def __datasource_to_csv(self, execution_date):
final_query = self.extract_query.\
replace("$EXECUTION_DATE", """'%s'""" % execution_date)
logging.info("QUERY : %s" % final_query)
cursor = PostgresHook(self.connection).get_conn().cursor()
cursor.execute(final_query)
result = cursor.fetchall()
# Write to CSV file
temp_path = self.file_path + \
self.table_name + \
'_' + execution_date + '.csv'
with open(temp_path, 'w') as fp:
a = csv.writer(fp, quoting=csv.QUOTE_MINIMAL, delimiter='|')
a.writerow([i[0] for i in cursor.description])
a.writerows(result)
# Read CSV file
full_path = temp_path + '.gz'
with open(temp_path, 'rb') as f:
data = f.read()
# Compress CSV file
with gzip.open(full_path, 'wb') as output:
try:
output.write(data)
finally:
output.close()
# Close file after reading
f.close()
# Delete csv file
os.remove(temp_path)
# Change access mode
os.chmod(full_path, 0o777)
def __csv_to_db(self, execution_date):
csv_file_path = self.file_path + \
self.table_name + \
'_' + execution_date + '.csv.gz'
with gzip.open(csv_file_path, 'rt') as f:
csvobj = csv.reader(f,delimiter = '|',quotechar='"')
create_query = """
CREATE TABLE etl.order (
id bigint primary key,
student_id bigint,
teacher_id bigint,
stage varchar(10),
status varchar(512),
created_at timestamp,
updated_at timestamp
);
"""
cursor = PostgresHook(self.connection).get_conn().cursor()
cursor.execute(final_query)
result = cursor.fetchall()
# Write to CSV file
temp_path = self.file_path + \
self.table_name + \
'_' + execution_date + '.csv'
with open(temp_path, 'w') as fp:
a = csv.writer(fp, quoting=csv.QUOTE_MINIMAL, delimiter='|')
a.writerow([i[0] for i in cursor.description])
a.writerows(result)
# Read CSV file
full_path = temp_path + '.gz'
with open(temp_path, 'rb') as f:
data = f.read()
# Compress CSV file
with gzip.open(full_path, 'wb') as output:
try:
output.write(data)
finally:
output.close()
# Close file after reading
f.close()
# Delete csv file
os.remove(temp_path)
# Change access mode
os.chmod(full_path, 0o777)
