Jakob Lykke Andersen

Some time ago I looked into regenerating the parser with generateDS both because the Doxygen XML has changed with newer versions, but also because it seems to use lxml now....

I'm not sure there is any functionality to do what you would like at the moment. There are two entities in play here: - the current scope of the C++...

Indeed, except in the second example Sphinx would not declare ``my_namespace::Foo`` but just ``Foo``, so a ``.. cpp:namespace:: my_namespace`` is needed in addition to the ``doxygenprefix``.

A lot has happened since that issue, and I wouldn't call the connection fragile nowadays. Many issues are not necessarily in the Breathe/Sphinx connection but rather the Doxygen/Breathe connection, and...

> I have the same issue with a macro declaration. (Invalid C++ declaration) Any workaround available yet? This issue it not really about a macro declaration, but rather a declaration...

This would be a very nice feature, and I have wanted to implement something like it in Sphinx for quite some time. I don't know much about the internals of...

I think we need two related mechanisms on the Sphinx side: - The alias role/directive which in some form renders a declaration without redeclaring it. My current thought is that...

Thanks for the report. As you indicate, the problem is that Breathe does not know about Java, so the declarations are passed to Sphinx as C++, which Sphinx rightfully complains...

The main issue: I tried looking a bit into this, and it seems that those entries are directly from the Doxygen XML. With a small test project with just ``"pke":...

The "P.S." issue: The ``.. cpp:type::`` is for declaring a type alias / typedef in the documentation. While you may repeat them in real code, you can't in Sphinx, as...