port) + ")"
doc = CorpusDocumentX(docname)
processed_doc = False
for sentence in doc.sentences():
words = " ".join([x.text for x in sentence])
process_sentence = False
for x in sentence:
if not (ns('dcoi') + 'pos' in x.attrib or ns('dcoi') + 'lemma'
in x.attrib or 'pos' in x.attrib or 'lemma' in x.attrib):
process_sentence = True
if process_sentence:
processed_doc = True
for i, (word, lemma, morph, pos) in enumerate(
frogclient.process_aligned(words, 'iso-8859-15')):
if word:
try:
word_id = sentence[i].attrib[ns('xml') + 'id']
except KeyError:
print >> sys.stderr, "ERROR: Unable to extract ID attribute!"
break
except IndexError:
print >> sys.stderr, "ERROR: words out of sync in " + sentence.attrib[
ns('xml') + 'id'] + ': Unable to resolve word ' + str(
i + 1) + ': ' + word.encode(
'utf-8') + '. Source has ' + str(
len(sentence)) + ' words.'
break
if pos:
doc[word_id].attrib['pos'] = pos
