def validate_user(self):
user_id = access_control.get_current_user_id()
if not user_id:
self.redirect(users.create_login_url(self.request.uri))
return False
if not may_see_home_view(user_id):
self.response.write(
'User ' + user_id + ' is not allowed to access this page! <a href="' + users.create_logout_url(
'/argunit/home') + '">Logout</a>')
return False
return True
def validate_user(self):
user_id = access_control.get_current_user_id()
if not user_id:
self.redirect(users.create_login_url(self.request.uri))
return False
if not may_see_home_view(user_id):
self.response.write(
'User ' + user_id +
' is not allowed to access this page! <a href="' +
users.create_logout_url('/argunit/home') + '">Logout</a>')
return False
return True
def post(self):
if not self.validate_user():
return
user_id = access_control.get_current_user_id()
if self.request.path.endswith('nextdoc'):
json_req = json.loads(self.request.body)
self.respond_with_next_document(user_id, json_req)
elif self.request.path.endswith('previousdoc'):
json_req = json.loads(self.request.body)
self.respond_with_previous_document(user_id, json_req)
elif self.request.path.endswith('setapproval'):
json_req = json.loads(self.request.body)
annotator = json_req['annotator']
document = json_req['document']
isApproved = json_req['approved']
ManageDataHandler().setApproval(annotator, document, isApproved)
self.response.headers['Content-Type'] = 'application/json'
self.response.write(json.dumps({'message': "0"}))
def post(self):
if not self.validate_user():
return
user_id = access_control.get_current_user_id()
if self.request.path.endswith('nextdoc'):
json_req = json.loads(self.request.body)
self.respond_with_next_document(user_id, json_req)
elif self.request.path.endswith('previousdoc'):
json_req = json.loads(self.request.body)
self.respond_with_previous_document(user_id, json_req)
elif self.request.path.endswith('setapproval'):
json_req = json.loads(self.request.body)
annotator = json_req['annotator']
document = json_req['document']
isApproved = json_req['approved']
ManageDataHandler().setApproval(annotator, document, isApproved)
self.response.headers['Content-Type'] = 'application/json'
self.response.write(json.dumps({'message': "0"}))
def get(self):
if not self.validate_user():
return
user_id = access_control.get_current_user_id()
if self.request.path.endswith('/compare'):
user_data = load_user_data(user_id)
topics = user_data.selected_topics
documents_sorted_by_id = get_docs_in_topics_by_id(topics);
template_values = {'user': user_id,
'logout_url': users.create_logout_url('/argunit/'),
'all_views': access_control.get_view_ids(user_id),
'current_view': access_control.COMPARE_VIEW_ID}
if not documents_sorted_by_id:
template_values.update({'has_document': False,
'message': "No documents in collection"})
else:
doc_filename = self.request.get('doc', None)
doc = mark_as_current_doc(user_id, doc_filename, documents_sorted_by_id[0])
all_users = sorted(access_control.get_all_users())
print(all_users)
annotator_id_1 = self.request.get("annotator1")
if not annotator_id_1:
annotator_id_1 = all_users[0]
annotator_id_2 = self.request.get("annotator2")
if not annotator_id_2:
annotator_id_2 = all_users[1]
showAnnotator3 = self.request.get("showAnnotator3") != ''
annotator_id_3 = self.request.get("annotator3")
if not annotator_id_3:
annotator_id_3 = all_users[2]
opening_time = time.strftime("%H:%M:%S %Z", time.localtime())
current_topic = get_current_topic(user_id)
all_topics = get_all_topics()
annotation1 = self.load_annotations(annotator_id_1, doc)
annotation2 = self.load_annotations(annotator_id_2, doc)
annotation3 = self.load_annotations(annotator_id_3, doc)
arg_units1 = annotation1.arg_units if annotation1 else []
arg_units2 = annotation2.arg_units if annotation2 else []
arg_units3 = annotation3.arg_units if annotation3 else []
approved1 = annotation1.approved if annotation1 else False
approved2 = annotation2.approved if annotation2 else False
approved3 = annotation3.approved if annotation3 else False
notes1 = annotation1.notes if annotation1 else ""
notes2 = annotation2.notes if annotation2 else ""
notes3 = annotation3.notes if annotation3 else ""
template_values.update({'navigation_docs': documents_sorted_by_id,
'documents_sorted_by_id': documents_sorted_by_id,
'text': doc.text,
'doc_url': doc.url,
'doc_filename': doc.filename,
'num_sentences': doc.num_sentences,
'num_tokens': doc.num_tokens,
'time': opening_time,
'showAnnotator3': showAnnotator3,
'annotator1': annotator_id_1,
'annotator2': annotator_id_2,
'annotator3': annotator_id_3,
'approved1': approved1,
'approved2': approved2,
'approved3': approved3,
'arg_units1': json.dumps(arg_units1),
'arg_units2': json.dumps(arg_units2),
'arg_units3': json.dumps(arg_units3),
'notes1': notes1,
'notes2': notes2,
'notes3': notes3,
'current_topic': current_topic,
'all_topics': all_topics,
'all_users': all_users,
'has_document': True,
'message': ""})
template = JINJA_ENVIRONMENT.get_template('compare.html')
self.response.write(template.render(template_values))
elif self.request.path.endswith('/selecttopic'):
topics = self.request.get_all('topic', None)
annotator1 = self.request.get('annotator1');
annotator2 = self.request.get('annotator2');
annotator3 = self.request.get('annotator3');
set_as_current_topics(user_id, topics)
self.redirect('/argunit/compare' + "?annotator1="
+ annotator1 + "&annotator2=" + annotator2 + "&annotator3=" + annotator3)
def get(self):
if not self.validate_user():
return
if self.request.path.endswith('/annotate'):
user_id = access_control.get_current_user_id()
# Allow to take the role of another user
if self.request.get('user', None) in access_control.get_all_users():
user_id = self.request.get('user', None)
documents_sorted_by_id = get_docs_ordered_by_id(user_id);
processing_state = get_processing_state(user_id)
# Per user_id annotation statistics
num_annotated, num_approved = self.get_evaluation_stats()
template_values = {'user': user_id,
'logout_url': users.create_logout_url('/argunit/'),
'all_views': access_control.get_view_ids(user_id),
'current_view': access_control.ANNOTATE_VIEW_ID}
if not documents_sorted_by_id:
template_values.update({
'has_document': False,
'navigation_docs': [],
'num_annotated': num_annotated,
'num_approved': num_approved,
'message': "No documents to display."
})
else:
doc_filename = self.request.get('doc', None)
doc = mark_as_current_doc(user_id, doc_filename, documents_sorted_by_id[0])
topics = get_all_topics()
current_topic = get_current_topic(user_id)
opening_time = time.strftime("%H:%M:%S %Z", time.localtime())
docs_with_processing_state = [(d, processing_state[d.filename]) for d in
documents_sorted_by_id]
# Get arg_units to show
arg_units = []
relations = []
concepts = []
doc_approved = False
notes = ""
personal_annotations = DocumentAnnotation.all().filter('user_id =', user_id).filter(
'document =', doc.filename).get()
if personal_annotations:
doc_approved = personal_annotations.approved
notes = personal_annotations.notes
arg_units = [str(item) for item in personal_annotations.arg_units]
relations = [str(item) for item in personal_annotations.relations]
concepts = [str(item) for item in personal_annotations.concepts]
template_values.update({'navigation_docs': docs_with_processing_state,
'documents_sorted_by_id': documents_sorted_by_id,
'text': doc.text,
'doc_url': doc.url,
'doc_filename': doc.filename,
'doc_approved': doc_approved,
'num_sentences': doc.num_sentences,
'num_tokens': doc.num_tokens,
'time': opening_time,
'arg_units': json.dumps(arg_units),
'relations': json.dumps(relations),
'concepts': json.dumps(concepts),
'all_topics': topics,
'current_topic': current_topic,
'num_annotated': num_annotated,
'num_approved': num_approved,
'notes': notes,
'has_document': True,
'message': ""})
template = JINJA_ENVIRONMENT.get_template('annotate.html')
self.response.write(template.render(template_values))
elif self.request.path.endswith('/selecttopic'):
user_id = access_control.get_current_user_id()
topics = self.request.get_all('topic', None)
set_as_current_topics(user_id, topics)
self.redirect('/argunit/annotate')
def get(self):
user_id = access_control.get_current_user_id()
if not self.validate_user():
return
elif self.request.path.endswith('dump'):
self.dump_corpus()
elif self.request.path.endswith('dropall'):
self.drop_all_data()
self.redirect(
'%s?%s' % (self.base_path(), urllib.urlencode({"message": 'Dropped all data.'})))
elif self.request.path.endswith('dropanno'):
self.drop_all_annotations()
self.redirect('%s?%s' % (
self.base_path(), urllib.urlencode({"message": 'Dropped all annotations.'})))
elif self.request.path.endswith('loaddata'):
response_text = self.load_documents()
self.redirect(
'%s?%s' % (self.base_path(), urllib.urlencode({"verbatim_message": response_text})))
elif self.request.path.endswith('forceupdate'):
response_text = self.load_documents(force_update=True)
self.redirect(
'%s?%s' % (self.base_path(), urllib.urlencode({"verbatim_message": response_text})))
elif self.request.path.endswith('unapprove'):
annotator = self.request.get("annotator")
document = self.request.get("doc")
self.setApproval(annotator, document, False);
response_text = "Unapproved: %s:%s" % (annotator, document)
self.redirect(
'%s?%s' % (self.base_path(), urllib.urlencode({"message": response_text})))
elif self.request.path.endswith('/managedata'):
all_documents = [doc.filename for doc in Document.all()]
all_documents.sort()
all_users = access_control.get_all_users()
all_users.sort()
status_table = dict()
for user in all_users:
status_table[user] = dict()
for doc in all_documents:
anno = DocumentAnnotation.all().filter("user_id =", user).filter("document =",
doc).get()
if not anno:
status_table[user][doc] = UNPROCESSED
elif not anno.approved:
status_table[user][doc] = IN_PROGRESS
else:
status_table[user][doc] = COMPLETE
documents_per_line = 44
num_docs = len(all_documents)
num_lines = (num_docs + documents_per_line - 1) / documents_per_line
partitioned_docs = []
for i in range(0, num_lines):
partitioned_docs.append(all_documents[i * documents_per_line:min(num_docs, (
i + 1) * documents_per_line)])
message = self.request.get('message', "")
verbatim_message = self.request.get('verbatim_message', "")
metadata = CorpusMetadata.all().get()
segmenter = "unknown"
preprocessing_date = "unknown"
if metadata:
segmenter = metadata.segmenter
preprocessing_date = metadata.preprocessing_date
template_values = {'user': user_id,
'logout_url': users.create_logout_url('/argunit/'),
'all_views': access_control.get_view_ids(user_id),
'current_view': access_control.MANAGE_DATA_VIEW_ID,
'num_documents': len(all_documents),
'segmenter': segmenter,
'preprocessing_date': preprocessing_date,
'all_documents': all_documents,
'docs_per_line': documents_per_line,
'partitioned_docs': partitioned_docs,
'all_users': all_users,
'status_table': status_table,
'message': message,
'verbatim_message': verbatim_message}
template = JINJA_ENVIRONMENT.get_template('managedata.html')
self.response.write(template.render(template_values))
else:
self.redirect('/argunit/managedata')
def get(self):
if not self.validate_user():
return
if self.request.path.endswith('/annotate'):
user_id = access_control.get_current_user_id()
# Allow to take the role of another user
if self.request.get('user',
None) in access_control.get_all_users():
user_id = self.request.get('user', None)
documents_sorted_by_id = get_docs_ordered_by_id(user_id)
processing_state = get_processing_state(user_id)
# Per user_id annotation statistics
num_annotated, num_approved = self.get_evaluation_stats()
template_values = {
'user': user_id,
'logout_url': users.create_logout_url('/argunit/'),
'all_views': access_control.get_view_ids(user_id),
'current_view': access_control.ANNOTATE_VIEW_ID
}
if not documents_sorted_by_id:
template_values.update({
'has_document': False,
'navigation_docs': [],
'num_annotated': num_annotated,
'num_approved': num_approved,
'message': "No documents to display."
})
else:
doc_filename = self.request.get('doc', None)
doc = mark_as_current_doc(user_id, doc_filename,
documents_sorted_by_id[0])
topics = get_all_topics()
current_topic = get_current_topic(user_id)
opening_time = time.strftime("%H:%M:%S %Z", time.localtime())
docs_with_processing_state = [(d, processing_state[d.filename])
for d in documents_sorted_by_id]
# Get arg_units to show
arg_units = []
relations = []
concepts = []
doc_approved = False
notes = ""
personal_annotations = DocumentAnnotation.all().filter(
'user_id =', user_id).filter('document =',
doc.filename).get()
if personal_annotations:
doc_approved = personal_annotations.approved
notes = personal_annotations.notes
arg_units = [
str(item) for item in personal_annotations.arg_units
]
relations = [
str(item) for item in personal_annotations.relations
]
concepts = [
str(item) for item in personal_annotations.concepts
]
template_values.update({
'navigation_docs': docs_with_processing_state,
'documents_sorted_by_id': documents_sorted_by_id,
'text': doc.text,
'doc_url': doc.url,
'doc_filename': doc.filename,
'doc_approved': doc_approved,
'num_sentences': doc.num_sentences,
'num_tokens': doc.num_tokens,
'time': opening_time,
'arg_units': json.dumps(arg_units),
'relations': json.dumps(relations),
'concepts': json.dumps(concepts),
'all_topics': topics,
'current_topic': current_topic,
'num_annotated': num_annotated,
'num_approved': num_approved,
'notes': notes,
'has_document': True,
'message': ""
})
template = JINJA_ENVIRONMENT.get_template('annotate.html')
self.response.write(template.render(template_values))
elif self.request.path.endswith('/selecttopic'):
user_id = access_control.get_current_user_id()
topics = self.request.get_all('topic', None)
set_as_current_topics(user_id, topics)
self.redirect('/argunit/annotate')
def get(self):
user_id = access_control.get_current_user_id()
if not self.validate_user():
return
elif self.request.path.endswith('dump'):
self.dump_corpus()
elif self.request.path.endswith('dropall'):
self.drop_all_data()
self.redirect('%s?%s' %
(self.base_path(),
urllib.urlencode({"message": 'Dropped all data.'})))
elif self.request.path.endswith('dropanno'):
self.drop_all_annotations()
self.redirect(
'%s?%s' %
(self.base_path(),
urllib.urlencode({"message": 'Dropped all annotations.'})))
elif self.request.path.endswith('loaddata'):
response_text = self.load_documents()
self.redirect(
'%s?%s' %
(self.base_path(),
urllib.urlencode({"verbatim_message": response_text})))
elif self.request.path.endswith('forceupdate'):
response_text = self.load_documents(force_update=True)
self.redirect(
'%s?%s' %
(self.base_path(),
urllib.urlencode({"verbatim_message": response_text})))
elif self.request.path.endswith('unapprove'):
annotator = self.request.get("annotator")
document = self.request.get("doc")
self.setApproval(annotator, document, False)
response_text = "Unapproved: %s:%s" % (annotator, document)
self.redirect('%s?%s' %
(self.base_path(),
urllib.urlencode({"message": response_text})))
elif self.request.path.endswith('/managedata'):
all_documents = [doc.filename for doc in Document.all()]
all_documents.sort()
all_users = access_control.get_all_users()
all_users.sort()
status_table = dict()
for user in all_users:
status_table[user] = dict()
for doc in all_documents:
anno = DocumentAnnotation.all().filter(
"user_id =", user).filter("document =", doc).get()
if not anno:
status_table[user][doc] = UNPROCESSED
elif not anno.approved:
status_table[user][doc] = IN_PROGRESS
else:
status_table[user][doc] = COMPLETE
documents_per_line = 44
num_docs = len(all_documents)
num_lines = (num_docs + documents_per_line -
1) / documents_per_line
partitioned_docs = []
for i in range(0, num_lines):
partitioned_docs.append(all_documents[
i * documents_per_line:min(num_docs, (i + 1) *
documents_per_line)])
message = self.request.get('message', "")
verbatim_message = self.request.get('verbatim_message', "")
metadata = CorpusMetadata.all().get()
segmenter = "unknown"
preprocessing_date = "unknown"
if metadata:
segmenter = metadata.segmenter
preprocessing_date = metadata.preprocessing_date
template_values = {
'user': user_id,
'logout_url': users.create_logout_url('/argunit/'),
'all_views': access_control.get_view_ids(user_id),
'current_view': access_control.MANAGE_DATA_VIEW_ID,
'num_documents': len(all_documents),
'segmenter': segmenter,
'preprocessing_date': preprocessing_date,
'all_documents': all_documents,
'docs_per_line': documents_per_line,
'partitioned_docs': partitioned_docs,
'all_users': all_users,
'status_table': status_table,
'message': message,
'verbatim_message': verbatim_message
}
template = JINJA_ENVIRONMENT.get_template('managedata.html')
self.response.write(template.render(template_values))
else:
self.redirect('/argunit/managedata')
def get(self):
if not self.validate_user():
return
user_id = access_control.get_current_user_id()
if self.request.path.endswith('/compare'):
user_data = load_user_data(user_id)
topics = user_data.selected_topics
documents_sorted_by_id = get_docs_in_topics_by_id(topics)
template_values = {
'user': user_id,
'logout_url': users.create_logout_url('/argunit/'),
'all_views': access_control.get_view_ids(user_id),
'current_view': access_control.COMPARE_VIEW_ID
}
if not documents_sorted_by_id:
template_values.update({
'has_document': False,
'message': "No documents in collection"
})
else:
doc_filename = self.request.get('doc', None)
doc = mark_as_current_doc(user_id, doc_filename,
documents_sorted_by_id[0])
all_users = sorted(access_control.get_all_users())
print(all_users)
annotator_id_1 = self.request.get("annotator1")
if not annotator_id_1:
annotator_id_1 = all_users[0]
annotator_id_2 = self.request.get("annotator2")
if not annotator_id_2:
annotator_id_2 = all_users[1]
showAnnotator3 = self.request.get("showAnnotator3") != ''
annotator_id_3 = self.request.get("annotator3")
if not annotator_id_3:
annotator_id_3 = all_users[2]
opening_time = time.strftime("%H:%M:%S %Z", time.localtime())
current_topic = get_current_topic(user_id)
all_topics = get_all_topics()
annotation1 = self.load_annotations(annotator_id_1, doc)
annotation2 = self.load_annotations(annotator_id_2, doc)
annotation3 = self.load_annotations(annotator_id_3, doc)
arg_units1 = annotation1.arg_units if annotation1 else []
arg_units2 = annotation2.arg_units if annotation2 else []
arg_units3 = annotation3.arg_units if annotation3 else []
approved1 = annotation1.approved if annotation1 else False
approved2 = annotation2.approved if annotation2 else False
approved3 = annotation3.approved if annotation3 else False
notes1 = annotation1.notes if annotation1 else ""
notes2 = annotation2.notes if annotation2 else ""
notes3 = annotation3.notes if annotation3 else ""
template_values.update({
'navigation_docs': documents_sorted_by_id,
'documents_sorted_by_id': documents_sorted_by_id,
'text': doc.text,
'doc_url': doc.url,
'doc_filename': doc.filename,
'num_sentences': doc.num_sentences,
'num_tokens': doc.num_tokens,
'time': opening_time,
'showAnnotator3': showAnnotator3,
'annotator1': annotator_id_1,
'annotator2': annotator_id_2,
'annotator3': annotator_id_3,
'approved1': approved1,
'approved2': approved2,
'approved3': approved3,
'arg_units1': json.dumps(arg_units1),
'arg_units2': json.dumps(arg_units2),
'arg_units3': json.dumps(arg_units3),
'notes1': notes1,
'notes2': notes2,
'notes3': notes3,
'current_topic': current_topic,
'all_topics': all_topics,
'all_users': all_users,
'has_document': True,
'message': ""
})
template = JINJA_ENVIRONMENT.get_template('compare.html')
self.response.write(template.render(template_values))
elif self.request.path.endswith('/selecttopic'):
topics = self.request.get_all('topic', None)
annotator1 = self.request.get('annotator1')
annotator2 = self.request.get('annotator2')
annotator3 = self.request.get('annotator3')
set_as_current_topics(user_id, topics)
self.redirect('/argunit/compare' + "?annotator1=" + annotator1 +
"&annotator2=" + annotator2 + "&annotator3=" +
annotator3)