def decode_pred(elem):
if elem.tag == 'gpred':
return Pred.grammarpred(elem.text)
elif elem.tag == 'realpred':
return Pred.realpred(elem.get('lemma'),
elem.get('pos'),
elem.get('sense'))
def _decode_pred(elem):
# <!ELEMENT realpred EMPTY>
# <!ATTLIST realpred
# lemma CDATA #REQUIRED
# pos (v|n|j|r|p|q|c|x|u|a|s) #REQUIRED
# sense CDATA #IMPLIED >
# <!ELEMENT gpred (#PCDATA)>
if elem.tag == 'gpred':
return Pred.grammarpred(elem.text)
elif elem.tag == 'realpred':
return Pred.realpred(elem.get('lemma'), elem.get('pos'),
elem.get('sense'))
def _decode_pred(elem):
# <!ELEMENT realpred EMPTY>
# <!ATTLIST realpred
# lemma CDATA #REQUIRED
# pos (v|n|j|r|p|q|c|x|u|a|s) #REQUIRED
# sense CDATA #IMPLIED >
# <!ELEMENT gpred (#PCDATA)>
if elem.tag == 'gpred':
return Pred.grammarpred(elem.text)
elif elem.tag == 'realpred':
return Pred.realpred(elem.get('lemma'),
elem.get('pos'),
elem.get('sense'))
def strip_category(cat):
if cat.endswith("u_unknown"):
lemma, pos_and_sense = cat.rsplit("/", 1)
pos_part, sense_part = pos_and_sense.split("_", 1)
lemma_part = "X"
else:
pred_obj = Pred.stringpred(cat)
lemma_part = "X" if cat.startswith("_") else pred_obj.lemma
pos_part = str(pred_obj.pos)
sense_part = str(pred_obj.sense)
return lemma_part + "_" + pos_part + "_" + sense_part
def _read_node(tokens):
if not tokens or tokens[0][0] not in {'string', 'symbol', 'nodeid'}:
return None
# A node can be a pred, a nodeid, or both (in that order). This
# means two 'if's, not 'if-else'.
mtype, mtext = tokens.popleft()
pred = nodeid = None
if mtype in ('string', 'symbol'):
if mtext == TOP or mtext == STAR:
pred = mtext
else:
pred = Pred.stringpred(mtext)
if tokens and tokens[0][0] == 'nodeid':
mtype, mtext = tokens.popleft()
if mtype == 'nodeid':
nodeid = int(mtext[1:]) # get rid of the initial # character
context = _read_context(tokens)
links = _read_links(tokens)
return XmrsPathNode(nodeid, pred, context=context, links=links)
def _read_node(tokens):
if not tokens or tokens[0][0] not in {'string', 'symbol', 'nodeid'}:
return None
# A node can be a pred, a nodeid, or both (in that order). This
# means two 'if's, not 'if-else'.
mtype, mtext = tokens.popleft()
pred = nodeid = None
if mtype in ('string', 'symbol'):
if mtext == TOP or mtext == STAR:
pred = mtext
else:
pred = Pred.stringpred(mtext)
if tokens and tokens[0][0] == 'nodeid':
mtype, mtext = tokens.popleft()
if mtype == 'nodeid':
nodeid = int(mtext[1:]) # get rid of the initial # character
context = _read_context(tokens)
links = _read_links(tokens)
return XmrsPathNode(
nodeid,
pred,
context=context,
links=links
)
def decode_pred(elem):
if elem.tag == 'gpred':
return Pred.grammarpred(elem.text)
elif elem.tag == 'realpred':
return Pred.realpred(elem.get('lemma'), elem.get('pos'),
elem.get('sense'))