
Research
/Security News
9 Malicious NuGet Packages Deliver Time-Delayed Destructive Payloads
Socket researchers discovered nine malicious NuGet packages that use time-delayed payloads to crash applications and corrupt industrial control systems.
moesifwsgi
Advanced tools
by Moesif, the API analytics and API monetization platform.
With Moesif middleware for Python WSGI-based frameworks, you can automatically log API calls and send them to Moesif for API analytics and monitoring.
If you're new to Moesif, see our Getting Started resources to quickly get up and running.
This middleware allows you to integrate Moesif's API analytics and API monetization features with minimal configuration into APIs that are built on Python WSGI-based (Web Server Gateway Interface) frameworks.
WSGI (Web Server Gateway Interface) is a standard (PEP 3333) that describes how a web server communicates with web applications. Many Python Frameworks are build on top of WSGI, such as Flask, Bottle, and Pyramid.
Before using this middleware, make sure you have the following:
After you log into Moesif Portal, you can get your Moesif Application ID during the onboarding steps. You can always access the Application ID any time by following these steps from Moesif Portal after logging in:
Install with pip using the following command:
pip install moesifwsgi
See the available configuration options to learn how to configure the middleware for your use case.
Wrap your wsgi_app with the Moesif middleware.
from moesifwsgi import MoesifMiddleware
moesif_settings = {
'APPLICATION_ID': 'YOUR_MOESIF_APPLICATION_ID',
'LOG_BODY': True,
# ... For other options see below.
}
app.wsgi_app = MoesifMiddleware(app.wsgi_app, moesif_settings)
Replace YOUR_MOESIF_APPLICATION_ID with your Moesif Application ID.
For an example with Flask, see the /examples/flask folder of this repository.
Wrap your Bottle application with the Moesif middleware:
from moesifwsgi import MoesifMiddleware
app = bottle.Bottle()
moesif_settings = {
'APPLICATION_ID': 'YOUR_MOESIF_APPLICATION_ID',
'LOG_BODY': True,
# ... For other options see below.
}
bottle.run(app=MoesifMiddleware(app, moesif_settings))
Replace YOUR_MOESIF_APPLICATION_ID with your Moesif Application ID.
For an example with Bottle, see the /examples/bottle folder of this repository.
from pyramid.config import Configurator
from moesifwsgi import MoesifMiddleware
if __name__ == '__main__':
config = Configurator()
config.add_route('hello', '/')
config.scan()
app = config.make_wsgi_app()
# configure your moesif settings
moesif_settings = {
'APPLICATION_ID': 'YOUR_MOESIF_APPLICATION_ID',
'LOG_BODY': True,
# ... For other options see below.
}
# Put middleware
app = MoesifMiddleware(app, moesif_settings)
server = make_server('0.0.0.0', 8080, app)
server.serve_forever()
Replace YOUR_MOESIF_APPLICATION_ID with your Moesif Application ID.
If you are using a framework that is built on top of WSGI, it should work just by adding the Moesif middleware. Please read the documentation for your specific framework on how to add middlewares.
In addition to your own APIs, you can also start capturing calls out to third party services through by setting the CAPTURE_OUTGOING_REQUESTS option:
from moesifwsgi import MoesifMiddleware
from flask import Flask
moesif_settings = {
'APPLICATION_ID': 'YOUR_MOESIF_APPLICATION_ID',
'LOG_BODY': True,
'CAPTURE_OUTGOING_REQUESTS': False
}
app = Flask(__name__)
app.wsgi_app = MoesifMiddleware(app.wsgi_app, moesif_settings)
For configuration options specific to capturing outgoing API calls, see Options For Outgoing API Calls.
For a general troubleshooting guide that can help you solve common problems, see Server Troubleshooting Guide.
Other troubleshooting supports:
This library manages a thread pool to send data to Moesif in the background without impacting your app's latency.
However, the preload feature of Gunicorn (or preload-app for Hypercorn) may interfere with the thread pool before the worker is forked. If you encounter issues, avoid setting preload to True.
When using Docker with Ubuntu-based image, events may not be captured if the image fails to find any timezone configuration. To solve this issue, add the following line to your Dockerfile:
ENV TZ=UTC
Otherwise, you can add RUN apt-get install tzdata in the Dockerfile.
.
├── BUILDING.md
├── examples/
├── images/
├── LICENSE
├── MANIFEST.in
├── moesifwsgi/
├── README.md
├── requirements.txt
├── setup.cfg
└── setup.py
The following sections describe the available configuration options for this middleware. You can set these options in a Python dictionary and then pass that as a parameter when you create the middleware instance. See the examples/ folder for better understanding.
Notice the following about the configuration options:
app is the original WSGI app instance.environ is a WSGI environ.Moesif also adds the following keys to environ:
environ['moesif.request_body']: a JSON object or base64 encoded string if couldn't parse the request body as JSON
environ["moesif.response_body_chunks"]: a response body chunks
environ["moesif.response_headers"]: a dictionary representing the response headers
APPLICATION_ID| Data type |
|---|
| String |
A string that identifies your application in Moesif.
SKIP| Data type | Parameters | Return type |
|---|---|---|
| Function |
(app, environ)
| Boolean |
Optional.
A function that takes a WSGI application and an environ object,
and returns True if you want to skip this particular event.
IDENTIFY_USER| Data type | Parameters | Return type |
|---|---|---|
| Function |
(app, environ, response_headers)
| String |
Optional, but highly recommended.
A function with the following arguments:
environ objectReturns returns a string that represents the user ID used by your system.
Moesif identifies users automatically. However, due to the differences arising from different frameworks and implementations, provide this function to ensure user identification properly.
IDENTIFY_COMPANY| Data type | Parameters | Return type |
|---|---|---|
| Function |
(app, environ, response_headers)
| String |
Optional.
A function with the following arguments:
environ objectReturns a string that represents the company ID for this event.
GET_METADATA| Data type | Parameters | Return type |
|---|---|---|
| Function |
(app, environ)
| Dictionary |
Optional.
A function that takes a WSGI application and an environ object, and
returns a dictionary.
This function allows you to add custom metadata that Moesif can associate with the event. The metadata must be a simple Python dictionary that can be converted to JSON.
For example, you may want to save a virtual machine instance ID, a trace ID, or a resource ID with the request.
GET_SESSION_TOKEN| Data type | Parameters | Return type |
|---|---|---|
| Function |
(app, environ)
| String |
Optional.
A function that takes a WSGI application and an environ, and returns a string that represents the session token for this event.
Similar to users and companies, Moesif tries to retrieve session tokens automatically. But if it doesn't work for your service, provide this function to help identify sessions.
MASK_EVENT_MODEL| Data type | Parameters | Return type |
|---|---|---|
| Function |
(EventModel)
|
EventModel
|
Optional.
A function that takes the final Moesif event model and returns an event model with desired data removed.
The return value must be a valid eventt model required by Moesif data ingestion API. For more information about the EventModel object, see the Moesif Python API documentation.
DEBUG| Data type |
|---|
| Boolean |
Optional.
Set to True to print debug logs if you're having integration issues.
LOG_BODY| Data type | Default |
|---|---|
| Boolean |
True
|
Optional.
Whether to log request and response body to Moesif.
EVENT_QUEUE_SIZE| Data type | Default |
|---|---|
int
|
1000_000
|
Optional.
The maximum number of event objects queued in memory pending upload to Moesif. For a full queue, additional calls to MoesifMiddleware returns immediately without logging the event. Therefore, set this option based on the event size and memory capacity you expect.
EVENT_WORKER_COUNT| Data type | Default |
|---|---|
int
|
2
|
Optional.
The number of worker threads to use for uploading events to Moesif.
If you have a large number of events being logged, increasing this number can improve upload performance.
BATCH_SIZE| Data type | Default |
|---|---|
int
|
100
|
An optional field name that specifies the maximum batch size when sending to Moesif.
EVENT_BATCH_TIMEOUT| Data type | Default |
|---|---|
int
|
1
|
Optional.
Maximum time in seconds to wait before sending a batch of events to Moesif when reading from the queue.
AUTHORIZATION_HEADER_NAME| Data type | Default |
|---|---|
| String |
authorization
|
Optional.
A request header field name used to identify the User in Moesif. It also supports a comma separated string. Moesif checks headers in order like "X-Api-Key,Authorization".
AUTHORIZATION_USER_ID_FIELD| Data type | Default |
|---|---|
| String |
sub
|
Optional.
A field name used to parse the user from authorization header in Moesif.
BASE_URI| Data type |
|---|
| String |
Optional.
A local proxy hostname when sending traffic through secure proxy. Remember to set this field when using secure proxy. For more information, see Secure Proxy documentation..
The following options apply to outgoing API calls. These are calls you initiate using the Python Requests library to third parties like Stripe or to your own services.
Several options use request and response as input arguments. These correspond to the Requests library's request or response objects.
If you are not using WSGI, you can import moesifpythonrequest directly.
CAPTURE_OUTGOING_REQUESTS| Data type | Default |
|---|---|
| Boolean |
False
|
Set to True to capture all outgoing API calls.
GET_METADATA_OUTGOING| Data type | Parameters | Return type |
|---|---|---|
| Function |
(req, res)
| Dictionary |
Optional.
A function that enables you to return custom metadata associated with the logged API calls.
Takes in the Requests request and response objects as arguments.
We recommend that you implement a function that returns a dictionary containing your custom metadata. The dictionary must be a valid one that can be encoded into JSON. For example, you may want to save a virtual machine instance ID, a trace ID, or a resource ID with the request.
SKIP_OUTGOING| Data type | Parameters | Return type |
|---|---|---|
| Function |
(req, res)
| Boolean |
Optional.
A function that takes a Requests request and response objects,
and returns True if you want to skip this particular event.
IDENTIFY_USER_OUTGOING| Data type | Parameters | Return type |
|---|---|---|
| Function |
(req, res)
| String |
Optional, but highly recommended.
A function that takes Requests request and response objects, and returns a string that represents the user ID used by your system.
While Moesif tries to identify users automatically, different frameworks and your implementation might vary. So we highly recommend that you accurately provide a user ID using this function.
IDENTIFY_COMPANY_OUTGOING| Data type | Parameters | Return type |
|---|---|---|
| Function |
(req, res)
| String |
Optional.
A function that takes Requests request and response objects, and returns a string that represents the company ID for this event.
GET_SESSION_TOKEN_OUTGOING| Data type | Parameters | Return type |
|---|---|---|
| Function |
(req, res)
| String |
Optional.
A function that takes Requests request and response objects, and returns a string that corresponds to the session token for this event.
Similar to user IDs, Moesif tries to get the session token automatically. However, if you setup differs from the standard, this function can help tying up events together and help you replay the events.
LOG_BODY_OUTGOING| Data type | Default |
|---|---|
| Boolean |
True
|
Optional.
Set to False to remove logging request and response body.
See the examples/ directory for example applications using Flask, Falcon, and Bottle frameworks.
Here's a Flask example:
def identify_user(app, environ, response_headers=dict()):
# Your custom code that returns a user id string
return "12345"
def identify_company(app, environ, response_headers=dict()):
# Your custom code that returns a company id string
return "67890"
def should_skip(app, environ):
# Your custom code that returns true to skip logging
return "health/probe" in environ.get('PATH_INFO', '')
def get_token(app, environ):
# If you don't want to use the standard WSGI session token,
# add your custom code that returns a string for session/API token
return "XXXXXXXXXXXXXX"
def mask_event(eventmodel):
# Your custom code to change or remove any sensitive fields
if 'password' in eventmodel.response.body:
eventmodel.response.body['password'] = None
return eventmodel
def get_metadata(app, environ):
return {
'datacenter': 'westus',
'deployment_version': 'v1.2.3',
}
moesif_settings = {
'APPLICATION_ID': 'YOUR_MOESIF_APPLICATION_ID',
'DEBUG': False,
'LOG_BODY': True,
'IDENTIFY_USER': identify_user,
'IDENTIFY_COMPANY': identify_company,
'GET_SESSION_TOKEN': get_token,
'SKIP': should_skip,
'MASK_EVENT_MODEL': mask_event,
'GET_METADATA': get_metadata,
'CAPTURE_OUTGOING_REQUESTS': False
}
app.wsgi_app = MoesifMiddleware(app.wsgi_app, moesif_settings)
The following examples demonstrate how to add and update customer information.
To create or update a user profile in Moesif, use the update_user() function.
api_client = MoesifAPIClient("Your Moesif Application Id").api
# Only user_id is required.
# Campaign object is optional, but useful if you want to track ROI of acquisition channels
# See https://www.moesif.com/docs/api#users for campaign schema
# metadata can be any custom object
user = {
'user_id': '12345',
'company_id': '67890', # If set, associate user with a company object
'campaign': {
'utm_source': 'google',
'utm_medium': 'cpc',
'utm_campaign': 'adwords',
'utm_term': 'api+tooling',
'utm_content': 'landing'
},
'metadata': {
'email': 'john@acmeinc.com',
'first_name': 'John',
'last_name': 'Doe',
'title': 'Software Engineer',
'sales_info': {
'stage': 'Customer',
'lifetime_value': 24000,
'account_owner': 'mary@contoso.com'
},
}
}
update_user = api_client.update_user(user)
The metadata field can contain any customer demographic or other info you want to store. Moesif only requires the user_id field.
For more information, see the function documentation in Moesif Python API Reference.
To update a list of users in one batch, use the update_users_batch() function.
api_client = MoesifAPIClient("Your Moesif Application Id").api
userA = {
'user_id': '12345',
'company_id': '67890', # If set, associate user with a company object
'metadata': {
'email': 'john@acmeinc.com',
'first_name': 'John',
'last_name': 'Doe',
'title': 'Software Engineer',
'sales_info': {
'stage': 'Customer',
'lifetime_value': 24000,
'account_owner': 'mary@contoso.com'
},
}
}
userB = {
'user_id': '54321',
'company_id': '67890', # If set, associate user with a company object
'metadata': {
'email': 'mary@acmeinc.com',
'first_name': 'Mary',
'last_name': 'Jane',
'title': 'Software Engineer',
'sales_info': {
'stage': 'Customer',
'lifetime_value': 48000,
'account_owner': 'mary@contoso.com'
},
}
}
update_users = api_client.update_users_batch([userA, userB])
The metadata field can contain any customer demographic or other info you want to store. Moesif only requires the user_id field.
For more information, see the function documentation in Moesif Python API Reference.
To update a single company, use the update_company() function.
api_client = MoesifAPIClient("Your Moesif Application Id").api
# Only company_id is required.
# Campaign object is optional, but useful if you want to track ROI of acquisition channels
# See https://www.moesif.com/docs/api#update-a-company for campaign schema
# metadata can be any custom object
company = {
'company_id': '67890',
'company_domain': 'acmeinc.com', # If domain is set, Moesif will enrich your profiles with publicly available info
'campaign': {
'utm_source': 'google',
'utm_medium': 'cpc',
'utm_campaign': 'adwords',
'utm_term': 'api+tooling',
'utm_content': 'landing'
},
'metadata': {
'org_name': 'Acme, Inc',
'plan_name': 'Free',
'deal_stage': 'Lead',
'mrr': 24000,
'demographics': {
'alexa_ranking': 500000,
'employee_count': 47
},
}
}
update_company = api_client.update_company(company)
The metadata field can contain any company demographic or other information you want to store. Moesif only requires the company_id field. For more information, see the function documentation in Moesif Python API Reference.
To update a list of companies in one batch, use the update_companies_batch() function.
api_client = MoesifAPIClient("Your Moesif Application Id").api
companyA = {
'company_id': '67890',
'company_domain': 'acmeinc.com', # If domain is set, Moesif will enrich your profiles with publicly available info
'metadata': {
'org_name': 'Acme, Inc',
'plan_name': 'Free',
'deal_stage': 'Lead',
'mrr': 24000,
'demographics': {
'alexa_ranking': 500000,
'employee_count': 47
},
}
}
companyB = {
'company_id': '09876',
'company_domain': 'contoso.com', # If domain is set, Moesif will enrich your profiles with publicly available info
'metadata': {
'org_name': 'Contoso, Inc',
'plan_name': 'Free',
'deal_stage': 'Lead',
'mrr': 48000,
'demographics': {
'alexa_ranking': 500000,
'employee_count': 53
},
}
}
update_companies = api_client.update_companies_batch([companyA, companyB])
The metadata field can contain any company demographic or other information you want to store. Moesif only requires the company_id field. For more information, see the function documentation in Moesif Python API Reference.
If you face any issues using this middleware, try the troubleshooting guidelines. For further assistance, reach out to our support team.
Explore other integration options from Moesif:
FAQs
Moesif Middleware for Python WSGI based platforms (Flask, Bottle & Others)
We found that moesifwsgi demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 3 open source maintainers collaborating on the project.
Did you know?

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Research
/Security News
Socket researchers discovered nine malicious NuGet packages that use time-delayed payloads to crash applications and corrupt industrial control systems.

Security News
Socket CTO Ahmad Nassri discusses why supply chain attacks now target developer machines and what AI means for the future of enterprise security.

Security News
Learn the essential steps every developer should take to stay secure on npm and reduce exposure to supply chain attacks.