def extract(self, r):
content = text.strip_enclosed_carrots(r["title"]) + " " + text.strip_enclosed_carrots(r["abstract"])
ctokens = content.split()
phrases = []
current = []
for ct in ctokens:
if text.is_capitalized(ct):
current += filter(lambda w: w not in self._querystops, text.tokenize(ct))
elif len(current) > 0:
if len(current) >= self._pl:
phrases.append(" ".join(current))
current = []
return phrases
def __init__(self, query, pl=1): self._querystops = set(text.tokenize(query)) self._pl = calc_pl(query)
