[PyQt] Overloaded functions with same Python signature

[Camden Mannett]

/Constrained/ is certainly correct, presumably SIP tried to convert
the enums to ints as there is already a Qt overload for that.  I
thought that /Out/ would be more appropriate as the right streaming
operator doesn't read the enum, but SIP just reported the original
error if /In/ wasn't present - so I settled on /Constrained, In, Out/.

The resulting sip***cmodule.cpp  __rshift__ method looks correct:

QDataStream* a0;
Sy_version* a1;

if (sipParsePair(&sipParseErr, sipArg0, sipArg1, "J9J9",
sipType_QDataStream, &a0, sipType_Sy_version, &a1))
{
    QDataStream*sipRes;
     try
     {
         sipRes = &(*a0 >> *a1);
      }

On 3 July 2014 13:54, Phil Thompson <phil at riverbankcomputing.com> wrote:
> On 03/07/2014 1:19 pm, Camden Mannett wrote:
>>
>> SIP/PyQt5 is reporting an error during the parsing of my sip file:
>>
>> __rshift__() has overloaded functions with the same Python signature
>>
>> The offending functions are:
>>
>> QDataStream& operator<<( QDataStream& stream,
>>                          const Sy_loggerInterface::DebugLevel& obj );
>> QDataStream& operator>>( QDataStream& stream,
>>                          Sy_loggerInterface::DebugLevel& obj );
>>
>> QDataStream& operator<<( QDataStream& stream,
>>                          const Sy_loggerInterface::RotationMetric& obj );
>> QDataStream& operator>>( QDataStream& stream,
>>                          Sy_loggerInterface::RotationMetric& obj );
>>
>> Sy_loggerInterface::DebugLevel and Sy_loggerInterface::RotationMetric
>> are enums, so they clearly don't have the same signature in C++. But I
>> don't know what the created Python signature would look like.
>>
>> Also note the left shift operators do not cause an error when I
>> comment out one of the right shift ones, and simply changing the names
>> of the parameters has no effect (as you would expect). Does anyone
>> know what I can do to fix this error?
>
>
> The only similar thing in PyQt applies /Constrained, In/ to the argument.
> However I'm not convinced it will do the right thing.
>
> Phil
> _______________________________________________
> PyQt mailing list    PyQt at riverbankcomputing.com
> http://www.riverbankcomputing.com/mailman/listinfo/pyqt