def _is_leaf(tree: ParentedTree):
"""
Checks whether the given tree is a leaf.
:param tree: a ParentedTree instance
:return: true if it is a leaf
"""
return tree.height() == 2
def traverse_and_store(self, tree: ParentedTree,
parse_tree_stored: List[Dict]):
label = tree.label()
words = [x.split('_')[0] for x in tree.leaves()]
indices = [int(x.split('_')[-1]) for x in tree.leaves()]
ngram_info = len(words)
words = " ".join(words)
if tree.height() > self.TREE_HEIGHT and ngram_info < self.NGRAM_LIMIT:
parse_tree_stored.append({
'phrase_label': label,
'phrase': words,
'ngram': ngram_info,
'indices': indices
})
for subtree in tree:
if type(subtree) == ParentedTree:
self.traverse_and_store(tree=subtree,
parse_tree_stored=parse_tree_stored)
return parse_tree_stored
