def test_bamboo_service(self):
# comment out when we can test or mock it differently
raise SkipTest
service_url = 'http://bamboo.io/'
service_name = 'bamboo'
xml_submission1 = os.path.join(self.this_directory, u'fixtures',
u'dhisform_submission1.xml')
xml_submission2 = os.path.join(self.this_directory, u'fixtures',
u'dhisform_submission2.xml')
xml_submission3 = os.path.join(self.this_directory, u'fixtures',
u'dhisform_submission3.xml')
# make sure xform doesnt have a bamboo dataset
self.xform.bamboo_dataset = ''
self.xform.save()
# make a first submission without the service
self._make_submission(xml_submission1)
self.assertEqual(self.response.status_code, 201)
# add rest service AFTER 1st submission
self._add_rest_service(service_url, service_name)
# submit another one.
self._make_submission(xml_submission2)
self.assertEqual(self.response.status_code, 201)
self.wait(5)
# it should have created the whole dataset
xform = XForm.objects.get(id=self.xform.id)
self.assertTrue(xform.bamboo_dataset != ''
and xform.bamboo_dataset is not None)
dataset = Dataset(connection=Connection(service_url),
dataset_id=xform.bamboo_dataset)
self.assertEqual(dataset.get_info()['num_rows'], 2)
# submit a third one. check that we have 3 records
self._make_submission(xml_submission3)
self.assertEqual(self.response.status_code, 201)
self.wait(5)
self.assertEqual(dataset.get_info()['num_rows'], 3)
# test regeneration
dsi = dataset.get_info()
regen_url = reverse(link_to_bamboo,
kwargs={
'username': self.user.username,
'id_string': self.xform.id_string
})
response = self.client.post(regen_url, {})
# deleting DS redirects to profile page
self.assertEqual(response.status_code, 302)
self.wait(5)
xform = XForm.objects.get(id=self.xform.id)
self.assertTrue(xform.bamboo_dataset)
dataset = Dataset(connection=Connection(service_url),
dataset_id=xform.bamboo_dataset)
new_dsi = dataset.get_info()
self.assertEqual(new_dsi['num_rows'], dsi['num_rows'])
self.assertNotEqual(new_dsi['id'], dsi['id'])
def test_merge_default_connection(self):
dataset = Dataset(path=self.CSV_FILE,
connection=self.default_connection)
other_dataset = Dataset(path=self.CSV_FILE,
connection=self.default_connection)
result = Dataset.merge([dataset, other_dataset])
self.assertTrue(isinstance(result, Dataset))
self._cleanup(dataset)
self._cleanup(other_dataset)
self._cleanup(result)
def test_join_default_connection(self):
dataset = Dataset(path=self.CSV_FILE,
connection=self.default_connection)
aux_dataset = Dataset(path=self.AUX_CSV_FILE,
connection=self.default_connection)
self.wait()
result = Dataset.join(dataset, aux_dataset, 'food_type')
self.wait()
self.assertTrue(isinstance(result, Dataset))
self._cleanup(dataset)
self._cleanup(aux_dataset)
self._cleanup(result)
def test_create_dataset_from_schema(self):
dataset = Dataset(schema_path=self.SCHEMA_FILE,
connection=self.connection)
self.assertTrue(dataset.id is not None)
self._cleanup(dataset)
# schema string
schema_str = open(self.SCHEMA_FILE).read()
dataset = Dataset(schema_content=schema_str,
connection=self.connection)
self.assertTrue(dataset.id is not None)
self._cleanup(dataset)
def run_test_suite(dataset_file_path_list):
print "running test suite for %s" % " ".join(dataset_file_path_list)
alldata = []
for dataset_name in dataset_file_path_list:
d = {}
d['hostname'] = os.uname()[1]
d['bamboo_url'] = URL
d['unix_time'] = time.time()
conn = Connection(url=URL)
d['commit'] = conn.version['commit']
d['branch'] = conn.version['branch']
dataset = Dataset(connection=conn, path=dataset_name)
d['import_time'] = time_till_import_is_finished(dataset)
info = dataset.get_info()
d['row'] = info['num_rows']
d['col'] = info['num_columns']
d['add_1_calculations_time'] = time_to_add_1_calculations(dataset)
d['add_5_calculations_1by1_time'] = time_to_add_5_calculations_1by1(
dataset)
d['add_5_calculations_batch_time'] = time_to_add_5_calculations_batch(
dataset)
d['update_1_time'] = time_to_add_1_update(dataset)
d['update_5_1by1_time'] = time_to_add_5_update_1by1(dataset)
d['update_5_batch_time'] = time_to_add_5_update_batch(dataset)
dataset.delete()
alldata.append(d)
return alldata
def update_sources(site):
sources = 'sources.json'
sources_dir = os.path.join(os.path.dirname(__file__), 'data')
if isinstance(site, basestring):
sources = os.path.join(os.path.dirname(__file__), 'data', site.lower(),
'sources.json')
sources_dir = os.path.join(sources_dir, site.lower())
else:
sources = os.path.join(os.path.dirname(__file__), 'sources.json')
if not os.path.exists(sources):
raise Exception(u"Please define a sources.json.")
f = open(sources)
sources_dict = json.loads(f.read())
f.close()
assert 'bamboo_server' in sources_dict
assert 'sources' in sources_dict
connection = Connection(sources_dict['bamboo_server'])
for k, v in sources_dict['sources'].iteritems():
if v == "":
path = os.path.join(sources_dir, k)
if not os.path.exists(path):
raise Exception(u"%s does not exist," % path)
try:
dataset = Dataset(path=path,
connection=connection,
na_values=["---", "None"],
data_format='csv')
except Exception, e:
print u"Exception: Publishing %s failed!\n\t%s" % (k, e)
else:
sources_dict['sources'][k] = dataset.id
def test_create_dataset_from_schema_with_data(self):
# schema + JSON data
dataset = Dataset(path=self.JSON_FILE,
data_format='json',
schema_path=self.SCHEMA_FILE,
connection=self.connection)
self.assertTrue(dataset.id is not None)
self._cleanup(dataset)
# schema + CSV data
dataset = Dataset(path=self.CSV_FILE,
data_format='csv',
schema_path=self.SCHEMA_FILE,
connection=self.connection)
self.assertTrue(dataset.id is not None)
self._cleanup(dataset)
def send(self, url, parsed_instance):
xform = parsed_instance.instance.xform
rows = [parsed_instance.to_dict_for_mongo()]
# prefix meta columns names for bamboo
prefix = (u'%(id_string)s_%(id)s' % {
'id_string': xform.id_string,
'id': xform.id
})
for row in rows:
for col, value in row.items():
if col.startswith('_') or col.startswith('meta_') \
or col.startswith('meta/'):
new_col = (u'%(prefix)s%(col)s' % {
'prefix': prefix,
'col': col
})
row.update({new_col: value})
del (row[col])
# create dataset on bamboo first (including current submission)
if not xform.bamboo_dataset:
dataset_id = get_new_bamboo_dataset(xform, force_last=True)
xform.bamboo_dataset = dataset_id
xform.save()
else:
dataset = Dataset(connection=Connection(url=get_bamboo_url(xform)),
dataset_id=xform.bamboo_dataset)
dataset.update_data(rows=rows)
def test_merge(self):
# already have one dataset in self.dataset
dataset = Dataset(path=self.CSV_FILE, connection=self.connection)
result = Dataset.merge([self.dataset, dataset],
connection=self.connection)
self.assertTrue(isinstance(result, Dataset))
self._cleanup(dataset)
self._cleanup(result)
def delete_bamboo_dataset(xform):
if not xform.bamboo_dataset:
return False
try:
dataset = Dataset(connection=Connection(url=get_bamboo_url(xform)),
dataset_id=xform.bamboo_dataset)
return dataset.delete()
except ErrorParsingBambooData:
return False
def test_na_values(self):
dataset = Dataset(path=self.CSV_FILE,
connection=self.connection,
na_values=['n/a'])
self.wait()
first_row = dataset.get_data(query={
'food_type': 'street_meat',
'amount': 2,
'rating': 'delectible',
'risk_factor': 'low_risk'
},
limit=1)[-1]
self.assertEqual(first_row.get('comments'), 'null')
self._cleanup(dataset)
def get_new_bamboo_dataset(xform, force_last=False):
dataset_id = u''
try:
content_data = get_csv_data(xform, force_last=force_last)
dataset = Dataset(connection=Connection(url=get_bamboo_url(xform)),
content=content_data,
na_values=['n/a'])
except (ErrorParsingBambooData, NoRecordsFoundError):
return dataset_id
if dataset.id:
return dataset.id
return dataset_id
def _get_sum(self, key, value, period):
sum_value = 0
for v in value[key]:
if 'aggregations' in v:
sum_value += self._get_aggregate('aggregations', v, period)
continue
dataset_id = v['dataset_id']
# dataset_id form sources.json is most recent
if dataset_id != self._sources[v['source']]\
and self._sources[v['source']] != "":
dataset_id = self._sources[v['source']]
dataset = Dataset(
dataset_id=dataset_id, connection=self.connection)
params = {}
if 'calculation' in v:
# check or create calculations
if isinstance(v['calculation'], list):
for calculation in v['calculation']:
self._add_calculation(calculation, dataset, period)
if isinstance(v['calculation'], dict):
self._add_calculation(v['calculation'], dataset, period)
if 'query' in v:
query_string = json.dumps(v['query'])
template = env.from_string(query_string)
query_string = template.render(period=period)
v['query'] = json.loads(query_string)
params['query'] = v['query']
if 'count' in v and 'query' in v:
params['count'] = v['count']
if 'distinct' in v:
params['distinct'] = v['distinct']
val = dataset.get_data(**params)
if isinstance(val, dict):
raise Exception("Bamboo Error: %s" % val)
sum_value += val
return sum_value
def test_create_dataset_from_url(self):
dataset = Dataset(
url='http://formhub.org/mberg/forms/good_eats/data.csv',
connection=self.connection)
self.assertTrue(self.dataset.id is not None)
self._cleanup(dataset)
def test_create_dataset_bad_data_format(self):
with self.assertRaises(PyBambooException):
Dataset(path=self.CSV_FILE,
data_format='BAD',
connection=self.connection)
def test_create_dataset_no_info(self):
with self.assertRaises(PyBambooException):
Dataset()
def test_create_dataset_default_connection(self):
dataset = Dataset(path=self.CSV_FILE,
connection=self.default_connection)
self._cleanup(dataset)
def _get_aggregate(self, key, value, period):
sum_value = 0
for v in value[key]:
dataset_id = v['dataset_id']
# dataset_id form sources.json is most recent
if dataset_id != self._sources[v['source']]\
and self._sources[v['source']] != "":
dataset_id = self._sources[v['source']]
dataset = Dataset(
dataset_id=dataset_id, connection=self.connection)
params = {}
if 'calculation' in v:
# check or create calculations
if isinstance(v['calculation'], list):
for calculation in v['calculation']:
self._add_calculation(calculation, dataset, period)
if isinstance(v['calculation'], dict):
self._add_calculation(v['calculation'], dataset, period)
if 'query' in v:
query_string = json.dumps(v['query'])
template = env.from_string(query_string)
query_string = template.render(period=period)
v['query'] = json.loads(query_string)
params['query'] = v['query']
# if 'count' in v and 'query' in v:
# params['count'] = v['count']
if 'distinct' in v:
params['distinct'] = v['distinct']
data = dataset.get_data(format='csv', **params)
if data.strip() == '':
# no data to create a dataset - skip
continue
# create a aggregate dataset
aggr_dataset = Dataset(
content=data,
data_format='csv',
connection=self.connection)
if 'aggregate' in v:
# check or create calculations
if isinstance(v['aggregate'], list):
for calculation in v['aggregate']:
calc = aggr_dataset.add_calculation(
name=calculation['name'],
formula=calculation['formula']
)
if calc:
aggr_ds = aggr_dataset.get_aggregations()['']
k = aggr_ds.get_data()
val = k[0][calculation['name']]
if isinstance(val, basestring):
raise ValueError("Dataset %s return %s"
% (aggr_ds.id, val))
sum_value += val
aggr_ds.delete()
if isinstance(v['aggregate'], dict):
calculation = v['aggregate']
calc = aggr_dataset.add_calculation(
name=calculation['name'],
formula=calculation['formula']
)
if calc:
aggr_ds = aggr_dataset.get_aggregations()['']
k = aggr_ds.get_data()
val = k[0][calculation['name']]
if isinstance(val, basestring):
raise ValueError("Dataset %s return %s"
% (aggr_ds.id, val))
sum_value += val
aggr_ds.delete()
aggr_dataset.delete()
return sum_value
def main():
dataset_url = "%sdatasets/%s.csv" % (BAMBOO_DEV_URL, args.dataset)
dataset = Dataset(url=dataset_url)
print dataset.id
# get state of current datasets
with open(bamboo_id_file) as f:
bamboo_ids = json.loads(f.read())
if not bamboo_ids:
print '"%s" not found: exiting' % bamboo_id_file
sys.exit(0)
print 'current dataset status:'
print json.dumps(bamboo_ids, indent=4, sort_keys=True)
# upload originals
for sector in bamboo_ids.keys():
for name, id in bamboo_ids[sector]['originals'].iteritems():
if not id:
print 'dataset: %s not uploaded, uploading %s.csv' % (name, name)
dataset = Dataset(connection=connection,
path='csvs/originals/%s.csv' % name)
state = dataset.get_info()['state']
while state != 'ready':
time.sleep(1)
state = dataset.get_info()['state']
print state
bamboo_ids[sector]['originals'][name] = dataset.id
with open(bamboo_id_file, 'wb') as f:
f.write(json.dumps(bamboo_ids))
# merge originals
for sector in bamboo_ids.keys():
if not bamboo_ids[sector]['merged']:
print 'no merged dataset for sector: %s' % sector
datasets = [
Dataset(connection=connection, dataset_id=id)
def test_create_dataset_from_json(self):
dataset = Dataset(path=self.JSON_FILE,
data_format='json',
connection=self.connection)
self.assertTrue(dataset.id is not None)
self._cleanup(dataset)
def _create_aux_dataset_from_file(self):
self.aux_dataset = Dataset(path=self.AUX_CSV_FILE,
connection=self.connection)
self.wait()
print 'Could not load %s' % BAMBOO_HASH_FILE
sys.exit(1)
# update the datasets
hash_updates = dict()
for name, content in bamboo_hash.iteritems():
filename = content['filename']
bamboo_id = content['bamboo_id']
sector = content.get('sector')
file_path = 'data/' + filename
print '%s -> %s' % (filename, bamboo_id)
if bamboo_id:
print '%s has bamboo id: %s. Updating bamboo dataset.' %\
(name, bamboo_id)
try:
dataset = Dataset(dataset_id=bamboo_id)
dataset.remove_calculation('sector')
dataset.reset(path=file_path)
if sector:
formula = '"%s"' % sector
print 'Adding column for sector: %s, formula: %s' %\
(sector, formula)
result = dataset.add_calculation('sector', formula)
if result:
print 'Calculation added successfully!'
else:
print 'Problem adding calculation!'
except PyBambooException:
print 'Error creating dataset for file: %s' % filename
else:
print '%s has no bamboo id. Adding file to bamboo.' % name
def test_merge_fail(self):
other_dataset = Dataset('12345', connection=self.connection)
result = Dataset.merge([self.dataset, other_dataset],
connection=self.connection)
self.assertFalse(result)
