Feature reusable normal user for tests

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰

[ebreton]

The main feature is the added fixture normaluser_token_headers which goes as a companion for the existing fixture superuser_token_headers.

The fixture does not create a brand new user everytime it is called, but the same user with its email defined from config.EMAIL_TEST_USER. This allows to limit the number of random users created in the database, and provides a control on (at least) one normal user.

Behind the hoods, the password for this 'pre-defined' user is reseted every time it is needed, which keeps the security up to the standards (no password needs to be defined anywhere for this particular test configuration)

As a side bonus, the function byemail_authentication_token(email) can be used to get a valid authorization token for the sole given argument of an email (changing the password on its way though)

[tiangolo]

Thanks! I'll review it soon.

[wassname]

Thanks for this. I'm not suggesting you do it this way, but another possibility to acheive this is to use pytest fixtures and yield to clean up the created users. It can sometimes lead to cleaner code and a cleaner db with less effort.

fromfakerimportFakerfaker=Faker()
@pytest.fixture(scope="module")defrandom_user():
password=faker.password()
user_in=UserCreate(username=faker.name(), email=faker.email(), password=password)
user=crud.user.create(
db_session=db_session, user_in=user_in
)
# Attach the fake password for testing purposesuser.password=passwordyielduser# clean up fake user after tests pass or failcrud.user.remove(db_session=db_session, id=user.id)

@pytest.fixture(scope="module")defrandom_user_auth_headers(random_user):
# one fixture can reuse anotherreturnuser_authentication_headers(get_server_api(), random_user.email, random_user.password)

[tiangolo]

Thank you @ebreton for your contribution! 🚀 🎉 🍰