a few more files. - Set ENABLE_JSC_MULTIPLE_THREADS=0 define on Mac due to r52355 in trunk. Reviewed-by: Simon Hausmann
* Perform latin1 conversion directly on JSC::UStringKent Hansen2010-03-011-0/+18
| | | | | | | | | This conversion needs to be as fast as possible since it's performed every time you access a QObject property from QtScript. Hence, we should avoid going via QString and instead do the conversion ourselves. Reviewed-by: Jedrzej Nowacki
* Avoid conversion between JavaScriptCore String type and QStringKent Hansen2010-03-011-10/+10
| | | | | | | Use JSC::UString internally. Only when the public API demands it should a JSC::UString be converted to a QString. Reviewed-by: Jedrzej Nowacki
* Merge branch '4.6' of scm.dev.nokia.troll.no:qt/qt-s60-public into master-s60axis2010-02-261-2/+10
|\ | | | | | | | | | | Conflicts: qmake/generators/symbian/initprojectdeploy_symbian.cpp qmake/generators/symbian/symmake_abld.h
| * Fix memory leak when lazily binding QScriptValue to an engineKent Hansen2010-02-221-2/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Avoid the engine's list of free script values from growing without bounds. When a QScriptValue is initially not bound, its private will be allocated from the normal heap (and not from the engine's pool of privates, because there is no engine at this point). But when a value is later bound (e.g. by setting it as a property of an object, or by passing it as argument to QScriptValue::call()) and is subsequently destroyed, its private will be handed to the engine, which will add it to its free-list (hence the memory is not freed). This allocation/deallocation asymmetry causes this list go keep growing. The solution is to limit the size of the free-list, and free the memory of the private immediately when the list has reached a certain size. Task-number: QTBUG-8400 Reviewed-by: Olivier Goffart
* | QScriptValue::isQMetaObject crash fix.Jedrzej Nowacki2010-02-241-1/+1
| | | | | | | | | | | | | | | | QScriptValue::isQMetaObject shouldn't crash for a value that internally is represented as a non JSObjects values. Additional check was added to QScriptEnginePrivate::isQMetaObject. Reviewed-by: Kent Hansen
* | Work around MSVC2008 compiler crashKent Hansen2010-02-191-1/+12
| | | | | | | | | | | | | | | | | | "e:\pulse\work\91088\src\script\api\qscriptengine.h(360) : fatal error C1001: An internal error has occurred in the compiler. (compiler file 'f:\dd\vctools\compiler\utc\src\p2\main.c[0x510A0530:0x00000007]', line 243) To work around this problem, try simplifying or changing the program near the locations listed above." Apparently the compiler doesn't like that a few functions are inlined.
* | Avoid calling out to public API in the QtScript implementationKent Hansen2010-02-181-16/+134
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There's no reason to construct QScriptValues when JSC::JSValues can be operated on directly. A benchmark showed that ~10% of the time for reading a QObject property from a script was spent just creating and destroying QScriptValue temporaries. It's time to finally get rid of this potential bottleneck, not just in the QObject integration but everywhere. This change refactors the code so that all internal operations are performed on JSC::JSValue (most importantly, conversion from/to Qt types), and the public API functions are just thin wrappers around these. E.g. instead of doing enginePrivate->scriptValueFromJSCValue(jsValue).toQObject() the implementation now does QScriptEnginePrivate::toQObject(jsValue) Other operations are delegated to the engine implementation in similar style. Task-number: QTBUG-8304 Reviewed-by: Jedrzej Nowacki
* | Move property helper functions to QScriptEnginePrivateKent Hansen2010-02-181-15/+84
| | | | | | | | | | | | More preparation for operating purely on JSC::JSValue internally. Reviewed-by: Jedrzej Nowacki
* | Move more script value conversion code to helper functionsKent Hansen2010-02-181-0/+131
| | | | | | | | | | | | In preparation of operating purely on JSC::JSValue internally. Reviewed-by: Jedrzej Nowacki
* | Cleanup: Move value conversion code to helper functionsKent Hansen2010-02-181-0/+50
| | | | | | | | | | | | In preparation of doing this conversion in more places. Reviewed-by: Jedrzej Nowacki
* | Move implementation of QScriptValue construction functions to private classKent Hansen2010-02-181-0/+39
| | | | | | | | | | | | | | | | In preparation of getting rid of QScriptValue construction internally; the implementation should only call the private functions that operate directly on JSC::JSValues. Reviewed-by: Jedrzej Nowacki
* | Cleanup: Move number conversion functions to QScriptEnginePrivateKent Hansen2010-02-181-0/+5
| | | | | | | | | | | | | | Also rename ToUint{16,32} to ToUInt{16,32} to follow the Qt naming (it takes precedence over the ECMA one). Reviewed-by: Jedrzej Nowacki
* | Move some helper function declarations outside QT_NO_QOBJECT guardKent Hansen2010-02-181-13/+14
| | | | | | | | | | | | | | These are not dependent on QObject, I don't know how they ended up inside there. Reviewed-by: Jedrzej Nowacki
* | Cleanup: Move exception helper functions to QScriptEnginePrivateKent Hansen2010-02-181-2/+5
|/ | | | | | Because that's where they belong. Reviewed-by: Jedrzej Nowacki
* Update copyright year to 2010Jason McDonald2010-01-061-1/+1
| | | | Reviewed-by: Trust Me
* Fixed the QtScript license information in the source files to referSimon Hausmann2009-11-171-19/+1
| | | | | | | | to the LGPL only. To do this I ran replace-licenses.zsh $QTDIR/src/script release Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
* Replace LGPL license tags with LGPL-ONLYSimon Hausmann2009-11-171-1/+1
| | | | Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
* Add QScriptDeclarativeClassAaron Kennedy2009-11-101-1/+1
| | | | | | | | | QScriptDeclarativeClass is a private, but exported, class used by the declarativeui module. It is very similar to QScriptClass, but slightly faster and provides a couple of "backdoor" extension mechanisms used by declarative. Reviewed-by: Warwick Allison
* Say hello to QScriptProgram :-)Kent Hansen2009-10-281-0/+5
| | | | | | | | | | | QScriptProgram encapsulates a Qt Script program (AKA a script). It retains the compiled representation of the script, so that repeated evaluation of the same script becomes faster. An overload of QScriptEngine::evaluate() that takes a QScriptProgram has been added. Reviewed-by: Olivier Goffart
* Inline QtScript exception helper functionsKent Hansen2009-10-231-0/+16
| | | | | | Makes QScriptValue::toNumber() ~50% faster. Reviewed-by: Olivier Goffart
* Inline internal property lookup functionKent Hansen2009-10-231-0/+22
| | | | | | | | | Also avoid looking up the object's own properties twice (before we called getOwnPropertySlot() and then getPropertySlot() on the same object). Makes QScriptValue::property() ~20% faster when calling it on an "empty" object. Reviewed-by: Olivier Goffart
* Inline scriptEngineFromExec() functionKent Hansen2009-10-231-2/+18
| | | | | | Makes QScriptContext::engine() 80% faster. Reviewed-by: Olivier Goffart
* Inline internal QtScript functions (frameForContext())Kent Hansen2009-10-231-2/+12
| | | | | | Makes QScriptContext::parentContext() 50% faster. Reviewed-by: Olivier Goffart
* Inline two internal QtScript functions (contextForFrame() and globalExec())Kent Hansen2009-10-231-2/+17
| | | | | | Makes QScriptEngine::currentContext() 25% faster. Reviewed-by: Olivier Goffart
* Use the qsreal type instead of double when working with QtScript numbersKent Hansen2009-10-221-1/+1
| | | | | | | | The idea is that qsreal can be typedef'ed to float on platforms where it's appropriate. Since the QScriptValue ctor takes a qsreal, we should not convert it to a double internally. Reviewed-by: Olivier Goffart
* Regression fix. Fix the hasUncaughtException() flag in debugger's event.Jedrzej Nowacki2009-10-191-0/+5
| | | | | | | | | | | The QScriptEngine::hasUncaughtException() flag should be set to true if returning from a JS function was caused by an exception. According to documentation, the flag had to be accessible from the QScriptEngineAgent::functionExit event. New autotest was added. Reviewed-by: Kent Hansen
* Inline QScriptValuePrivate operator new and deleteKent Hansen2009-10-071-0/+16
| | | | Make allocation faster.
* Fix column number provided to QScriptEngineAgentKent Hansen2009-10-011-1/+72
| | | | | | | | Introduced a helper function in our custom source provider, columnNumberFromOffset(), that maps an absolute offset in the source input to a relative column number. Reviewed-by: Jedrzej Nowacki
* QScript: Fix strange bugs and crashes.Olivier Goffart2009-09-281-1/+2
| | | | | | | | | | I was assuming that the default return value register was always set to 0 for native calls. But this is not the case. So we must ensure this. Also be consistend in the way the stackframe grow and shrink. This expose another bug in the way the call frame is created in JSC Reviewed-by: Kent Hansen
* Update src/3rdparty/javascriptcore and adapt src/script to the changes.Kent Hansen2009-09-241-1/+5
| | | | Reviewed-by: Simon Hausmann
* make sure the engine's currentFrame is in sync when calling out to public APIKent Hansen2009-09-221-0/+23
| | | | | | | The currentFrame pointer is used e.g. by QScriptValue::toString(). It needs to be in sync, otherwise we will crash. Reviewed-by: Olivier Goffart
* put the this-register calculation into a functionKent Hansen2009-09-181-0/+1
| | | | | | Avoid copy and paste. Reviewed-by: Olivier Goffart
* Fix compilation with winscwSimon Hausmann2009-09-101-2/+2
| | | | | | | | Winscw gets very confused when the name of an enum value is the same as the name of an entire namespace, JSC in this case. Renaming the enum value to JavaScriptCore fixes this. Rubber-stamped-by: Kent
* Update license headers again.Jason McDonald2009-09-091-4/+4
| | | | Reviewed-by: Trust Me
* improve memory management scheme of QScriptString(Private)Kent Hansen2009-09-021-5/+24
| | | | | | | | Get rid of QPointer. Use linked list of privates (like was recently done for QScriptValue). Allocate the private on the stack when we can. Reviewed-by: Olivier Goffart
* Optimize QScriptClassOlivier Goffart2009-09-021-0/+10
| | | | | | | Do not convert JSC::Identifier to QString to convert it later to JSC::Identivier again Reviewed-by: Kent Hansen
* Update tech preview license header for files that are new in 4.6.Jason McDonald2009-08-311-13/+13
| | | | Reviewed-by: Trust Me
* make some internal qtscript functions inlineKent Hansen2009-08-241-7/+111
| | | | Make it faster.
* compile fix with namespaced Qthjk2009-08-241-2/+2
|
* handle memory management of QScriptValuePrivate in engine if possibleKent Hansen2009-08-241-0/+4
| | | | | | | | | Avoid calling malloc() and free() so often. The premise is that QScriptValue is usually a short-lived type, and only a few QScriptValues exist at a time, so if we cache privates in the engine, QScriptValues will be much faster to create and destroy. Reviewed-by: Olivier Goffart
* Fix obsolete license headers.Jason McDonald2009-08-211-1/+1
| | | | Reviewed-by: Trust Me
generated by cgit v0.12 at 2025-09-19 21:04:18 (GMT)