A simple pcfg parser in python, using these sources as a guide
https://en.wikipedia.org/wiki/Cyk_algorithm https://en.wikipedia.org/wiki/Inside%E2%80%93outside_algorithm https://courses.cs.washington.edu/courses/cse590a/09wi/pcfg.pdf
A better counts source than the penn treebank sample included with nltk is needed. Or atleast a counts source that includes RARE tags.
Future steps to try:
implement this variation: http://nlp.stanford.edu/~manning/papers/unlexicalized-parsing.pdf
and this enhancement: http://www.socher.org/uploads/Main/SocherBauerManningNg_ACL2013.pdf
from the stanford parser