def setup_logging_client(log_name: str):
"""
This function will connect to gcluod and post all logging commands there as well as locally
The "Log Name" can then be set to the name of the service as a filter
https://console.cloud.google.com/logs?project=development-278003
:param log_name: The custom name of the log file that will be used in gcloud
:return:
"""
if os.path.exists("config/GCLOUD_LOGGING_SERVICE_KEY.json"):
# The GCLOUD_LOGGING_SERVICE_KEY exists in circle ci, and is passed through to the service
# There is one for each environment. eg. development:
# console.cloud.google.com/iam-admin/serviceaccounts/details/104042617795891603364?project=development-278003
os.environ[
"GOOGLE_APPLICATION_CREDENTIALS"] = "config/GCLOUD_LOGGING_SERVICE_KEY.json"
# Instantiates a client and handler for logging with gcloud
client = google.cloud.logging.Client()
handler = CloudLoggingHandler(name=log_name, client=client)
logging.getLogger().setLevel(logging.INFO) # defaults to WARN
setup_logging(handler)
logging.debug("Logging connected to GCloud")
else:
print("No GCLOUD_LOGGING_SERVICE_KEY detected, using native logging.")
def setup_logging(self,
log_level=logging.INFO,
excluded_loggers=EXCLUDED_LOGGER_DEFAULTS,
**kw):
"""Attach default Stackdriver logging handler to the root logger.
This method uses the default log handler, obtained by
:meth:`~get_default_handler`, and attaches it to the root Python
logger, so that a call such as ``logging.warn``, as well as all child
loggers, will report to Stackdriver logging.
:type log_level: int
:param log_level: (Optional) Python logging log level. Defaults to
:const:`logging.INFO`.
:type excluded_loggers: tuple
:param excluded_loggers: (Optional) The loggers to not attach the
handler to. This will always include the
loggers in the path of the logging client
itself.
:type kw: dict
:param kw: keyword args passed to handler constructor
"""
handler = self.get_default_handler(**kw)
setup_logging(handler,
log_level=log_level,
excluded_loggers=excluded_loggers)
def setup_logging(
self, log_level=logging.INFO, excluded_loggers=EXCLUDED_LOGGER_DEFAULTS, **kw
):
"""Attach default Stackdriver logging handler to the root logger.
This method uses the default log handler, obtained by
:meth:`~get_default_handler`, and attaches it to the root Python
logger, so that a call such as ``logging.warn``, as well as all child
loggers, will report to Stackdriver logging.
:type log_level: int
:param log_level: (Optional) Python logging log level. Defaults to
:const:`logging.INFO`.
:type excluded_loggers: tuple
:param excluded_loggers: (Optional) The loggers to not attach the
handler to. This will always include the
loggers in the path of the logging client
itself.
:type kw: dict
:param kw: keyword args passed to handler constructor
"""
handler = self.get_default_handler(**kw)
setup_logging(handler, log_level=log_level, excluded_loggers=excluded_loggers)
def setup():
logger = logging.getLogger()
# Set the region on log records.
default_factory = logging.getLogRecordFactory()
logging.setLogRecordFactory(partial(region_record_factory,
default_factory))
# Send logs directly via the logging client if possible. This ensures trace
# ids are propogated and allows us to send structured messages.
if environment.in_gae():
client = Client()
handler = StructuredAppEngineHandler(client)
handlers.setup_logging(handler, log_level=logging.INFO)
for handler in logger.handlers:
if not isinstance(handler, StructuredAppEngineHandler):
logger.removeHandler(handler)
else:
logging.basicConfig()
for handler in logger.handlers:
# If writing directly to Stackdriver, send a structured message.
if isinstance(handler, StructuredAppEngineHandler):
handler.setFormatter(StructuredLogFormatter())
# Otherwise, the default stream handler requires a string.
else:
handler.setFormatter(
logging.Formatter(
"(%(region)s) %(module)s/%(funcName)s : %(message)s"))
def _call_fut(self, handler, excludes=None):
from google.cloud.logging.handlers import setup_logging
if excludes:
return setup_logging(handler, excluded_loggers=excludes)
else:
return setup_logging(handler)
def create_app(config, debug=False, testing=False, config_overrides=None):
app = Flask(__name__)
app.config.from_object(config)
app.debug = debug
app.testing = testing
if config_overrides:
app.config.update(config_overrides)
# [START setup_logging]
if not app.testing:
client = google.cloud.logging.Client(app.config['PROJECT_ID'])
handler = CloudLoggingHandler(client)
# Attaches the handler to the root logger
setup_logging(handler)
logging.getLogger().setLevel(logging.INFO)
# [END setup_logging]
# Setup the data model.
with app.app_context():
model = get_model()
model.init_app(app)
# Initalize the OAuth2 helper.
oauth2.init_app(
app,
scopes=['email', 'profile'],
authorize_callback=_request_user_info)
# Add a logout handler.
@app.route('/logout')
def logout():
# Delete the user's profile and the credentials stored by oauth2.
del session['profile']
session.modified = True
oauth2.storage.delete()
return redirect(request.referrer or '/')
# Register the Bookshelf CRUD blueprint.
from .crud import crud
app.register_blueprint(crud, url_prefix='/books')
# Add a default root route.
@app.route("/")
def index():
return redirect(url_for('crud.list'))
# Add an error handler. This is useful for debugging the live application,
# however, you should disable the output of the exception for production
# applications.
@app.errorhandler(500)
def server_error(e):
return """
An internal error occurred: <pre>{}</pre>
See logs for full stacktrace.
""".format(e), 500
return app
def setup_logger() -> object:
'''Setup main application logger
'''
# Setup uncaught exception handler
default_exception_handler()
if app_config['app_log_factory'] == 'gcp' and os.getenv(
'GOOGLE_APPLICATION_CREDENTIALS') is None:
app_config[
'app_log_factory'] = 'file' # Prevent set up GCP logging without credentials ENV variable
logging.error(
'GOOGLE_APPLICATION_CREDENTIALS ENV variable is missing, logging set to file output'
)
if app_config['app_log_factory'] == 'gcp':
# Disable whole cherrypy console logging
cherrypy.log.screen = False
logging.getLogger("cherrypy").propagate = False
# Connect GCP logging to default Python logger
logger = logging.getLogger()
# Remove original log handlers
for handler in logger.handlers:
logger.removeHandler(handler)
# Setup Google Cloud Logging
client = google.cloud.logging.Client()
# Setup CloudLoggingHandler(logging.StreamHandler) handler explicitly with Custom GCP Formatter
handler = CloudLoggingHandler(client,
labels={
'app_name': app_config['app_name'],
'app_version':
app_config['app_version'],
'app_environment':
app_config['app_env']
})
handler.setFormatter(CustomGCPFormatter())
# Setup Python logger explicitly with custom handler
setup_logging(handler)
elif app_config['app_log_factory'] == 'file':
# Load log configuration
logging.config.dictConfig(LOG_CONFIG)
# Custom app logger
logger = logging.getLogger('app')
else:
# Load log configuration
logging.config.dictConfig(LOG_CONFIG)
# Custom app logger
logger = logging.getLogger()
return logger
def create_app(config, debug=False, testing=False, config_overrides=None):
app = Flask(__name__)
app.config.from_object(config)
app.debug = debug
app.testing = testing
if config_overrides:
app.config.update(config_overrides)
# [START setup_logging]
if not app.testing:
client = google.cloud.logging.Client(app.config['PROJECT_ID'])
handler = CloudLoggingHandler(client)
# Attaches the handler to the root logger
setup_logging(handler)
logging.getLogger().setLevel(logging.INFO)
# [END setup_logging]
# Setup the data model.
with app.app_context():
model = get_model()
model.init_app(app)
# Initalize the OAuth2 helper.
oauth2.init_app(app,
scopes=['email', 'profile'],
authorize_callback=_request_user_info)
# Add a logout handler.
@app.route('/logout')
def logout():
# Delete the user's profile and the credentials stored by oauth2.
del session['profile']
session.modified = True
oauth2.storage.delete()
return redirect(request.referrer or '/')
# Register the Bookshelf CRUD blueprint.
from .crud import crud
app.register_blueprint(crud, url_prefix='/books')
# Add a default root route.
@app.route("/")
def index():
return redirect(url_for('crud.list'))
# Add an error handler. This is useful for debugging the live application,
# however, you should disable the output of the exception for production
# applications.
@app.errorhandler(500)
def server_error(e):
return """
An internal error occurred: <pre>{}</pre>
See logs for full stacktrace.
""".format(e), 500
return app
def configure(global_config, **settings):
gcp_project = settings.get('idris.google_cloud_project')
gcp_auth = settings.get('idris.google_application_credentials')
if gcp_project and gcp_auth:
os.environ['GOOGLE_CLOUD_PROJECT'] = gcp_project
os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = os.path.abspath(
gcp_auth)
if settings.get('idris.use_google_cloud_logging') == 'true':
if 'GAE_INSTANCE' in os.environ:
client = google.cloud.logging.Client()
else:
client = google.cloud.logging.Client().from_service_account_json(
settings['idris.google_application_credentials'])
handler = CloudLoggingHandler(client)
logging.getLogger().setLevel(logging.INFO)
setup_logging(handler)
config = Configurator(settings=settings, root_factory=root_factory)
config.add_tween('idris.token_tween_factory')
config.include('cornice')
config.include('cornice_swagger')
config.include('pyramid_chameleon')
config.include('pyramid_jwt')
config.include('idris.storage')
config.include('idris.blob')
config.include('idris.services.lookup')
config.include('idris.services.cache')
config.include('idris.services.download_counter')
config.include('idris.services.auditlog')
config.include('idris.apps.base')
config.include('idris.apps.course')
config.set_authorization_policy(ACLAuthorizationPolicy())
config.set_jwt_authentication_policy(settings['idris.secret'],
http_header='Authorization',
auth_type='Bearer',
expiration=3600,
callback=add_role_principals)
config.scan("idris.views")
config.add_route('liveness_check', '/_live')
config.add_route('readiness_check', '/_ready')
config.add_route('api_without_slash', '/api')
config.add_view(lambda _, __: HTTPFound('/api/'),
route_name='api_without_slash')
config.add_static_view('api', path='idris:static/dist/swagger')
config.add_static_view('static', path='idris:static/dist/web')
config.add_route('edit_without_slash', '/edit')
config.add_view(lambda _, __: HTTPFound('/edit/'),
route_name='edit_without_slash')
return config
def init_gcloud_log(project_id, logger_name):
log_format = "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
if IS_DEV_ENV or IS_TEST_ENV:
logging.basicConfig(filename=logger_name + u'.log', level=logging.DEBUG, format=log_format)
else:
# see: https://googlecloudplatform.github.io/google-cloud-python/latest/logging-usage.html#cloud-logging-handler
# and https://github.com/GoogleCloudPlatform/getting-started-python/blob/master/6-pubsub/bookshelf/__init__.py#L40
client = google.cloud.logging.Client(project_id)
# NB: we should use AppEngineHandler for server at next google.cloud API update
# https://googlecloudplatform.github.io/google-cloud-python/latest/logging-handlers-app-engine.html
handler = CloudLoggingHandler(client, logger_name)
handler.setFormatter(logging.Formatter(log_format))
setup_logging(handler)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("readability.readability").setLevel(logging.WARNING) # very verbose package
def setup_gce_logging(gunicorn_access_logger,
gunicorn_error_logger): # pragma: no cover
if application.config['SQLALCHEMY_DATABASE_URI'][:22] in [
'postgresql://localhost', 'db://localhost/test_db'
]:
return
import google.cloud.logging
from google.cloud.logging.handlers import CloudLoggingHandler, setup_logging
client = google.cloud.logging.Client()
handler = CloudLoggingHandler(client, name=get_env())
setup_logging(handler)
gunicorn_access_logger.addHandler(handler)
gunicorn_error_logger.addHandler(handler)
def setup() -> None:
"""Setup logging"""
# Set the region on log records.
logging.setLogRecordFactory(ContextualLogRecord)
logger = logging.getLogger()
# Send logs directly via the logging client if possible. This ensures trace
# ids are propagated and allows us to send structured messages.
if environment.in_gcp():
client = Client()
structured_handler = StructuredAppEngineHandler(client)
handlers.setup_logging(structured_handler, log_level=logging.INFO)
before_request_handler = StructuredAppEngineHandler(
client, name=BEFORE_REQUEST_LOG)
logging.getLogger(BEFORE_REQUEST_LOG).addHandler(
before_request_handler)
# Streams unstructured logs to stdout - these logs will still show up
# under the appengine.googleapis.com/stdout Stackdriver logs bucket,
# even if other logs are stalled on the global interpreter lock or some
# other issue.
stdout_handler = logging.StreamHandler(sys.stdout)
handlers.setup_logging(stdout_handler, log_level=logging.INFO)
for handler in logger.handlers:
if not isinstance(
handler,
(StructuredAppEngineHandler, logging.StreamHandler)):
logger.removeHandler(handler)
else:
logging.basicConfig()
for handler in logger.handlers:
# If we aren't writing directly to Stackdriver, prefix the log with important
# context that would be in the labels.
if not isinstance(handler, StructuredAppEngineHandler):
handler.setFormatter(
logging.Formatter(
"[pid: %(process)d] (%(region)s) %(module)s/%(funcName)s : %(message)s"
))
# Export gunicorn errors using the same handlers as other logs, so that they
# go to Stackdriver in production.
gunicorn_logger = logging.getLogger("gunicorn.error")
gunicorn_logger.handlers = logger.handlers
def setup():
"""Setup logging"""
# Set the region on log records.
default_factory = logging.getLogRecordFactory()
logging.setLogRecordFactory(partial(region_record_factory,
default_factory))
logger = logging.getLogger()
# Send logs directly via the logging client if possible. This ensures trace
# ids are propogated and allows us to send structured messages.
if environment.in_gcp():
client = Client()
handler = StructuredAppEngineHandler(client)
handlers.setup_logging(handler, log_level=logging.INFO)
# Streams unstructured logs to stdout - these logs will still show up
# under the appengine.googleapis.com/stdout Stackdriver logs bucket,
# even if other logs are stalled on the global interpreter lock or some
# other issue.
stdout_handler = logging.StreamHandler(sys.stdout)
handlers.setup_logging(stdout_handler, log_level=logging.INFO)
for handler in logger.handlers:
if not isinstance(
handler,
(StructuredAppEngineHandler, logging.StreamHandler)):
logger.removeHandler(handler)
else:
logging.basicConfig()
for handler in logger.handlers:
# If writing directly to Stackdriver, send a structured message.
if isinstance(handler, StructuredAppEngineHandler):
handler.setFormatter(StructuredLogFormatter())
# Otherwise, the default stream handler requires a string.
else:
handler.setFormatter(
logging.Formatter(
"(%(region)s) %(module)s/%(funcName)s : %(message)s"))
# Export gunicorn errors using the same handlers as other logs, so that they
# go to Stackdriver in production.
gunicorn_logger = logging.getLogger("gunicorn.error")
gunicorn_logger.handlers = logger.handlers
def getLogger(name, level=__loglevel):
formatter = logging.Formatter('%(nasip)s %(nasid)s %(siteid)s %(msgtype)s %(message)s', '%a, %d %b %Y %H:%M:%S', )
logger = logging.getLogger(name)
logger.setLevel(level)
logger.addFilter(log_ctx)
if GRAYLOG_SERVER:
client = google.cloud.logging.Client()
cloud_handler = CloudLoggingHandler(client, name="radiusd")
logger.addHandler(cloud_handler)
setup_logging(cloud_handler)
cloud_handler.addFilter(log_ctx)
cloud_handler.setFormatter(formatter)
else:
stream_handler = logging.StreamHandler(sys.stdout)
stream_handler.setFormatter(formatter)
logger.addHandler(stream_handler)
return logger
def setup_stackdriver(log_level, name, log_format, excluded_loggers=None):
try:
from google.cloud.logging import Client
from google.cloud.logging import handlers as google_logging_handlers
from google.cloud.logging.handlers.handlers import \
EXCLUDED_LOGGER_DEFAULTS, \
CloudLoggingHandler
except ImportError:
raise ValueError("google-cloud-logging is not properly installed")
if not excluded_loggers:
excluded_loggers = EXCLUDED_LOGGER_DEFAULTS
client = Client()
# the docstring of CloudLoggingHandler point to client instead of Client
# noinspection PyTypeChecker
handler = CloudLoggingHandler(client, name)
handler.setFormatter(logging.Formatter(log_format, None, "%"))
google_logging_handlers.setup_logging(handler,
log_level=log_level,
excluded_loggers=excluded_loggers)
def _initialize_stackdriver_logging(self):
stackdriver_client = google.cloud.logging.Client()
stackdriver_handler = CloudLoggingHandler(stackdriver_client,name=__name__, resource=self.log_resource, labels={})
setup_logging(stackdriver_handler)
import logging
logger = logging.getLogger("gs-scroller.main")
if __name__ == '__main__':
logging.getLogger("gs-scroller").setLevel(logging.INFO)
else:
logging.getLogger("gs-scroller").setLevel(logging.INFO)
try:
import google.cloud.logging
logging_client = google.cloud.logging.Client()
from google.cloud.logging.handlers import CloudLoggingHandler, setup_logging
logging_handler = CloudLoggingHandler(logging_client)
except ImportError:
logger.warning("Google's logging module could not be imported")
else:
setup_logging(logging_handler)
logger.info("Google's logging module was imported")
logger.debug("Debug messages will be logged")
try:
from google.appengine.api import wrap_wsgi_app
except ImportError:
logger.warning("Google's appengine wrapper could not be imported")
wrap_wsgi_app = None
else:
logger.info("Google's appengine wrapper was imported")
from converters import (Base64Converter, DigitsConverter, DigitListConverter)
from cache import temporary_cache
import urlread
def runflow(sample_size, runner, argv=None):
from pipeline_package import pipe_module as pm
parser = argparse.ArgumentParser()
known_args, pipeline_args = parser.parse_known_args(argv)
config_object = ConfigParser()
config_object.read("config.ini")
session_info = config_object["SESSION_INFO"]
bigquery_config = config_object["BIGQUERY_CONFIG"]
db_name = session_info["session_name"]
bq_project = bigquery_config["project_name"]
bq_account = bigquery_config["account_name"]
bq_auth = bigquery_config["google_credential_path"]
model_bucket = bigquery_config["model_bucket"]
model_path = bigquery_config["model_path"]
model_dest = bigquery_config["model_dest"]
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = bq_auth
session_id = bq_account + '_' + db_name
client = google.cloud.logging.Client()
handler = CloudLoggingHandler(client, name=session_id)
logging.getLogger().setLevel(logging.INFO) # defaults to WARN
setup_logging(handler)
dataset_name = bq_account + '_' + db_name
base_table_id = bq_project + '.' + dataset_name + '.' + 'base_tweets'
pred_table_id = bq_project + '.' + dataset_name + '.' + 'tweet_predictions'
# Construct a BigQuery client object.
client = bigquery.Client()
# create the tweet_predictions table if does not exist
try:
client.get_table(pred_table_id) # Make an API request.
msg = "Table {} already exists.".format(pred_table_id)
print(msg)
logging.info(msg)
except NotFound:
msg = "Table {} is not found. Creating table...".format(pred_table_id)
print(msg)
logging.info(msg)
schema = [
bigquery.SchemaField("tweet_id", "INTEGER", mode="REQUIRED"),
bigquery.SchemaField("prediction", "STRING", mode="REQUIRED"),
]
table = bigquery.Table(pred_table_id, schema=schema)
table = client.create_table(table) # Make an API request.
msg = f'Created table {pred_table_id}'
logging.info(msg)
source_query = pm.get_source_query(sample_size)
sample_size_desc = pm.get_sample_size_desc(sample_size)
# options = PipelineOptions(
# flags=[],
# project=bq_project,
# job_name='sentiment-prediction-job-pt',
# temp_location='gs://sentiment-model-ja/temp',
# staging_location='gs://sentiment-model-ja/temp',
# region='us-central1',
# max_num_workers=2)
with beam.Pipeline(runner, argv=pipeline_args) as pipeline:
table_schema = beam_bq.TableSchema()
# Fields that use standard types.
id_schema = beam_bq.TableFieldSchema()
id_schema.name = 'tweet_id'
id_schema.type = 'integer'
id_schema.mode = 'required'
table_schema.fields.append(id_schema)
predict_schema = beam_bq.TableFieldSchema()
predict_schema.name = 'prediction'
predict_schema.type = 'string'
predict_schema.mode = 'required'
table_schema.fields.append(predict_schema)
(
pipeline
| 'Read from BigQuery {}'.format(sample_size_desc) >> beam.io.ReadFromBigQuery(query=source_query, use_standard_sql=True)
| 'predict' >> beam.ParDo(pm.Predict_bert(project=bq_project, bucket_name=model_bucket,
model_path=model_path,
destination_file_name=model_dest))
| "Write data to BQ" >> beam.io.WriteToBigQuery(table='tweet_predictions', dataset=dataset_name, schema=table_schema,
project=bq_project,
write_disposition=beam.io.BigQueryDisposition.WRITE_APPEND)
)
job = pipeline.run()
if runner == 'DataflowRunner':
job.wait_until_finish()
account_name = bigquery_config["account_name"]
bq_auth = bigquery_config["google_credential_path"]
session_id = account_name + '_' + db_name
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = bq_auth
if not os.path.exists(os.path.join(os.path.dirname(__file__), db_name)):
os.makedirs(db_name)
logging.basicConfig(filename=os.path.join(db_name, 'tweets_capture.log'),
level=logging.INFO)
client = google.cloud.logging.Client()
handler = CloudLoggingHandler(client, name=session_id)
logging.getLogger().setLevel(logging.INFO) # defaults to WARN
setup_logging(handler)
log_name = '_' + db_name + '_to_process.txt'
error_file_log_name = '_' + db_name + '_error_files.txt'
log_path = os.path.join(db_name, log_name)
error_file_log_path = os.path.join(db_name, error_file_log_name)
db_backlog = open(os.path.join(db_name, log_name), 'a+')
db_backlog.close()
error_file_log = open(os.path.join(db_name, error_file_log_name), 'a+')
error_file_log.close()
# inherit from StreamListener class
class SListener(StreamListener):
def do_logging_setup(log_level=logging.DEBUG):
client = google.cloud.logging.Client()
handler = CloudLoggingHandler(client)
logging.getLogger().setLevel(log_level)
setup_logging(handler)
def main():
os.chdir(os.path.dirname(__file__))
#Read config.ini file
config_object = ConfigParser()
config_object.read("config.ini")
session_info = config_object["SESSION_INFO"]
postgres_config = config_object["POSTGRES_CONFIG"]
bigquery_config = config_object["BIGQUERY_CONFIG"]
db_type = session_info["database"]
db_name = session_info["session_name"]
psql_server = postgres_config["host"]
psql_port = postgres_config["port"]
psql_user = postgres_config["username"]
psql_password = postgres_config["password"]
bq_project = bigquery_config["project_name"]
bq_account = bigquery_config["account_name"]
bq_auth = bigquery_config["google_credential_path"]
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = bq_auth
session_id = bq_account + '_' + db_name
log_name = '_' + db_name + '_to_process.txt'
error_file_log_name = '_' + db_name + '_error_files.txt'
backlog_file = os.path.join(db_name, log_name)
error_file_log = os.path.join(db_name, error_file_log_name)
client = google.cloud.logging.Client()
handler = CloudLoggingHandler(client, name=session_id)
logging.getLogger().setLevel(logging.INFO) # defaults to WARN
setup_logging(handler)
print(
f'Begin:\nProcessing json files in {log_name}\n{db_type}:\nProject: {bq_project}\nDataset: {bq_account}.{db_name}'
)
# for handler in logging.root.handlers[:]:
# logging.root.removeHandler(handler)
logging.basicConfig(filename=os.path.join(db_name,
'tweets_processing.log'),
level=logging.INFO)
while True:
while Path(backlog_file).stat().st_size == 0:
wait_msg = f'WAITING {str(datetime.datetime.now())}: waiting for json data'
print(wait_msg)
logging.info(wait_msg)
time.sleep(60)
if Path(backlog_file).stat().st_size > 0:
with open(backlog_file, 'r') as f:
lines = f.readlines()
file_raw = lines[0]
file = file_raw.strip('\n')
with open(backlog_file, 'w') as update:
for line in lines:
if line != file_raw:
update.write(line)
if os.path.isfile(file):
check_conn = False
while check_conn == False:
try:
msg = f'Connecting to {db_type}...'
print(msg)
logging.info(msg)
if db_type == 'postgres':
url = make_url(
f'postgresql://{psql_user}:{psql_password}@{psql_server}:{psql_port}/{db_name}'
)
engine = sqlalchemy.create_engine(url)
with engine.connect() as con:
rs = con.execute('SELECT 1')
print(rs)
check_conn = True
msg = 'Postgres connection successful!'
logging.info(msg)
print(msg)
if not database_exists(
engine.url): #db does not exist
# create a new database
create_database(engine.url)
msg = f"{db_name} postgres database not found. Creating new database."
logging.info(msg)
print(msg)
elif db_type == 'bigquery':
project_name = bq_project
dataset_name = bq_account + '_' + db_name
os.environ[
"GOOGLE_APPLICATION_CREDENTIALS"] = bq_auth
client = bigquery.Client()
# set pandas_gbq project
pandas_gbq.context.project = project_name
# Set dataset_id to the ID of the dataset to determine existence.
dataset_id = project_name + '.' + dataset_name
try:
client.get_dataset(
dataset_id) # Make an API request.
msg = "BigQuery: Dataset {} already exists".format(
dataset_id)
print(msg)
logging.info(msg)
check_conn = True
except NotFound:
msg = "BigQuery: Dataset {} is not found. Creating dataset.".format(
dataset_id)
print(msg)
logging.info(msg)
dataset = bigquery.Dataset(dataset_id)
# TODO(developer): Specify the geographic location where the dataset should reside.
dataset.location = "US"
# Send the dataset to the API for creation, with an explicit timeout.
# Raises google.api_core.exceptions.Conflict if the Dataset already
# exists within the project.
dataset = client.create_dataset(
dataset,
timeout=30) # Make an API request.
msg = "BigQuery: Created dataset {}.{}".format(
client.project, dataset.dataset_id)
logging.info(msg)
print(msg)
except Exception as e:
# error_msg = f'ERROR {str(datetime.datetime.now())}: Connection to postgres could not be established'
# logging.info(error_msg)
template = "An exception of type {0} occurred. Arguments:\n{1!r}"
message = template.format(type(e).__name__, e.args)
print(message)
print(traceback.format_exc())
logging.info(message)
check_conn = False
with open(error_file_log, 'a+') as f:
f.write(file_raw)
sys.exit()
try:
file_lst = glob.glob(file)
for file_idx, file_name in enumerate(file_lst):
if file_idx == 0:
option = 'append'
else:
option = 'append'
process_start_msg = f'PROCESSING: {str(datetime.datetime.now())}: processing {file}'
print(process_start_msg)
logging.info(process_start_msg)
try:
to_sql_g = json2df(file_name)
print('Conversion of json to df complete.')
except Exception as e:
template = "An exception of type {0} occurred during json to df conversion or prediction. Arguments:\n{1!r}"
message = template.format(
type(e).__name__, e.args)
print(message)
print(traceback.format_exc())
logging.info(message)
else:
for df_name, df in to_sql_g.items():
table_id = project_name + '.' + dataset_name + '.' + df_name
msg = f'{str(datetime.datetime.now())}: uploading {df_name} --> {db_type}:{table_id}'
print(msg)
logging.info(msg)
if db_type == 'postgres':
df.to_sql(df_name,
con=engine,
if_exists=option)
elif db_type == 'bigquery':
table_name = df_name
# create new table and write df
dest_table = dataset_name + '.' + table_name
pandas_gbq.to_gbq(df,
dest_table,
if_exists='append')
msg = f'{str(datetime.datetime.now())}: uploading {df_name} --> {db_type} complete'
print(msg)
logging.info(msg)
if db_type == 'postgres':
engine.dispose()
cdate = str(datetime.datetime.now())
success_msg = 'SUCCESS {}: (DB {}) FILE {}'.format(
cdate, db_type, file_name)
logging.info(success_msg)
except Exception as e:
#cdate = str(datetime.datetime.now())
#error_msg = 'ERROR {}: FILE {}: {}'.format(cdate, file_name, e)
template = "An exception of type {0} occurred. Arguments:\n{1!r}"
message = template.format(type(e).__name__, e.args)
print(message)
print(traceback.format_exc())
logging.info(message)
with open(error_file_log, 'a+') as f:
f.write(file_raw)
else:
msg = f'{file} does not exist'
print(msg)
logging.info(msg)
with open(error_file_log, 'a+') as f:
f.write(file_raw)
sys.exit()
