dependency_overrides does not override dependency - python

The following FastApi test should use my get_mock_db function instead of the get_db function, but it dosen't. Currently the test fails because it uses the real Database.
def get_mock_db():
example_todo = Todo(title="test title", done=True, id=1)
class MockDb:
def query(self, _model):
mock = Mock()
mock.get = lambda _param: example_todo
def all(self):
return [example_todo]
def add(self):
pass
def commit(self):
pass
def refresh(self, todo: CreateTodo):
return Todo(title=todo.title, done=todo.done, id=1)
return MockDb()
client = TestClient(app)
app.dependency_overrides[get_db] = get_mock_db
def test_get_all():
response = client.get("/api/v1/todo")
assert response.status_code == 200
assert response.json() == [
{
"title": "test title",
"done": True,
"id": 1,
}
]

Key is to understand that dependency_overrides is just a dictionary. In order to override something, you need to specify a key that matches the original dependency.
def get_db():
return {'db': RealDb()}
def home(commons: dict= Depends(get_db))
commons['db'].doStuff()
app.dependency_overrides[get_db] = lambda: {'db': MockDb()}
Here you have inside the Depends function call a reference to get_db function. Then you are referring to the exact same function with dependency_overrides[get_db]. Therefore it gets overridden. Start by verifying that 'xxx' in these two match exactly: Depends(xxx) and dependency_overrides[xxx].
It took some time to wrap my head around the fact that whatever is inside the Depends call is actually the identifier for the dependency. So in this example the identifier is function get_db and the same function is used as key in the dictionary.
So this means the following example does not work since you are overriding something else than what's specified for Depends.
def get_db(connection_string):
return {'db': RealDb(connection_string)}
def home(commons: dict= Depends(get_db(os.environ['connectionString']))
commons['db'].doStuff()
# Does not work
app.dependency_overrides[get_db] = lambda: {'db': MockDb()}

Related

Python - list comprehension as a decorator (including self)

I have two functions:
job_status is getting a response from boto3 api.
jobs_detailsis a list comprehension that performs job_status on each element of the input list.
I want to change jobs_details into a decorator of jobs_status but below solutions throws inner() takes 1 positional argument but 2 were given error.
Appreciate any comment/alternative approach to my issue. Thanks!
import boto3
class GlueClient:
def __init__(self):
self.glue_client = boto3.client('glue')
#self.envs = envs
def jobs_list(self):
response = self.glue_client.list_jobs()
result = response["JobNames"]
while "NextToken" in response:
response = self.glue_client.list_jobs(NextToken=response["NextToken"])
result.extend(response["JobNames"])
return [e for e in result if "jobs_xyz" in e]
#WHAT IS CURRENTLY
def job_status(self, job_name):
paginator = self.glue_client.get_paginator('get_job_runs')
response = paginator.paginate(JobName=job_name)
return response
def jobs_details(self, jobs):
return [self.job_status(e) for e in jobs]
#WHAT IS EXPECTED
def pass_by_list_comprehension(func):
def inner(list_of_val):
return [func(value) for value in list_of_val ]
return inner
#pass_by_list_comprehension
def job_status(self, job_name):
paginator = self.glue_client.get_paginator('get_job_runs')
response = paginator.paginate(JobName=job_name)
return response
glue_client = GlueClient()
jobs = glue_client.jobs_list()
jobs_status = glue_client.job_status(jobs)
print(jobs)
You want something like:
import boto3
from typing import Callable
def handle_iterable_input(func):
def inner(self, list_of_val):
return [func(self, value) for value in list_of_val]
return inner
class GlueClient:
def __init__(self):
self.glue_client = boto3.client('glue')
#self.envs = envs
def jobs_list(self):
response = self.glue_client.list_jobs()
result = response["JobNames"]
while "NextToken" in response:
response = self.glue_client.list_jobs(NextToken=response["NextToken"])
result.extend(response["JobNames"])
return [e for e in result if "jobs_xyz" in e]
#handle_iterable_input
def job_status(self, job_name):
paginator = self.glue_client.get_paginator('get_job_runs')
response = paginator.paginate(JobName=job_name)
return response
glue_client = GlueClient()
jobs = glue_client.jobs_list()
jobs_status = glue_client.job_status(jobs)
print(jobs)
This is the most basic way to make your decorator handle methods properly, by explicitly handling the passing of self. Note, it assumes the function being decorated will only take a single argument.
If all you want to do is make job_status iterate through a list of job names instead of operating on just one, something like this should work:
def jobs_status(self, job_names):
paginator = self.glue_client.get_paginator('get_job_runs')
return [paginator.paginate(JobName=job_name) for job_name in job_names]
Using a decorator to change what parameters a method expects seems like a bad idea.
Also, naming your class GlueClient would imply that it is a glue client. The fact that it has an attribute named glue_client makes me suspect you could probably choose a clearer name for one or both of them. (However, I'm not familiar with the package you're using.)

How to call the right function (as a string) based on an argument?

I have a class which is intended to create an IBM Cloud Object Storage object. There are 2 functions I can use for initialization : resource() and client(). In the init function there is an object_type parameter which will be used to decide which function to call.
class ObjectStorage:
def __init__(self, object_type: str, endpoint: str, api_key: str, instance_crn: str, auth_endpoint: str):
valid_object_types = ("resource", "client")
if object_type not in valid_object_types:
raise ValueError("Object initialization error: Status must be one of %r." % valid_object_types)
method_type = getattr(ibm_boto3, object_type)()
self._conn = method_type(
"s3",
ibm_api_key_id = api_key,
ibm_service_instance_id= instance_crn,
ibm_auth_endpoint = auth_endpoint,
config=Config(signature_version="oauth"),
endpoint_url=endpoint,
)
#property
def connect(self):
return self._conn
If I run this, I receive the following error:
TypeError: client() missing 1 required positional argument: 'service_name'
If I use this in a simple function and call it by using ibm_boto3.client() or ibm_boto3.resource(), it works like a charm.
def get_cos_client_connection():
COS_ENDPOINT = "xxxxx"
COS_API_KEY_ID = "yyyyy"
COS_INSTANCE_CRN = "zzzzz"
COS_AUTH_ENDPOINT = "----"
cos = ibm_boto3.client("s3",
ibm_api_key_id=COS_API_KEY_ID,
ibm_service_instance_id=COS_INSTANCE_CRN,
ibm_auth_endpoint=COS_AUTH_ENDPOINT,
config=Config(signature_version="oauth"),
endpoint_url=COS_ENDPOINT
)
return cos
cos = get_cos_client_connection()
It looks like it calls the client function on this line, but I am not sure why:
method_type = getattr(ibm_boto3, object_type)()
I tried using:
method_type = getattr(ibm_boto3, lambda: object_type)()
but it was a silly move.
The client function looks like this btw:
def client(*args, **kwargs):
"""
Create a low-level service client by name using the default session.
See :py:meth:`ibm_boto3.session.Session.client`.
"""
return _get_default_session().client(*args, **kwargs)
which refers to:
def client(self, service_name, region_name=None, api_version=None,
use_ssl=True, verify=None, endpoint_url=None,
aws_access_key_id=None, aws_secret_access_key=None, aws_session_token=None,
ibm_api_key_id=None, ibm_service_instance_id=None, ibm_auth_endpoint=None,
auth_function=None, token_manager=None,
config=None):
return self._session.create_client(
service_name, region_name=region_name, api_version=api_version,
use_ssl=use_ssl, verify=verify, endpoint_url=endpoint_url,
aws_access_key_id=aws_access_key_id,
aws_secret_access_key=aws_secret_access_key,
aws_session_token=aws_session_token,
ibm_api_key_id=ibm_api_key_id, ibm_service_instance_id=ibm_service_instance_id,
ibm_auth_endpoint=ibm_auth_endpoint, auth_function=auth_function,
token_manager=token_manager, config=config)
Same goes for resource()
If you look at the stracktrace, it will probably point to this line:
method_type = getattr(ibm_boto3, object_type)()
And not the one after where you actually call it. The reason is simple, those last two parenthese () mean you're calling the function you just retrieved via getattr.
So simply do this:
method_type = getattr(ibm_boto3, object_type)
Which means that method_type is actually the method from the ibm_boto3 object you're interested in.
Can confirm that by either debugging using import pdb; pdb.set_trace() and inspect it, or just add a print statement:
print(method_type)

Pytest Fixtures - Parameterisation - Call Fixture Once

I have a fixture that returns the endpoint for the name of that endpoint (passed in)
The name is a string set in the test. I have messed up by calling the endpoint each time in the tests (parameterised) and now I can't figure out how to get the same functionality working without calling the endpoint each time.
Basically I just need to call the endpoint once and then pass that data between all my tests in that file (Ideally without anything like creating a class and calling it in the test. I have about 12 files each with similar tests and I want to reduce the boiler plate. Ideally if it could be done at the fixture/parametrisation level with no globals.
Here's what I have so far:
#pytest.mark.parametrize('field', [('beskrivelse'), ('systemId')])
def test_intgra_001_elevforhold_req_fields(return_endpoint, field):
ep_to_get = 'get_elevforhold'
ep_returned = return_endpoint(ep_to_get)
apiv2 = Apiv2()
apiv2.entity_check(ep_returned, field, ep_to_get, False)
#pytest.fixture()
def return_endpoint():
def endpoint_initialisation(ep_name):
apiv2 = Apiv2()
ep_data = apiv2.get_ep_name(ep_name)
response = apiv2.get_endpoint_local(ep_data, 200)
content = json.loads(response.content)
apiv2.content_filt(content)
apiv2_data = content['data']
return apiv2_data
return endpoint_initialisation
Create return_endpoint as a fixture with scope session and store data in a dictionary after it is fetched. The fixture doesn't return the initialization function, but a function to access the dictionary.
#pytest.mark.parametrize('field', [('beskrivelse'), ('systemId')])
def test_intgra_001_elevforhold_req_fields(return_endpoint, field):
ep_to_get = 'get_elevforhold'
ep_returned = return_endpoint(ep_to_get)
apiv2 = Apiv2()
apiv2.entity_check(ep_returned, field, ep_to_get, False)
#pytest.fixture(scope='session')
def return_endpoint():
def endpoint_initialisation(ep_name):
apiv2 = Apiv2()
ep_data = apiv2.get_ep_name(ep_name)
response = apiv2.get_endpoint_local(ep_data, 200)
content = json.loads(response.content)
apiv2.content_filt(content)
apiv2_data = content['data']
return apiv2_data
ep_data = dict()
def access(ep_name):
try:
return ep_data[ep_name] # or use copy.deepcopy
except KeyError:
ep_data[ep_name] = endpoint_initialisation(ep_name)
return ep_data[ep_name] # or use copy.deepcopy
return access
There are some caveats here. If the object returned by endpoint_initialisation() is mutable, then you potentially create unwanted dependencies between your tests. You can avoid this by returning a (deep) copy of the object. You can use the copy module for that.

Python client authentication with decorators

I want to build a python client on top of a REST API that uses authentication with a api_token. Hence all api calls require the api_token. As it is pretty ugly to add a field
'token=...'
e.g.
a = f1(5, token='token')
b = f2(6, 12, token='token')
c = f3(2, 'a', token='token')
where internally f1 and f2 delegate to the REST api
to each function call. What I would like to have is something like:
auth = authenticate('token')
a = f1(5)
b = f2(6, 12,)
c = f3(2, 'a')
What I can do is to create a class and make all functions member functions. Hence, we would have:
auth = calculator('token')
a = auth.f1(5)
b = auth.f2(6, 12,)
c = auth.f3(2, 'a')
but that would also be somewhat ugly. I am trying to get this to work with decorators, but to no avail so far.
class authenticate:
def __init__(self, token):
self.token = token
def __call__(self, func):
def functor(*args, **kwargs):
return func(*args, **kwargs, key=self.authentication)
return functor
#authenticate
def f1(a, key):
data = a
result = requests.get(1, data, key)
return result
However, this seems to be going nowhere. I am also wondering whether this might work at all as decorators are executed at import time and the token is added at runtime.
Any suggestions on how to make this work or anyone know if there is another standard pattern for this?
So after some hacking around we came up with the following:
class authenticate:
# start empty key
key = None
#classmethod
""" add the token """
def set_key(cls, token):
cls.token = token
def __init__(self, func=None):
if func is not None:
self.func = func
else:
print('no function')
def __call__(self, *arg):
"""
add authentication to function func
"""
ret = self.func(*arg, auth_key=self.key)
return ret
#authenticate
def f1(a, key):
data = a
result = requests.get(1, data, key)
return result
Then you can run code like:
authentication_key = 'token'
print('Initiate class')
authenticate().set_key(key=authentication_key)
print('Run f1(5)')
a1 = f1(5) # no token needed!
a2 = f2(6, 12) # again no token needed as it is in the decorator
print(a1)
This works more or less as I hoped and I find it cleaner than the class methods. If anyone has a better suggestion or improvements let me know.

flask: How to store and retrieve POST calls?

#app.route('/path/<user>/<id>', methods=['POST'])
#cache.cached(key_prefix='/path/<user>', unless=True)
def post_kv(user, id):
cache.set(user, id)
return value
#app.route('/path/<user>', methods=['GET'])
#cache.cached(key_prefix='/path/<user>', unless=True)
def getkv(user):
cache.get(**kwargs)
I want to be able to make POST calls to the path described, store them, and GET their values from the user. The above code runs, but has bugs and doesn't perform as needed. Frankly, the flask-cache docs aren't helping. How can I properly implement cache.set and cache.get to perform as needed?
In Flask, implementing your custom cache wrapper is very simple.
from werkzeug.contrib.cache import SimpleCache, MemcachedCache
class Cache(object):
cache = SimpleCache(threshold = 1000, default_timeout = 100)
# cache = MemcachedCache(servers = ['127.0.0.1:11211'], default_timeout = 100, key_prefix = 'my_prefix_')
#classmethod
def get(cls, key = None):
return cls.cache.get(key)
#classmethod
def delete(cls, key = None):
return cls.cache.delete(key)
#classmethod
def set(cls, key = None, value = None, timeout = 0):
if timeout:
return cls.cache.set(key, value, timeout = timeout)
else:
return cls.cache.set(key, value)
#classmethod
def clear(cls):
return cls.cache.clear()
...
#app.route('/path/<user>/<id>', methods=['POST'])
def post_kv(user, id):
Cache.set(user, id)
return "Cached {0}".format(id)
#app.route('/path/<user>', methods=['GET'])
def get_kv(user):
id = Cache.get(user)
return "From cache {0}".format(id)
Also, the simple memory cache is for single process environments and the SimpleCache class exists mainly for the development server and is not 100% thread safe. In production environments you should use MemcachedCache or RedisCache.
Make sure you implement logic when item is not found in the cache.

Categories