def claims_from_html(content):
text = ht.html_to_text(nt.convert_entities(content))
return claims_from_body(text)
def claims_from_html(content): text = ht.html_to_text(nt.convert_entities(content)) return claims_from_body(text)
def bodys_from_tab_file(f):
for line in f:
row = line.strip().split("\t")
if len(row) > 3:
yield ht.html_to_text(nt.convert_entities(row[3]))
def bodys_from_tab_file(f):
for line in f:
row = line.strip().split("\t")
if len(row) > 3:
yield ht.html_to_text(nt.convert_entities(row[3]))
