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 setUp(self):
self.bamboo_url = self.TEST_BAMBOO_URL
self.connection = Connection(self.bamboo_url)
self.default_connection = Connection(DEFAULT_BAMBOO_URL)
# these two datasets (if created) will automatically
# get deleted by the test harness
# NOTE: do not reuse these names for tests, they
# should only be created through the helper functions
self.dataset = None
self.aux_dataset = None
# add any additional datasets should be added
# to this list and they will be deleted as well
self.datasets_to_delete = []
def join(cls, left_dataset, right_dataset, on, connection=None):
"""
Create a new dataset that is the result of a join, where this
left_dataset is the lefthand side and *right_dataset* is the
righthand side and *on* is the column on which to join.
The column that is joined on must be unique in the righthand side
and must exist in both datasets.
"""
if connection is None:
connection = Connection()
if not isinstance(left_dataset, Dataset) or\
not isinstance(right_dataset, Dataset):
raise PyBambooException(
'datasets must be an instances of Dataset.')
data = {
'dataset_id': left_dataset.id,
'other_dataset_id': right_dataset.id,
'on': on,
}
result = connection.make_api_request(
'POST', '/datasets/join', data=data)
if 'id' in result.keys():
return Dataset(result['id'], connection=connection)
return False
def merge(cls, datasets, connection=None):
"""
Create a new dataset that is a row-wise merge of those in *datasets*.
Returns the new merged dataset.
"""
if connection is None:
connection = Connection()
# TODO: allow list of dataset_ids?
checked_datasets = []
for dataset in datasets:
if not isinstance(dataset, Dataset):
raise PyBambooException(
'Datasets need to be instances of Dataset.')
checked_datasets.append(dataset.id)
data = {'dataset_ids': safe_json_dumps(
checked_datasets,
PyBambooException('datasets is not JSON-serializable.'))}
result = connection.make_api_request(
'POST', '/datasets/merge', data=data)
if 'id' in result.keys():
return Dataset(result['id'], connection=connection)
# this is never reached...
# see TestDataset.test_merge_fail()
return False
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 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 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 _set_sources(self, site=None):
path = os.path.join(os.path.dirname(__file__), 'sources.json')
if isinstance(site, basestring):
path = os.path.join(
os.path.dirname(__file__), 'data',
site.lower(), 'sources.json'
)
f = open(path)
self._sources_dict = json.loads(f.read())
self._sources = self._sources_dict['sources']
self.connection = Connection(self._sources_dict['bamboo_server'])
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 __init__(self, dataset_id=None, url=None,
path=None, content=None, data_format='csv',
schema_path=None, schema_content=None,
na_values=None, connection=None, reset=False):
"""
Create a new pybamboo.Dataset from one of the following:
* dataset_id - the id of an existing bamboo.Dataset
* url - url to a .csv file
* path - path to a local .csv or .json file
* content - a CSV or JSON string
* data_format - whether path or content is csv | json
* schema_path - path to a JSON SDF schema
* schema_content - a JSON SDF string
One can also pass in a pybamboo.Connection object. If this is not
supplied one will be created automatically with the default options.
"""
if dataset_id is None and url is None \
and path is None and content is None \
and schema_path is None and schema_content is None:
raise PyBambooException(
'Must supply dataset_id, url, content, schema or file path.')
if data_format not in self.DATA_FORMATS:
raise PyBambooException('Illegal data_format: %s. data_format'
' must be one of %s' %
(data_format, self.DATA_FORMATS))
req_data = {}
if reset:
req_data.update({'dataset_id': self._id})
if na_values is not None:
if not isinstance(na_values, (list, tuple, set)):
raise PyBambooException('N/A values must be a list.')
self.NA_VALUES = na_values
req_data.update({'na_values':
safe_json_dumps(na_values,
PyBambooException('na_values '
'are not JSON-serializable'))})
if connection is None:
self._connection = Connection()
else:
self._connection = connection
if dataset_id is not None:
# TODO: check if this dataset exists?
self._id = dataset_id
return
if url is not None:
# TODO: check valid url?
req_data.update({'url': url})
self._id = self._connection.make_api_request(
'POST', '/datasets', req_data).get('id')
return
# files might be overloaded by schema or path/content
files = {}
if schema_path is not None or schema_content is not None:
# TODO: check for bad file stuff?
schema_data = schema_content if schema_content is not None \
else open(schema_path)
files.update({'schema': ('data.schema.json', schema_data)})
if path is not None or content is not None:
# TODO: check for bad file stuff?
data = content if content is not None else open(path)
files.update({'%s_file' % data_format:
('data.%s' % data_format, data)})
self._id = self._connection.make_api_request('POST', '/datasets',
files=files,
data=req_data).get('id')
import json
import sys
import time
from pybamboo.connection import Connection
from pybamboo.dataset import Dataset
PROD_BAMBOO_ID_FILE = 'ids/ids.prod.json'
DEV_BAMBOO_ID_FILE = 'ids/ids.dev.json'
DEV_BAMBOO = True
# set bamboo instance
if DEV_BAMBOO:
connection = Connection('http://localhost:8080')
bamboo_id_file = DEV_BAMBOO_ID_FILE
else:
connection = Connection()
bamboo_id_file = PROD_BAMBOO_ID_FILE
# 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():
def __init__(self, site=None):
self.connection = Connection(BAMBOO_URL)
self._set_sources(site)
self._db = MongoClient().bamboo_dev
