14.01.2011, 21:24, "Benjamin Poulain" <[email protected]>:
> On 01/14/2011 06:06 PM, ext Konstantin Tokarev wrote:
>
>>  I've found that QtWebKit does not actually depend on QtScript module (at 
>> least if configured without
>>  V8 support).
>>
>>  The only thing actually used is enum QScriptEngine::ValueOwnership. If its 
>> definition is included into
>>  Source/WebCore/bridge/qt/qt_instance.h and WebKit/qt/Api/qwebframe.h, 
>> QtWebKit compiles fine if
>>  QtScript module is disabled in Qt configuration.
>>
>>  Will it be OK if I add new header with the next code
>>
>>  #ifndef QT_NO_SCRIPT
>>  #include<QtScript/qscriptengine.h>
>>  #else
>>  QT_BEGIN_NAMESPACE
>>  namespace QScriptEngine {
>>       enum ValueOwnership {
>>           QtOwnership,
>>           ScriptOwnership,
>>           AutoOwnership
>>       };
>>  }
>>  QT_END_NAMESPACE
>>  #endif
>>
>>  This header needs to be "public" because it's needed for qwebframe.h
>
> Long term QtWebKit will depends on QtScript for the bridge. To avoid
> inconsistencies and regressions, I think it would be more reasonable to
> guard the functions of the bridge with #ifndef QT_NO_SCRIPT.

Do you mean it's OK to disable whole class QtInstance? (AFAIU, its main 
method is getQtInstance, and from your words it should be disabled)

>
> cheers,
> Benjamin
> _______________________________________________
> webkit-qt mailing list
> [email protected]
> http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt

-- 
Regards,
Konstantin
_______________________________________________
webkit-qt mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt

Reply via email to