Tables are not created for test_database while running behave test

[SALIMA534]

We are using Behave tool for running BDD tests and when we run the tests using Behave only the tests are running fine. Here is an example to show how we are running test

behave login.feature

When we use the 'behave-django' to integrate BDD tests with Django, tests can't be run. We followed the steps available from these pages
https://behave-django.readthedocs.io/en/latest/installation.html
&
https://behave-django.readthedocs.io/en/latest/usage.html

and the error what we are hitting is like this:
Option 1:

python manage.py behave
.........
..........
Creating test database for alias 'default'...
Traceback (most recent call last):
  File "C:\Users\USER 1\PycharmProjects\behaveproject\venv\lib\site-packages\django\db\backends\utils.py", line 84, in _execute
    return self.cursor.execute(sql, params)
psycopg2.errors.UndefinedTable: relation "auth_user" does not exist
The above exception was the direct cause of the following exception:
.............
.............
django.db.utils.ProgrammingError: relation "auth_user" does not exist

Option 2:

python manage.py behave --keepdb
.........
.........
Using existing test database for alias 'default'...
Traceback (most recent call last):
  File "C:\Users\USER 1\PycharmProjects\behaveproject\venv\lib\site-packages\django\db\backends\utils.py", line 84, in _execute
    return self.cursor.execute(sql, params)
psycopg2.errors.UndefinedTable: relation "auth_user" does not exist
The above exception was the direct cause of the following exception:
..........
..........
django.db.utils.ProgrammingError: relation "auth_user" does not exist

This is really blocking our deployments because tests can't be run. Could someone look at this issue and help us unblocking it.

[bittner]

It runs with just behave but not with python manage.py behave? Sounds like an issue with your Behave setup. Not with behave-django.

Do you have some special code in your environment.py, maybe, that makes the tests work? You have to share more details about your setup.

As an alternative, you could start with a simple setup from the tutorial, convince yourself that it's working, and add parts of your current setup to it until the test execution breaks.

[SALIMA534]

Thank you @bittner. Yes, like I mentioned before running the tests just with behave works. As per document steps for behave-django we have to run python manage.py behave. Please check the instructions from this page

https://behave-django.readthedocs.io/en/latest/usage.html

Yes, To start browser we added code in environment.py ,

from selenium import webdriver

def before_all(context):
    # PhantomJS is used there (headless browser - meaning we can execute tests in a command-line environment, which is what we want for use with SemaphoreCI
    # For debugging purposes, you can use the Firefox driver instead.
    context.driver = webdriver.Chrome()

    # context.browser = webdriver.PhantomJS()
    context.browser.implicitly_wait(1)
    context.server_url = 'http://127.0.0.1:8000'


def after_all(context):
    # Explicitly quits the browser, otherwise it won't once tests are done
    context.browser.quit()


def before_feature(context, feature):
    # Code to be executed each time a feature is going to be tested
    pass

[bittner]

Is there a reason why you set context.server_url? What do you do with that value?

Can you try to use context.get_url() instead, as described in the instructions in the manual? Django's test client knows the server URL that the Django runserver uses.

Please, try to follow the manual closely, and only when that works fine, only then go about integrating Selenium (if you have to).

Aside, the description of your setup is still incomplete. You have to provide a minimal, reproducible example, otherwise it's very hard to help you.