The Berkeley parser wrapper has the following limitations:
1- The Berkeley parser wrapper needs the text be tokenized and pos tagged before parsing.
2- The parser does not parse some sentences properly. Specially sentences with tokens needed to be normalized to the Penn Treebank convention (e.g. '(' should be converted to '-LRB-' before the parsing step).
The Berkeley parser wrapper has the following limitations: 1- The Berkeley parser wrapper needs the text be tokenized and pos tagged before parsing. 2- The parser does not parse some sentences properly. Specially sentences with tokens needed to be normalized to the Penn Treebank convention (e.g. '(' should be converted to '-LRB-' before the parsing step).