On Mon, 7 Mar 2011, Bill Janssen wrote:
Andi Vajda <va...@apache.org> wrote:
On Thu, 3 Mar 2011, Andi Vajda wrote:
On Mar 3, 2011, at 22:09, Bill Janssen <jans...@parc.com> wrote:
Here's one of the generated lines that's causing me grief.
DECLARE_TYPE(RankIterator, t_RankIterator, ::java::lang::Object, RankIterator,
t_RankIterator_init_, PyObject_SelfIter, ((PyObject *(*)(t_RankIterator *))
get_next<t_RankIterator,t_JArray< jint >,JArray< jint >>),
Ah yes, that's invalid c++. Nested generics need to insert a space between '>'.
Otherwise, the c++ parser gets it as the bit shifting operator, believe it or not.
Should be easy enough to fix in jcc.
Fixed in trunk/branch_3x rev 1077828.
Yeah, but not for me. After installing the latest JCC, I get this:
So you found not one bug but two. The first one (about >>) is fixed, as you
can see in the new error message.
build/_PPD/__wrap03__.cpp:15149: error: ?t_JArray? was not declared in this
scope
build/_PPD/__wrap03__.cpp:15149: error: parse error in template argument list
build/_PPD/__wrap03__.cpp:15149: error: insufficient contextual information to
determine type
build/_PPD/__wrap03__.cpp:15149: error: expected primary-expression before ?>?
token
build/_PPD/__wrap03__.cpp:15149: error: expected primary-expression before ?)?
token
build/_PPD/__wrap03__.cpp:15149: error: too many initializers for ?PyTypeObject?
And the line it's complaining about is now:
DECLARE_TYPE(RankIterator, t_RankIterator, ::java::lang::Object, RankIterator,
t_RankIterator_init_, PyObject_SelfIter, ((PyObject *(*)(t_RankIterator *)) get_next<
t_RankIterator,t_JArray< jint >,JArray< jint > >), t_RankIterator__fields_, 0, 0);
Where does t_JArray get defined? I can't find it.
I'm not sure there is one. If you can provide me with a piece of Java to
reproduce this, I can fix it faster.
Andi..
