def test_ApostropheInTheBeginning(self):
r = get_surrounding("came", "ApostropheBeginningSentence.txt", "test", 32, 4, 4)
self.assertEqual(r, "sure enough in he came and made a great")
def test_ConnectedWordsSentence(self):
r = get_surrounding("to-do", "ConnectedWordsSentence.txt", "test", 55, 4, 4)
self.assertEqual(r, "and made a great to-do about the way the")
def test_ComplicatedSentence(self):
r = get_surrounding("said", "ComplicatedSentence.txt", "test", 44, 4, 4)
self.assertEqual(r, self.result_text)
def test_SemicolonSentence(self):
r = get_surrounding("try", "SemicolonSentence.txt", "test", 31, 4, 4)
self.assertEqual(r, "Wind had the first try and gathering up all")
def test_HyphenSentence(self):
r = get_surrounding("said", "HyphenSentence.txt", "test", 43, 4, 4)
self.assertEqual(r, self.result_text)
def test_EventAtEnd(self):
r = get_surrounding("so", "EasySentence.txt", "test", 66, 4, 4)
self.assertEqual(r, "said or rather shouted so")
def test_EventAtBeginning(self):
r = get_surrounding("Holding", "EasySentence.txt", "test", 0, 4, 4)
self.assertEqual(r, "Holding it between his finger")
def parse_XML(filename, dirname):
"""Takes xml filename and name of directory as arguments an returns all Text objects bundled in a Holder instance."""
tree = etree.parse(filename)
root = tree.getroot()
data = Holder()
for i, txt in enumerate(root.iterdescendants("file")):
# Create Text object
text = Text(i, txt.get("name"))
for j, ann in enumerate(txt.iterdescendants("annotator")):
# Create an Annotator object
annotator = Annotator(j, text, ann.get("id"))
# Create link from Text object
text.append_annotator(annotator)
for k, ev in enumerate(ann.iterdescendants("event")):
event_text = ev.get("text")
# Get the surrounding text and sentence for this event
sentence, num_words_as_event_before_event = get_sentence(event_text, text.name, dirname, int(ev.get("begin")))
surrounding = get_surrounding(event_text, text.name, dirname, int(ev.get("begin")), Event.surrounding_words_left, Event.surrounding_words_right)
pos_surrounding = get_surrounding(event_text, text.name, dirname, int(ev.get("begin")), Event.pos_surrounding_words_left, Event.pos_surrounding_words_right)
# Get the polarity of the event
polarity = get_polarity(surrounding)
# Get the modality of the event
modality = get_modality(surrounding)
# Create an Event object
event = Event(annotator, k, event_text, sentence, num_words_as_event_before_event, surrounding, pos_surrounding, polarity, modality, ev.get("begin"), ev.get("end"))
# Create link from Annotator object
annotator.events.append(event)
for k, tlink in enumerate(ann.iterdescendants("tlink")):
# Connect corresponding event objects to this relation object
# Doing this by going through all possible events
source = tlink[1]
begin = source.get("begin")
end = source.get("end")
# Search through the events of this annotator
source_event = None
for event in annotator.events:
if event.begin == int(begin) and event.end == int(end):
source_event = event
break
target = tlink[2]
begin = target.get("begin")
end = target.get("end")
target_event = None
for event in annotator.events:
if event.begin == int(begin) and event.end == int(end):
target_event = event
break
# Create a Relation object
temporal_relation = tlink[0].get("type")
relation = Relation(annotator, source_event, target_event, temporal_relation)
# Create link from Annotator object
annotator.relations.append(relation)
# Include text to data structure
data.textfiles.append(text)
# Create Text_structure objects, remove not needed relations and then all other possible relations
for text in data.textfiles:
for ann in text.annotator:
ann.create_text_structure()
ann.remove_none_relations()
ann.create_other_relations()
return data
def parse_XML(filename, dirname):
"""Takes xml filename and name of directory as arguments an returns all Text objects bundled in a Holder instance."""
tree = etree.parse(filename)
root = tree.getroot()
data = Holder()
for i, txt in enumerate(root.iterdescendants("file")):
# Create Text object
text = Text(i, txt.get("name"))
for j, ann in enumerate(txt.iterdescendants("annotator")):
# Create an Annotator object
annotator = Annotator(j, text, ann.get("id"))
# Create link from Text object
text.append_annotator(annotator)
for k, ev in enumerate(ann.iterdescendants("event")):
event_text = ev.get("text")
# Get the surrounding text and sentence for this event
sentence, num_words_as_event_before_event = get_sentence(
event_text, text.name, dirname, int(ev.get("begin")))
surrounding = get_surrounding(event_text, text.name, dirname,
int(ev.get("begin")),
Event.surrounding_words_left,
Event.surrounding_words_right)
pos_surrounding = get_surrounding(
event_text, text.name, dirname, int(ev.get("begin")),
Event.pos_surrounding_words_left,
Event.pos_surrounding_words_right)
# Get the polarity of the event
polarity = get_polarity(surrounding)
# Get the modality of the event
modality = get_modality(surrounding)
# Create an Event object
event = Event(annotator, k, event_text, sentence,
num_words_as_event_before_event, surrounding,
pos_surrounding, polarity, modality,
ev.get("begin"), ev.get("end"))
# Create link from Annotator object
annotator.events.append(event)
for k, tlink in enumerate(ann.iterdescendants("tlink")):
# Connect corresponding event objects to this relation object
# Doing this by going through all possible events
source = tlink[1]
begin = source.get("begin")
end = source.get("end")
# Search through the events of this annotator
source_event = None
for event in annotator.events:
if event.begin == int(begin) and event.end == int(end):
source_event = event
break
target = tlink[2]
begin = target.get("begin")
end = target.get("end")
target_event = None
for event in annotator.events:
if event.begin == int(begin) and event.end == int(end):
target_event = event
break
# Create a Relation object
temporal_relation = tlink[0].get("type")
relation = Relation(annotator, source_event, target_event,
temporal_relation)
# Create link from Annotator object
annotator.relations.append(relation)
# Include text to data structure
data.textfiles.append(text)
# Create Text_structure objects, remove not needed relations and then all other possible relations
for text in data.textfiles:
for ann in text.annotator:
ann.create_text_structure()
ann.remove_none_relations()
ann.create_other_relations()
return data