.................................................FF...
======================================================================
FAIL: test_assert_not_process_the_template (tests.test_utils.TestRenderTemplates)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/mnt/gen2/TmpDir/portage/dev-python/flask-testing-0.4.2/work/Flask-Testing-0.4.2/tests/test_utils.py", line 192, in test_assert_not_process_the_template
    assert "" != response.data
AssertionError

======================================================================
FAIL: test_assert_the_real_render_template_is_restored (tests.test_utils.TestRestoreTheRealRender)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/mnt/gen2/TmpDir/portage/dev-python/flask-testing-0.4.2/work/Flask-Testing-0.4.2/tests/test_utils.py", line 209, in test_assert_the_real_render_template_is_restored
    assert "" != response.data
AssertionError

----------------------------------------------------------------------
Ran 54 tests in 5.295s

FAILED (failures=2)

Occurs under python2.7 and pypy. CAn you replicate? Do you require anything further?

When the test client is named self.app, i.e. you have a situation like

    def setUp(self):
        self.app = application.test_client()

Then when assertRedirects tries to access the Flask instance at self.app it instead gets the FlaskClient instance and crashes as below.

Traceback (most recent call last):
  File "/usr/local/Cellar/python/2.7.9/Frameworks/Python.framework/Versions/2.7/lib/python2.7/unittest/case.py", line 329, in run
    testMethod()
  File "/Users/bistenes/Code/socos/jive/server/src/tests/test_pages.py", line 118, in test_logs_in_user
    self.assertRedirects(rv, '/manage')
  File "/Users/bistenes/Code/socos/jive/server/src/lib/flask_testing/utils.py", line 278, in assertRedirects
    server_name = self.app.config.get('SERVER_NAME') or 'localhost'
AttributeError: 'FlaskClient' object has no attribute 'config'

I'm pretty surprised this wasn't a problem before 0.6.0 -- you would think that resolving self.app elsewhere, e.g. for self.app.get(...) would also be a problem, but it apparently wasn't. The obvious solution is to not name the test client self.app, but rather, say, self.client.

This is mostly a request for documentation -- somewhere it should advise against doing this. It's also, of course, a note to those running into this confusing error in the future. Actual code change is probably not called for here.

If no messages are flashed to compare against, then assertMessageFlashed will always fail silently without raising AssertionError.

I'm getting the error below when I use to flask.ext.testing.LiveServerTestCase class with python 3.3

Traceback (most recent call last):
  File "C:\Users\Reuber\env_stormy_flask\lib\site-packages\flask_testing\utils.py", line 345, in __call__
    self._spawn_live_server()
  File "C:\Users\Reuber\env_stormy_flask\lib\site-packages\flask_testing\utils.py", line 366, in _spawn_live_server
    self._process.start()
  File "C:\Python33\Lib\multiprocessing\process.py", line 111, in start
    self._popen = Popen(self)
  File "C:\Python33\Lib\multiprocessing\forking.py", line 248, in __init__
    dump(process_obj, to_child, HIGHEST_PROTOCOL)
  File "C:\Python33\Lib\multiprocessing\forking.py", line 166, in dump
    ForkingPickler(file, protocol).dump(obj)
_pickle.PicklingError: Can't pickle <class '_thread.lock'>: attribute lookup _thread.lock failed

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "C:\Python33\Lib\runpy.py", line 160, in _run_module_as_main
    "__main__", fname, loader, pkg_name)
  File "C:\Python33\Lib\runpy.py", line 73, in _run_code
    exec(code, run_globals)
  File "C:\Python33\Lib\unittest\__main__.py", line 19, in <module>
    main(module=None)
  File "C:\Python33\Lib\unittest\main.py", line 125, in __init__
    self.runTests()
  File "C:\Python33\Lib\unittest\main.py", line 265, in runTests
    self.result = testRunner.run(self.test)
  File "C:\Python33\Lib\unittest\runner.py", line 168, in run
    test(result)
  File "C:\Python33\Lib\unittest\suite.py", line 67, in __call__
    return self.run(*args, **kwds)
  File "C:\Python33\Lib\unittest\suite.py", line 105, in run
    test(result)
  File "C:\Python33\Lib\unittest\suite.py", line 67, in __call__
    return self.run(*args, **kwds)
  File "C:\Python33\Lib\unittest\suite.py", line 105, in run
    test(result)
  File "C:\Python33\Lib\unittest\suite.py", line 67, in __call__
    return self.run(*args, **kwds)
  File "C:\Python33\Lib\unittest\suite.py", line 105, in run
    test(result)
  File "C:\Users\Reuber\env_stormy_flask\lib\site-packages\flask_testing\utils.py", line 348, in __call__
    self._terminate_live_server()
  File "C:\Users\Reuber\env_stormy_flask\lib\site-packages\flask_testing\utils.py", line 373, in _terminate_live_server
    self._process.terminate()
  File "C:\Python33\Lib\multiprocessing\process.py", line 119, in terminate
    self._popen.terminate()
AttributeError: 'NoneType' object has no attribute 'terminate'

I attempted to follow a workflow with logging in as described in the Flask testing documentation . It seems to not be persisting the user login. I am not sure if this has to do with the scoped sessions or perhaps some trick that I am not aware of.

My '/login' endpoint works normally when running the server. API calls are being executed with the @login_required decorator during functional testing, but using this framework I am unable to test my @login_required calls. Let me know if I am missing something. Thanks

This is probably related to the change in #24.

Using python 2.7.5 and Flask-Testing==0.4.1

tests.py

#!/usr/bin/env python2.7
import unittest

if __name__ == "__main__":
    suite = unittest.TestLoader().discover('tests')
    unittest.TextTestRunner(verbosity=2).run(suite)

./tests/testApi.py

from flask.ext.testing import TestCase

class TestApi(TestCase):

    def create_app(self):
        app = Flask(__name__)
        app.config['TESTING'] = True
        return app

    def test_server_is_up_and_running(self):
        pass

> ./tests.py

test_helloworld (testHelloWorld.TestHelloWorld) ... ok
Traceback (most recent call last):
  File "./tests.py", line 6, in <module>
    unittest.TextTestRunner(verbosity=2).run(suite)
  File "/usr/lib/python2.7/unittest/runner.py", line 151, in run
    test(result)
  File "/usr/lib/python2.7/unittest/suite.py", line 70, in __call__
    return self.run(*args, **kwds)
  File "/usr/lib/python2.7/unittest/suite.py", line 108, in run
    test(result)
  File "/usr/lib/python2.7/unittest/suite.py", line 70, in __call__
    return self.run(*args, **kwds)
  File "/usr/lib/python2.7/unittest/suite.py", line 108, in run
    test(result)
  File "/usr/lib/python2.7/unittest/suite.py", line 70, in __call__
    return self.run(*args, **kwds)
  File "/usr/lib/python2.7/unittest/suite.py", line 108, in run
    test(result)
  File "/home/bhafichuk/.virtualenvs/kyc/local/lib/python2.7/site-packages/flask_testing/utils.py", line 90, in __call__
    self._post_teardown()
  File "/home/bhafichuk/.virtualenvs/kyc/local/lib/python2.7/site-packages/flask_testing/utils.py", line 116, in _post_teardown
    if self._ctx is not None:
AttributeError: 'TestApi' object has no attribute '_ctx'

https://github.com/jarus/flask-testing/blob/master/flask_testing/utils.py#L209

This suppose you run test with localhost. In case one use subdomains, this doesn't work because subdomains required a FQDN like myserver.local.

Latest version of pip makes pip install flask-testing fail because the current 0.4 version (from 07/2012) on PyPi fail because twill is required and twill is not on PyPi. The latest version of Flask-Testing works fine without twill (if you don't use those features) and this can be fixed simply by providing the latest version of Flask-Testing on PyPi.

In Testing with SQLAlchemy section of the document:

def setUp(self):
        db.create_all()

will cause error RuntimeError: application not registered on db instance and no application bound to current context with class LiveServerTestCase

which can be fixed by:

def setUp(self):
    with self.app.app_context():
        db.create_all()

Twill alters sys.path
( http://twill.idyll.org/problems.html )
and put its other_packages directory in front of everything. When I import subprocess module, instead of stdlib it imports from twill. When I change the order of my imports it works fine. It shouldn't be that way.

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/8
• Originally Reported By: Cenk Alti
• Originally Created At: 2012-01-13 11:26:46

It's my first time doing Python testing.

In the docs, I saw the following at the end of a testing file:

if __name__ == '__main__':
unittest.main()

However, in the example given, it does not have anything like this, so I decided to not put the above code. Do I need this code to run this test? Or does it work without?

Thanks for the help.

The documentation does not specify how the tests are to be run. Perhaps this should be added.

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/6
• Originally Reported By: Kiran Jonnalagadda
• Originally Created At: 2011-03-09 15:46:07

This isn't a bug in LiveServerTestCase itself but with running the flask application using wekzeug with multiprocessing when the use_reloader option set to true. It can be reproduced with the following snippet

import unittest
import multiprocessing
import flask
import time
import random

class FlaskAppMultiProcTest(unittest.TestCase):
    def runTest(self):
        app = flask.Flask(__name__)
        proc = multiprocessing.Process(target=app.run
                , kwargs = {"use_reloader":True})
        proc.start()
        time.sleep(1)
        proc.terminate()

if __name__ == "__main__":
    unittest.main()

• Changing use_reloader to False results in single execution.
• Running the above sample will show output from the double execution.
• Replacing unittest.main() with FlaskAppMultiProcTest().runTest() will also result in double execution.
• Running the test with nosetests also results in double execution.

Not sure if this should be a comment, or a suggestion, but it's a bit lost on me, why a testing framework would not be displaying the testing status (I found them on travis-ci, they do github badges), or the code coverage; which doesn't seem to be generated on travis, or exist on coveralls. Would it be possible to request that this gets added?

Can we use mongodb instead of SQL? Does the framework support mongodb?

The documentation does not specify how the tests are to ne run. Perhaps this should be added.

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/5
• Originally Reported By: Kiran Jonnalagadda
• Originally Created At: 2011-03-09 15:46:00

Hi,

There is an error in the documentation of the assert_context and assertContext functions.

The rst parser didn't find the right param keyword and the rendering is bad.

Stéphane

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/1
• Originally Reported By: Stéphane Wirtel
• Originally Created At: 2010-07-27 11:07:11

I add this to my base TestCase:

import re
from functools import partial

def __getattr__(self, name):
    name = re.sub(r'_([a-z])', lambda m: m.group(1).upper(), name)
    return partial(getattr(TestCase, name), self)
                         # ^ The parent class

Lets me do self.assert_equals etc. Adding it should also mean you can drop all the internal snake case aliases. Doesn't give you API documentation though you could add a note about this behaviour in the docs with some examples. Alternatively loop dir(TestCase) and dynamically alias any camel cased attributes, should take care of API docs.

I've tested that it avoids recursive stack exhaustion and that it properly raises an AttributeError when the camel case version doesn't exist. This is why we're doing getattr on the parent class and using partial to inject self.

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/2
• Originally Reported By: Dag Odenhall
• Originally Created At: 2010-10-19 14:51:09

It looks like a new, possibly buggy version of Twill was released on 2014-04-24, which doesn't seem to install cleanly with pip.

Until this is resolved, would it be possible to pin the twill requirement in this package to the previous version (0.9)?

My tests (without any actual use of Twill) appear to have all broken, and re-running pip install -r requirements.txt (which contains Flask-Testing) doesn't fix it.

One of my tests look like this:

def test_get_network_info(self):
    with patch('subprocess.check_output', Mock(return_value=values)):

And gives the following error:

======================================================================
ERROR: test_get_network_info (tests.test_tools.ToolsTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/tim/Documents/overseer/app/tests/test_tools.py", line 21, in test_get_network_info
    with patch('subprocess.check_output', Mock(return_value=mock_network_info.read())):
  File "/usr/local/lib/python2.7/dist-packages/mock.py", line 1268, in __enter__
    original, local = self.get_original()
  File "/usr/local/lib/python2.7/dist-packages/mock.py", line 1242, in get_original
    "%s does not have the attribute %r" % (target, name)
AttributeError: <module 'subprocess' from '/usr/local/lib/python2.7/dist-packages/twill/other_packages/subprocess.pyc'> does not have the attribute 'check_output'

My understanding of the problem is that mock is trying to mock twill's subprocess module instead of the python one.

Am I doing something wrong ? Should I try to patch the subprocess module differently ?

Here is the twill's version of subprocess wich looks like it has been copy pasted from an old version of python ?
https://github.com/ctb/twill/blob/master/twill/other_packages/subprocess.py

pallets/flask#1135

Can you please update the available version in PyPI? The last update was done in 2014 and there are some major fixes. I'm totally ok using a nasty-looking line in requirements.txt, but Flask-Testing==0.5.0 or 0.4.3 would be really nice :)

Thanks all!

Because of the pushed app context before each test case, multiple client requests within a test will see each other's (un-committed) sessions. This goes against the docs:

Another gotcha is that Flask-SQLAlchemy also removes the session instance at the end of every request (as should any threadsafe application using SQLAlchemy with scoped_session). Therefore the session is cleared along with any objects added to it every time you call client.get() or another client method.

Here's a test case that fails:

python
from flask import Flask
from flask.ext.sqlalchemy import SQLAlchemy
from flask.ext.testing import TestCase

app = Flask(name)
app.config['SQLALCHEMY_DATABASE_URI'] = 'sqlite:////tmp/test.db'
db = SQLAlchemy(app)
@app.route('/')
def index():
u = db.session.query(User).first()
u.name = 'bob'
return ''

class User(db.Model):
id = db.Column(db.Integer, primary_key=True)
name = db.Column(db.String)

class SessionTest(TestCase):

def create_app(self):
return app

def test_remove(self):
db.drop_all()
db.create_all()

u = User()
u.name = 'joe'
db.session.add(u)
db.session.commit()
client = app.test_client()
client.get('/')
assert u.name == 'joe' # fails!
assert u not in db.session # would fail if it reached here

assertContext, assertRedirects in class TestCase does not have a message parameter for customizing error message.
Since every other assert functions have them I see no reason not to implement it.

I'm trying to test with Flask-Testing a Flask-SQLAlchemy model. More precisely a static method of this model that uses first_or_404() and I cannot find a way to make my test work.

Here a self contained example that highlight the issue:

from flask import Flask
from flask_sqlalchemy import SQLAlchemy
from flask.ext.testing import TestCase

db = SQLAlchemy()

class ModelToTest(db.Model):
    __tablename__ = 'model_to_test'
    identifier = db.Column(db.String(80), unique=True, nullable=False, primary_key=True)

    @staticmethod
    def get_by_identifier(identifier):
        return ModelToTest.query.filter_by(identifier=identifier).first_or_404()

class Config:
    DEBUG = True
    TESTING = True
    SQLALCHEMY_DATABASE_URI = 'sqlite:///'
    SQLALCHEMY_TRACK_MODIFICATIONS = False

class TestGetByIdentifier(TestCase):

    def create_app(self):
        app = Flask('test')
        app.config.from_object(Config())
        db.init_app(app)
        return app

    def setUp(self):
        db.create_all()

    def tearDown(self):
        db.session.remove()
        db.drop_all()

    def test_get_by_identifier(self):
        self.assert404(ModelToTest.get_by_identifier('identifier'))

I got the error:

(my_env) PS C:\Dev\Test\Python\test_flask> nosetests-3.4.exe
E
======================================================================
ERROR: test_get_by_identifier (test_flask.TestGetByIdentifier)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Dev\Test\Python\test_flask\test_flask.py", line 37, in test_get_by_identifier
    self.assert404(ModelToTest.get_by_identifier('identifier'))
  File "C:\Dev\Test\Python\test_flask\test_flask.py", line 13, in get_by_identifier
    return ModelToTest.query.filter_by(identifier=identifier).first_or_404()
  File "c:\\my_env\lib\site-packages\flask_sqlalchemy\__init__.py", line 431, in first_or_404
    abort(404)
  File "c:\\my_env\lib\site-packages\werkzeug\exceptions.py", line 646, in __call__
    raise self.mapping[code](*args, **kwargs)
werkzeug.exceptions.NotFound: 404: Not Found

----------------------------------------------------------------------
Ran 1 test in 0.913s

So the line self.assert404(ModelToTest.get_by_identifier('identifier')) does generate an exception in the first_or_404() call and this exception is a werkzeug.exceptions.NotFound, it does not seems to be what's expected by self.assert404().

Requirements to run this test are:

• flask
• flask-sqlalchemy
• flask-testing

It is worth noting that when I use the function in the application it behaves as expected.

Thanks in advance.

I have a flask app which uses multiprocessing when computing stuff. When I add import flask.ext.testing at the top of my test module (and nothing else, not using it yet), I get this when I run my tests:

Traceback (most recent call last):
  File "mcbench_tests.py", line 5, in <module>
    import app
  File "/Users/isbadawi/code/py/mcbench/app.py", line 5, in <module>
    import mcbench.client
  File "/Users/isbadawi/code/py/mcbench/mcbench/client.py", line 4, in <module>
    from mcbench.benchmark import Benchmark, BenchmarkSet
  File "/Users/isbadawi/code/py/mcbench/mcbench/benchmark.py", line 5, in <module>
    import multiprocessing
  File "/usr/local/Cellar/python/2.7.5/Frameworks/Python.framework/Versions/2.7/lib/python2.7/multiprocessing/__init__.py", line 65, in <module>
    from multiprocessing.util import SUBDEBUG, SUBWARNING
  File "/usr/local/Cellar/python/2.7.5/Frameworks/Python.framework/Versions/2.7/lib/python2.7/multiprocessing/util.py", line 40, in <module>
    from subprocess import _args_from_interpreter_flags
ImportError: cannot import name _args_from_interpreter_flags

In _pre_setup (see commti 93c6365), a reference is set on the test case to the application object:

self.app = self.create_app()

But this reference is not removed on _post_teardown. That means that we're leaking applications in situations where test cases are kept track of (e.g. with nose).

It looks like Flask-Testing use assert which is bad

#!python

======================================================================
FAIL: test_login_post (tests.tests.LoginTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/amirouche/projects/flask-door/tests/tests.py", line 50, in test_login_post
    self.assertRedirects(response, '/')
  File "/home/amirouche/projects/flask-door/Flask_Testing-0.3-py2.7.egg/flaskext/testing.py", line 164, in assertRedirects
    assert response.location == "http://localhost" + location
AssertionError

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/7
• Originally Reported By: Amirouche Boubekki
• Originally Created At: 2011-07-24 17:10:33

Specifically I'm testing for status 400 for my web API, but others might be useful as well.

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/3
• Originally Reported By: Dag Odenhall
• Originally Created At: 2010-11-05 22:09:56

I have a test case in which I do a couple requests using the test client; after every request I need to check the context and I use the get_context_variable method. The problem I'm experiencing is that the value returned by this function is always the expected value from the first requests, subsequent requests don't affect the value returned by this method.

In the TestCase class, there's this method:

def __call__(self, result=None):
    """
    Does the required setup, doing it here
    means you don't have to call super.setUp
    in subclasses.
    """
    try:
        self._pre_setup()
        super(TestCase, self).__call__(result)
    finally:
        self._post_teardown()

The problem is that if create_app() throws an exception when called in _pre_setup(), then the self._orig_response_class field is never set/created. And because of the exception, the 'finally' block is executed, and on line 101, _post_teardown() is trying to reference that field, and that throws another exception (saying that the field doesn't exist). So:

1. The original exception is never reported.
2. The user sees a completely random exception instead.

I'm trying to install your package directly from github, using
pip install https://github.com/jarus/flask-testing.

However, when I do so, I get an error from pip:

Exception:
Traceback (most recent call last):
  File "/Library/Python/2.7/site-packages/pip/basecommand.py", line 122, in main
    status = self.run(options, args)
  File "/Library/Python/2.7/site-packages/pip/commands/install.py", line 278, in run
    requirement_set.prepare_files(finder, force_root_egg_info=self.bundle, bundle=self.bundle)
  File "/Library/Python/2.7/site-packages/pip/req.py", line 1197, in prepare_files
    do_download,
  File "/Library/Python/2.7/site-packages/pip/req.py", line 1375, in unpack_url
    self.session,
  File "/Library/Python/2.7/site-packages/pip/download.py", line 582, in unpack_http_url
    unpack_file(temp_location, location, content_type, link)
  File "/Library/Python/2.7/site-packages/pip/util.py", line 627, in unpack_file
    and is_svn_page(file_contents(filename))):
  File "/Library/Python/2.7/site-packages/pip/util.py", line 210, in file_contents
    return fp.read().decode('utf-8')
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/encodings/utf_8.py", line 16, in decode
    return codecs.utf_8_decode(input, errors, True)
UnicodeDecodeError: 'utf8' codec can't decode byte 0x8b in position 1: invalid start byte

Storing debug log for failure in /Users/Peter/Library/Logs/pip.log

Does anyone know how I can fix this? Thanks!

when i post username and password to def login();
it will raise a error
AssertionError: Popped wrong request context. (<RequestContext 'http://localhost/login' [POST] of adminapp> instead of <RequestContext 'http://localhost/' [GET] of adminapp>)

then i comment this line #login_user(Role.query.get(username=username))

test passing....

If you set render_templates to False, then flask's templating render gets monkeypatched with an empty render that doesn't do anything. Unfortunately, it doesn't seem to restore the original render, which causes all tests afterwards that require rendering to fail due to an empty response.

I have spotted the following code inside of TestCase._post_teardown:

        if hasattr(self, '_true_render'):
            templating._render = self._true_render

Presumably, this was intended to restore the render? If so, _true_render never gets created.. instead it's called _original_template_render.

-robodude666

I custome my TestCase's teardown method, and it case a error.
When I run this case, it could raise a error about Flask-testing's self._ctx.pop() error (link) .
That make me can't find real problem except I note r85 .

When the teardown method raise a error, flask-testing execute the _post_teardown will case this error :

AssertionError: Popped wrong request context.  (<RequestContext 'http://localhost/connect/signup' [POST] of run> instead of <RequestContext 'http://localhost/' [GET] of run>)

Twill itself has not been updated in 6 years and is not compatible with python 3. This means that flask-testing will never be able to support python 3 as long as it depends on twill.

The requests library has a callable json attribute.

To make TestResponse.json "optionally callable", we use the following work-around:

class JsonResponseMixin(object):
    """
    Mixin with testing helper methods
    """

    @cached_property
    def json(self):
        """Return response data as a Python object.

        The data is deserialized using the JSON format.
        """

        # Note that during testing, this response class may already
        # provide a JSON property. We re-implement it here for
        # consistency.
        value = loads(self.data)

        # Adapt value to make it callable. Alas, the 'requests'
        # library changed its response class such that the JSON
        # property is callable.
        cls = type(value)
        return type(cls.__name__, (cls, ), {
            '__call__': lambda v: cls(v),
            '__module__': cls.__module__,
            '__reduce__': lambda v: (cls, (cls(v), )),
        })(value)

That is, you can call it or use it directly – because it's a dict whose callable returns the dict.

Hello,

I am relatively new to Flask development, and I began writing tests cases simply using unittest, but I recently found this package and became interested. Meanwhile, there are no full examples of testing a page with a form and verifying the result on the server's side. The strange intermediary that twill serves perhaps is confusing, as it is a wrapper around mechanize, and twill development has halted, whilst mechanize development is still ongoing...

I would very much appreciate a more full fledged example, such as with flaskr. Or perhaps simply validating that test_manually actually produced an expected outcome server side.

Running the tests using: nosetests -e twill, I get:

.......................................F..F...
======================================================================
FAIL: test_assert_not_process_the_template (tests.test_utils.TestNotRenderTemplates)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/tmp/flask-testing/tests/test_utils.py", line 174, in test_assert_not_process_the_template
    assert "" == response.data
AssertionError

======================================================================
FAIL: test_assert_the_real_render_template_is_restored (tests.test_utils.TestRestoreTheRealRender)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/tmp/flask-testing/tests/test_utils.py", line 205, in test_assert_the_real_render_template_is_restored
    assert test_result.wasSuccessful()
AssertionError

----------------------------------------------------------------------
Ran 46 tests in 4.396s

FAILED (failures=2)

I'm using nose, and apparently, some of it's functionality, like the ability to yield tests, are not supported when inheriting from unittest.TestCase.

How do you feel about exposing the actual functionality of flask-testing's TestCase class as a bare TestCaseMixin class (inheriting from object) which could be used with nose.

• Bitbucket: https://bitbucket.org/danjac/flask-testing/issue/4
• Originally Reported By: Michael Elsdörfer
• Originally Created At: 2010-11-26 17:44:57

I had to replace the terminate method with the following to get coverage data from the child server process to be written to the file. *nix specific, but it works

    def _terminate_live_server(self):
        os.kill(self._process.pid, signal.SIGINT)
        self._process.join()

Without this, none of the code run by the server during selenium tests would be counted in coverage reports.

On py2.7.3 and following the code from https://pythonhosted.org/Flask-Testing/#testing-with-sqlalchemy I have created a test file as follows:

#test.py
import unittest
from flask.ext.testing import TestCase
from flask_issue import create_app
from model import User, db


class SomeTest(TestCase):

    SQLALCHEMY_DATABASE_URI = "sqlite:////home/mez/code/flask_issue/test.db"
    TESTING = True

    def create_app(self):

        # pass in test configuration
        return create_app(self)

    def setUp(self):

        db.create_all()

    def tearDown(self):

        db.session.remove()
        db.drop_all()

    def test_something(self):

        user = User()
        db.session.add(user)
        db.session.commit()

        # this works
        assert user in db.session

        response = self.client.get("/")
        print response

        # this raises an AssertionError
        assert user in db.session


if __name__ == '__main__':
    unittest.main()

I also have the following application and model files:

# flask_issue.py
from flask import Flask
from flask.ext.sqlalchemy import SQLAlchemy


def create_app(config_obj=None):
    new_app = Flask(__name__)
    if config_obj is not None:
        new_app.config.from_object(config_obj)
    from model import db
    db.init_app(new_app)
    return new_app

app = create_app()

@app.route('/')
def hello_world():
    return 'Hello World!'


if __name__ == '__main__':
    app.run()

and

# model.py
from flask.ext.sqlalchemy import SQLAlchemy

db = SQLAlchemy()


class User(db.Model):
    id = db.Column(db.Integer, primary_key=True)
    username = db.Column(db.String(80), unique=True)
    email = db.Column(db.String(120), unique=True)

In the given test.py setup code, the overriden create_app(self) method calls the imported create_app(self) from flask_issue. This piece of code assumptions about implementation details. This code only works if the user

• implements create_app in their own app (which is not normally required by Flask/SQLAlchemy)
• updates the config using new_app.config.from_object(config_obj) instead of new_app.config.update(config_dict) or new_app.config.from_pyfile

These assumptions should be explicitly mentioned.

Then, running these tests produces the following output:

mez@red:~/code/flask_issue$ python test.py 
<TestResponse streamed [404 NOT FOUND]>
.
----------------------------------------------------------------------
Ran 1 test in 0.456s

OK

This output does not match the tutorial's comments. The comment about assert user in db.session raising an AssertionError seems to be wrong.

Additionally, the given self.client.get("/") is not functioning. The output is a 404 because the routes are not being registered. Adding print self.app.url_map after the print response confirms this: Map([<Rule '/static/<filename>' (HEAD, OPTIONS, GET) -> static>])

The work-around is to manually register the routes when returning the newly created app:

class SomeTest(TestCase):

    SQLALCHEMY_DATABASE_URI = "sqlite:////home/mez/code/flask_issue/test.db"
    TESTING = True

    def create_app(self):

        # pass in test configuration
        new_app = create_app(self)
        from flask_issue import hello_world
        new_app.add_url_rule('/', 'hello_world', hello_world)
        return new_app

Yields the output:

mez@red:~/code/flask_issue$ python test.py 
<TestResponse streamed [200 OK]>
Map([<Rule '/' (HEAD, OPTIONS, GET) -> hello_world>,
 <Rule '/static/<filename>' (HEAD, OPTIONS, GET) -> static>])
.
----------------------------------------------------------------------
Ran 1 test in 0.473s

OK

The resulting functioning code has some key differences to the given tutorial code. Additionally, adding the routes manually is not maintainable. Is this the intended behavior?

Not an issue, more of a question.

My Flask apps tend to use route decorators (not all of them in blueprints), which means they use a global app object. My create_app function imports my view module. But if that create_app function is called more than once, as it is in Flask-Testing, the views import only works once and my routes aren't picked up.

What's the recommended way of combining route decorators with Flask-Testing?

Thanks.

The test matrix in CI currently uses multiple versions of Flask but does not cover the cases where components (ie. blinker) are installed and then not installed. Because of this, we were unable to catch the issue reported in: #91 but it would have been immediately obvious with the proper facilities.

Investigate a solution such that the build matrix can have different configurations of flask installations

Hi

Great plugin. Do you have any schedule for cutting a new release? I'm missing #45 in particular.

I'm trying to use unittest2 to add some of the newer unittest functionality to code running on python 2.6. It looks like flask.ext.testing.TestCase inherits directly from unittest.TestCase though, so some newer methods like assetIsNotNone are not present.

Is there some way I can make this work out of the box?

The change to flask.g has made flask test cases get the same g variable in subsequent calls to the test_client. Here is a test case:

from flask import Flask, g
app = Flask(__name__)
from flask.ext.testing import TestCase
import flask.ext.testing

@app.route("/")
def hello():
    if hasattr(g, 'foo'):
        return "g.foo is set to:" + g.foo
    else:
        g.foo = "bar"
        return "Hello World!"

class GTest(TestCase):
    def create_app(self):
        return app
    def test_g(self):
        tc = self.client
        r = tc.get('/')
        print r.data
        r = tc.get('/')
        print r.data
        self.assertEqual(r.data, "Hello World!")

if __name__ == "__main__":
    flask.ext.testing.unittest.main()

Hi,

Since 0.6.0 I can't get my tests succeed anymore. They all fail with the following error:

======================================================================
ERROR: test_live_monitor (test_burpui.BurpuiRoutesTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/tmp/burp-ui/py/lib/python2.7/site-packages/nose/case.py", line 133, in run
    self.runTest(result)
  File "/tmp/burp-ui/py/lib/python2.7/site-packages/nose/case.py", line 151, in runTest
    test(result)
  File "/tmp/burp-ui/py/lib/python2.7/site-packages/flask_testing/utils.py", line 112, in __call__
    self._post_teardown()
  File "/tmp/burp-ui/py/lib/python2.7/site-packages/flask_testing/utils.py", line 170, in _post_teardown
    message_flashed.disconnect(self._add_flash_message)
  File "/tmp/burp-ui/py/lib/python2.7/site-packages/flask/signals.py", line 32, in _fail
    raise RuntimeError('signalling support is unavailable '
RuntimeError: signalling support is unavailable because the blinker library is not installed.

• nose-1.3.7
• Flask-Testing-0.6.0
• Flask-0.11.1

How do I test this line?
return render_template(u'http_error.html', error_detail=detail)

I am a beginner as far as flask testing is concerned.

There is an unanswered question similar to this on stackoverflow.
http://stackoverflow.com/questions/23987564/test-flask-render-template-context