def update_dataset(self, source_url_hash, s3_path=None):
md = session.query(MetaTable).get(source_url_hash)
if md.result_ids:
ids = md.result_ids
ids.append(self.request.id)
else:
ids = [self.request.id]
with engine.begin() as c:
c.execute(MetaTable.__table__.update()\
.where(MetaTable.source_url_hash == source_url_hash)\
.values(result_ids=ids))
etl = PlenarioETL(md.as_dict())
etl.update(s3_path=s3_path)
return 'Finished updating {0} ({1})'.format(md.human_name, md.source_url_hash)
def update_dataset(self, source_url_hash, s3_path=None):
md = session.query(MetaTable).get(source_url_hash)
if md.result_ids:
ids = md.result_ids
ids.append(self.request.id)
else:
ids = [self.request.id]
with engine.begin() as c:
c.execute(
MetaTable.__table__.update().where(MetaTable.source_url_hash == source_url_hash).values(result_ids=ids)
)
etl = PlenarioETL(md.as_dict())
etl.update(s3_path=s3_path)
return "Finished updating {0} ({1})".format(md.human_name, md.source_url_hash)
def update_dataset(source_url_hash, s3_path=None):
md = session.query(MetaTable).get(source_url_hash)
etl = PlenarioETL(md.as_dict())
etl.update(s3_path=s3_path)
return 'Finished updating %s' % md.human_name
def add_dataset(source_url_hash, s3_path=None, data_types=None):
md = session.query(MetaTable).get(source_url_hash)
etl = PlenarioETL(md.as_dict(), data_types=data_types)
etl.add(s3_path=s3_path)
return 'Finished adding %s' % md.human_name
from sqlalchemy import Table, MetaData
from sqlalchemy.sql.sqltypes import DATE
from plenario.database import session, app_engine
from plenario.models import MetaTable
from sqlalchemy.exc import NoSuchTableError
from plenario.utils.helpers import slugify
from plenario.utils.etl import PlenarioETL
import traceback
if __name__ == "__main__":
them = session.query(MetaTable).all()
meta = MetaData()
for t in them:
try:
table = Table('dat_{0}'.format(t.dataset_name), meta,
autoload=True, autoload_with=app_engine, keep_existing=True)
try:
date_col = getattr(table.c, slugify(t.observed_date))
if type(date_col.type) == DATE:
e = PlenarioETL(t.as_dict())
e._get_or_create_data_table()
e._add_weather_info()
print 'added weather for {0}'.format(t.dataset_name)
except AttributeError, e:
raise e
print 'no col {0}'.format(t.observed_date)
pass
except NoSuchTableError:
print 'no table {0}'.format(t.dataset_name)
pass
def setUpClass(cls):
# Assume there exists a test database with postgis at the connection string specified in test_settings.py
tables_to_drop = [
"census_blocks",
"dat_flu_shot_clinic_locations",
"dat_master",
"meta_master",
"meta_shape",
"plenario_user",
]
drop_tables(tables_to_drop)
# Create meta, master, user tables
init_master_meta_user()
# Ingest the census blocks
init_census()
# TODO: support local ingest of csv
# For now, ingest Chicago's csv of 2013 flu shot locations from the data portal.
# It's a nice little Chicago dataset that won't change.
# So, adding the dataset to meta_table happens in view.py.
# I don't want to mock out a whole response object with form data and such,
# so here's a janky way.
url = "https://data.cityofchicago.org/api/views/g5vx-5vqf/rows.csv?accessType=DOWNLOAD"
url_hash = md5(url).hexdigest()
d = {
"dataset_name": u"flu_shot_clinic_locations",
"human_name": u"flu_shot_clinic_locations",
"attribution": u"foo",
"description": u"bar",
"source_url": url,
"source_url_hash": url_hash,
"update_freq": "yearly",
"business_key": u"Event",
"observed_date": u"Date",
"latitude": u"Latitude",
"longitude": u"Longitude",
"location": u"Location",
"contributor_name": u"Frederick Mcgillicutty",
"contributor_organization": u"StrexCorp",
"contributor_email": u"*****@*****.**",
"contributed_data_types": None,
"approved_status": True,
"is_socrata_source": False,
}
# add this to meta_master
md = MetaTable(**d)
session.add(md)
session.commit()
meta = {
"dataset_name": u"flu_shot_clinic_locations",
"source_url": url,
"business_key": u"Event",
"observed_date": u"Date",
"latitude": u"Latitude",
"longitude": u"Longitude",
"location": u"Location",
"source_url_hash": url_hash,
}
point_etl = PlenarioETL(meta)
point_etl.add()
cls.app = create_app().test_client()
def update_dataset(source_url_hash, s3_path=None):
md = session.query(MetaTable).get(source_url_hash)
etl = PlenarioETL(md.as_dict())
etl.update(s3_path=s3_path)
return 'Finished updating %s' % md.human_name
def add_dataset(source_url_hash, s3_path=None, data_types=None):
md = session.query(MetaTable).get(source_url_hash)
etl = PlenarioETL(md.as_dict(), data_types=data_types)
etl.add(s3_path=s3_path)
return 'Finished adding %s' % md.human_name
from plenario.models import MetaTable
from sqlalchemy.exc import NoSuchTableError
from plenario.utils.helpers import slugify
from plenario.utils.etl import PlenarioETL
import traceback
if __name__ == "__main__":
them = session.query(MetaTable).all()
meta = MetaData()
for t in them:
try:
table = Table('dat_{0}'.format(t.dataset_name),
meta,
autoload=True,
autoload_with=app_engine,
keep_existing=True)
try:
date_col = getattr(table.c, slugify(t.observed_date))
if type(date_col.type) == DATE:
e = PlenarioETL(t.as_dict())
e._get_or_create_data_table()
e._add_weather_info()
print 'added weather for {0}'.format(t.dataset_name)
except AttributeError, e:
raise e
print 'no col {0}'.format(t.observed_date)
pass
except NoSuchTableError:
print 'no table {0}'.format(t.dataset_name)
pass