On Tue, Jun 7, 2011 at 12:25 PM, Gonzalo Tornaria <torna...@math.utexas.edu> wrote: > In sage/structure/parent_base.pyx there is a function > "check_old_coerce" with the following definition: > > cdef inline check_old_coerce(parent.Parent p): > if p._element_constructor is not None: > raise RuntimeError, "%s still using old coercion framework" % p > > > This function is called from a few methods in class ParentWithBase > (e.g. from _richcmp, from _coerce_c_impl, and from base_extend). > > In my class (which inherits from ParentWithBase) I've defined > _element_constructor_ instead, but the init for ParentWithBase insist > on copying this method onto _element_constructor -- thus the > "check_old_coerce" function always fails... It's impossible to > avoid!!! > > What is the base class that one is supposed to use instead of ParentWithBase?
Just Parent, if one can. > The only reference for the coercion model that I was able to find is > http://www.sagemath.org/doc/reference/coercion.html#example, which > even shows an example, but the example inherits from Ring, which > ultimately inherits from ParentWithBase anyway. What this is really telling you is that you're using an old coercon method from a class that's trying to do new-style coercion. (I was hoping to have all this ripped out years ago, but it lingers on...) In particular, _coerce_c_impl should never be called, nor should _richcmp or the generic base_extend. A full traceback would help see what the problem is. - Robert -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
