Welcome to Mockify!¶
Welcome to Mockify library documentation!
Mockify is a mocking toolkit for Python inspired by GMock (Google Mock) C++
framework. I was using GMock a lot during my 5 years of work as a C++ developer
and really liked it for its expressive API. During that days I was still
writing some Python code (mostly in Python 2.x) and for testing it I was using
hand-written stubs when needed. When I used unittest.mock
for the first
time I noticed that it uses a very different approach than GMock I got used to,
so I decided to start writing my own toolkit.
Currently, Mockify is supplied with following features:
- Creating mocks of standalone functions (more will come soon…)
- Recording call expectations with fixed arguments and using matchers
- Checking if expectations are satisfied using one single
assert_satisfied
assertion method
- Configuring recorded expectations:
- setting expected call count
- recording single and repeated actions (a.k.a. side effects)
- chaining actions
I hope you will find this library useful :-)
User’s Guide¶
Installation¶
From PyPI using virtualenv and pip¶
Mockify can be installed by simply invoking this inside active virtual Python environment:
$ pip install mockify
This will install most recently deployed version of the library.
You can also add Mockify to your requirements.txt file if your project already has one. After that, you can install all dependencies at once using this command:
$ pip install -r requirements.txt
Directly from source using virtualenv and pip¶
You can also install Mockify directly from source code by simply invoking this command inside active virtual Python environment:
$ pip install git+https://gitlab.com/zef1r/mockify.git@[branch-or-tag]
This will allow you to install most recent version of the library that may not be released yet to PyPI. And also you will be able to install from any branch and tag.
Verifying installation¶
Once Mockify is installed, you can simply check if it works by invoking this code to print version of installed Mockify library:
import mockify
print(mockify.version)
And you’re now ready to use Mockify for mocking things in your tests. Please proceed to Tutorial section of this documentation to learn how to use it.
Tutorial¶
Creating mocks¶
Function mocks¶
To create function mock you need to import function mock utility:
>>> from mockify.mock.function import Function
Now you can create function mock using following boilerplate pattern:
>>> foo = Function('foo')
Most examples in this tutorial use function mocks.
Recording and verifying expectations¶
Mocks with no expectations¶
When mock is created, it has no expectations set, so it already is satisfied:
>>> foo = Function('foo')
>>> foo.assert_satisfied()
Mockify requires each mock to have all needed expectations recorded. But since
foo
has no expectations recorded yet, it cannot be called with any
arguments and doing so will result in mockify.exc.UninterestedCall
exception being raised when call is made. For example:
>>> foo(1, 2)
Traceback (most recent call last):
...
mockify.exc.UninterestedCall: foo(1, 2)
In order to allow foo
to be called with (1, 2)
as parameters, a
matching expectation have to be recorded.
Mocks with one expectation¶
Let’s go back to our mock foo
defined in previous example and record a
matching expectation:
>>> foo.expect_call(1, 2)
<mockify.Expectation: foo(1, 2)>
Now we’ve recorded that foo
is expected to be called once with (1, 2)
as positional arguments. Since the mock now has expectation, it is not
satisfied now, as the expectation was not yet satisfied (previous failed call
does not count):
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo(1, 2)
Expected: to be called once
Actual: never called
As you can see, Mockify is presenting explanatory assertion message. You will know that only one expectation has failed and will no exactly which expectation it is as exact file and line number where the expectation was created are presented. Besides, you will also know how many times the mock is expected to be called with params matching Pattern and how many times it was actually called.
Each expectation can be in one of three states:
- unsatisfied,
- satisfied
- and oversaturated.
Currently, expectation from example above is in unsatisfied state, as it can still be satisfied by adequate number of matching mock calls. Let’s then call a mock once to make it satisfied:
>>> foo(1, 2)
>>> foo.assert_satisfied()
Calling a mock more times than expected is possible and will not cause
mockify.exc.UninterestedCall
exception, as this is only used to point
out that there were no expectations found that match given call parameters. But
if expectation is already satisfied and is called again, it becomes
oversaturated and the mock will stay unsatisfied for entire its lifetime:
>>> foo(1, 2)
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo(1, 2)
Expected: to be called once
Actual: called twice
>>> foo(1, 2)
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo(1, 2)
Expected: to be called once
Actual: called 3 times
Mocks with many expectations¶
Usually each mock will have many expectations recorded, as the code being under test will usually use its dependencies more than once and with many different parameters. Let’s have a look at following simple function:
>>> def example(count, callback):
... for i in range(count):
... callback(i)
This function is simply calling callback
given number of times and passes
current loop index as an argument on each iteration. If we want to test such
function we basically need 3 tests:
- Check if
callback
is not called whencount
is 0- Check if
callback
is called once with 0 whencount
is 1- Check if
callback
is triggered with 0, 1, …, N-1 ifcount
is N
First test can be written as simple as this one:
>>> callback = Function('callback')
>>> example(0, callback)
>>> callback.assert_satisfied()
If callback
gets called, the test will fail with
mockify.exc.UninterestedCall
exception. There is also a nicer way to
expect something to not happen but we’ll talk about this a bit later.
Second test will look similar to what we’ve already used in previous examples:
>>> callback = Function('callback')
>>> callback.expect_call(0)
<mockify.Expectation: callback(0)>
>>> example(1, callback)
>>> callback.assert_satisfied()
And third test would look like this. For the sake of simplicity let’s test our
example
function for N=2:
>>> callback = Function('callback')
>>> callback.expect_call(0)
<mockify.Expectation: callback(0)>
>>> callback.expect_call(1)
<mockify.Expectation: callback(1)>
>>> example(2, callback)
>>> callback.assert_satisfied()
As you can see, we have recorded two expectations. Mockify by default does not care about order of expectations, so the same can also be achieved if those expectations are reversed:
>>> callback = Function('callback')
>>> callback.expect_call(1)
<mockify.Expectation: callback(1)>
>>> callback.expect_call(0)
<mockify.Expectation: callback(0)>
>>> example(2, callback)
>>> callback.assert_satisfied()
Note
There are plans of implementing ordered expectations in future releases of Mockify.
Let’s now leave our example
function for a while and have a look at how
unsatisfied assertion is rendered in case of multiple failed expectations.
Let’s create another mock with two expectations and call assert_satisfied
on it:
>>> foo = Function('foo')
>>> foo.expect_call(1)
<mockify.Expectation: foo(1)>
>>> foo.expect_call(2)
<mockify.Expectation: foo(2)>
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following 2 expectations are not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo(1)
Expected: to be called once
Actual: never called
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo(2)
Expected: to be called once
Actual: never called
If you now call a mock for the first time and check if it is satisfied, you’ll see that only one unsatisfied expectation has left:
>>> foo(1)
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo(2)
Expected: to be called once
Actual: never called
And if call one remaining expected call, the mock will become satisfied:
>>> foo(2)
>>> foo.assert_satisfied()
Using matchers¶
Sometimes you will need to write single expectation that is supposed to match
multiple argument values. For this purpose, you will need matchers.
Matchers are simple objects with overloaded object.__eq__()
method.
Thanks to matchers you will be able to write expectations that match entire
classes of values, not exact ones. You will find predefined matchers in
mockify.matchers
module.
Let’s now use mockify.matchers.Any
matcher to show how it would look
in practice:
>>> from mockify.matchers import _
>>> foo = Function('foo')
>>> foo.expect_call(_)
<mockify.Expectation: foo(_)>
>>> foo.expect_call(_)
<mockify.Expectation: foo(_)>
We’ve just recorded that we expect foo
to be called twice with exactly one
argument of any kind. So, for example, we can satisfy our mock with this:
>>> foo([])
>>> foo('spam')
>>> foo.assert_satisfied()
Matchers will also allow us to write complex patterns. For example, if mock is called with dict as an argument and the dict represents JSONRPC request (see: https://www.jsonrpc.org/specification), we could write expectation that we want our mock to be execute with request object, but no matter what is the method, params and ID:
>>> foo = Function('foo')
>>> foo.expect_call({'jsonrpc': '2.0', 'method': _, 'params': _, 'id': _})
<mockify.Expectation: foo({...})>
>>> foo({'jsonrpc': '2.0', 'method': 'spam', 'params': 123, 'id': 1})
>>> foo.assert_satisfied()
But if now the mock is called with different dict structure, the call will fail:
>>> foo({'jsonrpc': '2.0'})
Traceback (most recent call last):
...
mockify.exc.UninterestedCall: foo({'jsonrpc': '2.0'})
Configuring expectation objects¶
So far, we’ve done nothing with mockify.engine.Expectation
object
expect_call
method returns. But it has a lot of very handy features that we
are going to discuss right now.
Expecting a mock to be never called¶
It is very tricky to expect something to never happen as there are infinite number of possibilities. Besides, especially if it takes time to execute test, after how many seconds should we say that somethid did not happpen? But sometimes you may need to expect a mock to be never called.
Let’s go back to our example
function defined before. There was a test that
callback is never called. The test looked like this:
>>> callback = Function('callback')
>>> example(0, callback)
>>> callback.assert_satisfied()
Although it works fine, there is not visible what we are expecting. Same test can be done like this:
>>> from mockify.matchers import _
>>> callback = Function('callback')
>>> callback.expect_call(_).times(0)
<mockify.Expectation: callback(_)>
>>> example(0, callback)
>>> callback.assert_satisfied()
As you can see, we’ve used mockify.engine.Expectation.times()
method and
called it with 0, meaning that we expect callback
to be called 0 times. Now
the test looks more expressive, but as stated in the beginning, expecting
something to never happen is tricky. No matter if we call example
function,
other function or even nothing instead, the test will still pass:
>>> from mockify.matchers import _
>>> callback = Function('callback')
>>> callback.expect_call(_).times(0)
<mockify.Expectation: callback(_)>
>>> callback.assert_satisfied()
Just like normally expectation has expected call count set to one, modifying it
with times(0)
sets this counter to 0, so mock is already satisfied.
Situtation changes when mock gets called:
>>> callback(0)
>>> callback.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: callback(_)
Expected: to be never called
Actual: called once
Expecting a mock to be called given number of times¶
So far, if we needed to expect a mock to be called more than once we’ve recorded two or more expectations with same parameters. But there is a better way of doing this.
Let’s go back to our example
function and third test. We can rewrite it in
following way:
>>> callback = Function('callback')
>>> callback.expect_call(_).times(2)
<mockify.Expectation: callback(_)>
>>> example(2, callback)
>>> callback.assert_satisfied()
But actually we’ve verified only that mock is called twice each time with any
argument. So in fact, if example
calls a mock with fixed argument, then the
test above will still pass. Therefore, we need another matcher to ensure that
mock is called with valid arguments. For that purpose, we’ll use
mockify.matchers.SaveArg
:
>>> from mockify.matchers import SaveArg
>>> count = SaveArg()
>>> callback = Function('callback')
>>> callback.expect_call(count).times(2)
<mockify.Expectation: callback(SaveArg)>
>>> example(2, callback)
>>> callback.assert_satisfied()
>>> count.called_with == [0, 1]
True
Using mockify.matchers.SaveArg
you will also have to do some
additional assertions like in example above.
Method mockify.engine.Expectation.times()
allows to configure more then
just fixed expected number of calls. For more information go to the
mockify.times
module documentation.
Single actions¶
Besides setting how many times each mock is expected to be called and with what
arguments, you can also record actions to be executed on each mock call. For
example, we can tell a mock to return given value when it gets called. To do
this, we need to use mockify.engine.Expectation.will_once()
method:
>>> from mockify.actions import Return
>>> foo = Function('foo')
>>> foo.expect_call().will_once(Return(1))
<mockify.Expectation: foo()>
If you now check if mock is satisfied, you’ll notice that there is additional information of what action is going to be executed next:
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(1)
Expected: to be called once
Actual: never called
So if you now call a mock, it will return 1 and will be satisfied:
>>> foo()
1
>>> foo.assert_satisfied()
But if you now call a mock again it will end up with an exception:
>>> foo()
Traceback (most recent call last):
...
mockify.exc.OversaturatedCall: at <doctest tutorial.rst[...]>:1: foo(): no more actions recorded for call: foo()
This is a very special situation, as when actions are recorded it is assumed that the mock should always return something. Therefore, failing to do that is treated as exception currently.
Note
There are plans to implement default actions, so there will be no such exception in that case, but a default action will be executed instead. But mock will not be satisfied anyway.
For more actions please proceed to the mockify.actions
documentation.
Action chains¶
You can chain mockify.engine.Expectation.will_once()
method invocations
to end up with action chains being recorded, so each time when mock is called,
next action in a chain is executed. For example, you can record expectation
that mock is going to be called twice, returning 1 on first call and 2 on
second call:
>>> foo = Function('foo')
>>> foo.expect_call().will_once(Return(1)).will_once(Return(2))
<mockify.Expectation: foo()>
When you now check if mock is satisfied, you will be informed that it is
expected to be called twice and that next action is Return(1)
:
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(1)
Expected: to be called twice
Actual: never called
If you now call a mock, it will return 1:
>>> foo()
1
If you now check if it is satisfied, you will notice that one more call is
needed and that next action will be Return(2)
:
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(2)
Expected: to be called twice
Actual: called once
Finally, if you call a mock for the second time it will return 2 and mock will become satisfied:
>>> foo()
2
>>> foo.assert_satisfied()
You can of course record different actions type for each call. For list of
available built-in actions or instructions of how to make custom ones please
refer to the mockify.actions
module documentation.
Repeated actions¶
Repeated actions allow to set single action that will keep being executed each
time the mock is called. By default, if mock has repeated action set it can be
called any number of times, so mock with repeated action set is initially
satisfied. Repeated actions are recorded using
mockify.engine.Expectation.will_repeatedly()
method:
>>> foo = Function('foo')
>>> foo.expect_call().will_repeatedly(Return(1))
<mockify.Expectation: foo()>
>>> foo.assert_satisfied()
And you can call mock with such defined expectation any times you want. For example, lets call it 3 times. The mock will return 1 on each call and still will be satisfied:
>>> for _ in range(3):
... foo()
1
1
1
>>> foo.assert_satisfied()
You can also use mockify.engine.Expectation.times()
method to set
expected call count on a repeated action. For example, if you want to record
repeated action that can be executed at most twice, you would write following:
>>> from mockify.times import AtMost
>>> foo = Function('foo')
>>> foo.expect_call().will_repeatedly(Return(1)).times(AtMost(2))
<mockify.Expectation: foo()>
Such expectation is already satisfied (as at most twice is 0, 1 or 2 calls):
>>> foo.assert_satisfied()
But right now if you call a mock 3 times, the mock will no longer be satisfied:
>>> for _ in range(3):
... foo()
1
1
1
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(1)
Expected: to be called at most twice
Actual: called 3 times
Recording complex expectations¶
Currently we’ve used all of the features independently, but actually it is possible to record expectations that are combination of those. For example, you can record few single actions, and one repeated:
>>> foo = Function('foo')
>>> foo.expect_call().will_once(Return(1)).will_once(Return(2)).will_repeatedly(Return(3))
<mockify.Expectation: foo()>
Such mock will be expected to be called at least twice, as there are two single actions in the chain recorded:
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(1)
Expected: to be called at least twice
Actual: never called
If now the mock is called for the fist time it will return 1, for the second time - 2, and after that it will keep returning 3. And of course it will be satisfied, as all single actions were consumed:
>>> foo()
1
>>> foo()
2
>>> for _ in range(3):
... foo()
3
3
3
>>> foo.assert_satisfied()
You can also set expected call count for repeated action:
>>> foo = Function('foo')
>>> foo.expect_call().will_once(Return(1)).will_repeatedly(Return(2)).times(2)
<mockify.Expectation: foo()>
Now the mock will have to be called exactly 3 times:
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(1)
Expected: to be called 3 times
Actual: never called
>>> foo()
1
>>> foo()
2
>>> foo()
2
>>> foo.assert_satisfied()
Even such combinations are possible:
>>> foo = Function('foo')
>>> foo.expect_call().will_once(Return(1)).will_repeatedly(Return(2)).times(2).will_once(Return(3))
<mockify.Expectation: foo()>
And this time the mock is expected to be called 4 times:
>>> foo.assert_satisfied()
Traceback (most recent call last):
...
mockify.exc.Unsatisfied: following expectation is not satisfied:
at <doctest tutorial.rst[...]>:1
-----------------------------...
Pattern: foo()
Action: Return(1)
Expected: to be called 4 times
Actual: never called
>>> foo()
1
>>> for _ in range(2):
... foo()
2
2
>>> foo()
3
>>> foo.assert_satisfied()
API Reference¶
Mocking frontends¶
Mocking frontends are used to provide easy to use interface on top of classes
from mockify.engine
module that act like a backend.
So instead of doing this:
>>> from mockify.engine import Call, Registry
>>> reg = Registry()
>>> reg.expect_call(Call('foo', (1, 2)), 'foo.py', 123)
<mockify.Expectation: foo(1, 2)>
>>> reg(Call('foo', (1, 2)))
>>> reg.assert_satisfied()
You use a frontend to do the same much easier:
>>> from mockify.mock.function import Function
>>> foo = Function('foo')
>>> foo.expect_call(1, 2)
<mockify.Expectation: foo(1, 2)>
>>> foo(1, 2)
>>> foo.assert_satisfied()
mockify.mock.function
- Frontends for mocking functions¶
-
class
mockify.mock.function.
Function
(name, registry=None)¶ Bases:
object
Class for mocking Python functions.
Example usage:
>>> foo = Function('foo') >>> foo.expect_call(1, 2).times(2) <mockify.Expectation: foo(1, 2)> >>> for _ in range(2): ... foo(1, 2) >>> foo.assert_satisfied()
Parameters: - name – Mock function name
- registry –
This is optional.
Use this to pass custom instance of
mockify.engine.Registry
class if you need to share it between multiple frontends. Sharing is useful for example to check if all mocks are satisfied using oneassert_satisfied
call:>>> from mockify.engine import Registry >>> reg = Registry() >>> foo = Function('foo', registry=reg) >>> bar = Function('bar', registry=reg) >>> foo.expect_call() <mockify.Expectation: foo()> >>> bar.expect_call() <mockify.Expectation: bar()> >>> foo() >>> bar() >>> reg.assert_satisfied()
-
assert_satisfied
()¶ Assert that this function mock is satisfied.
This method just calls
mockify.engine.Registry.assert_satisfied()
withname
given via constructor as an argument.
-
expect_call
(*args, **kwargs)¶ Record call expectation.
This method creates
mockify.engine.Call
instance giving itargs
andkwargs
, fetches file and line number from current call stack and triggersmockify.engine.Registry.expect_call()
and returns expectation object it produces.
mockify.engine
- Library core¶
This module contains set of classes that provides backend mechanism for storing and tracking call expectations.
-
class
mockify.engine.
Call
(name, args=None, kwargs=None)¶ Bases:
object
Binds mock name with arguments it was called with or it is expected to be called with.
Call objects are created in mock frontends (like
mockify.mock.function.Function
mock class) by methodsexpected_call
and__call__
by simply passing their argument toCall
constructor.Instances of this class are comparable. Two
Call
objects are equal if and only if all attributes (name
,args
andkwargs
) are the same. For example:>>> Call('foo') == Call('foo') True >>> Call('foo') != Call('bar') True >>> Call('foo', (1, 2), {'c': 3}) == Call('foo', (1, 2), {'c': 3}) True
Call objects can also be created with use of matchers, for example
mockify.matchers.Any
, that will match any value:>>> from mockify.matchers import _ >>> Call('foo', (_, _)) == Call('foo', (1, 2)) True >>> Call('foo', (_, _)) == Call('foo', (3, 4)) True
Parameters: - name – Function or method name.
- args – Positional arguments
- kwargs – Named arguments
-
args
¶ Mock positional args.
-
kwargs
¶ Mock named args.
-
name
¶ Mock name.
-
class
mockify.engine.
Expectation
(expected_call, filename, lineno)¶ Bases:
object
Class representing single expectation.
Instances of this class are normally created by registry objects using
Registry.expect_call()
method. Each instance of this class is correlated with exactly onemockify.engine.Call
object representing expected mock call pattern.After
Expectation
object is created by call to someexpect_call
method, it can be mutated using following methods:Parameters: - call – Instance of
mockify.engine.Call
representing expected mock call pattern - filename – File name were this expectation was created
- lineno – Line number where this expectation was created
-
__call__
(call)¶ Call this expectation object.
If given
call
object does not matchexpected_call
then this method will raiseTypeError
exception.Otherwise, total call count is increased by one and:
- if actions are recorded, then next action is executed and its
result returned or
mockify.exc.OversaturatedCall
exception is raised if there are no more actions - if there are no actions recorded, just
None
is returned
- if actions are recorded, then next action is executed and its
result returned or
-
expected_call
¶ Instance of
mockify.engine.Call
representing expected mock call pattern.This basically is exactly the same
Call
object as was passed toExpectation
constructor.
-
format_action
()¶ Return textual representation of next action to be executed.
This method uses action’s
__str__
method to render action name.Returns
None
if there were no actions recorded or all were consumed.This is used by
mockify.exc.Unsatisfied
exception when rendering error message.
-
format_actual
()¶ Return textual representation of how many times this expectation was called so far.
This is used by
mockify.exc.Unsatisfied
exception when rendering error message.
-
format_expected
()¶ Return textual representation of how many times this expectation is expected to be called.
This is used by
mockify.exc.Unsatisfied
exception when rendering error message.
-
format_location
()¶ Return textual representation of place (filename and lineno) where this expectation was created.
Basically, it just returns
[filename]:[lineno]
string, wherefilename
andlineno
are given viaExpectation
constructor.
-
is_satisfied
()¶ Check if this expectation is satisfied.
-
match
(call)¶ Check if
expected_call
matchescall
.
-
times
(expected_count)¶ Record how many times this expectation is expected to be called.
Parameters: expected_count – Expected call count.
This can be either integer number (exact call count) or instance of one of classes from
mockify.times
module.
-
will_once
(action)¶ Attach action to be executed when this expectation gets consumed.
This method can be used several times, making action chains. Once expectation is consumed, next action is executed and removed from the list. If there are no more actions, another call will fail with
mockify.exc.OversaturatedCall
exception.After this method is used, you can also use
will_repeatedly()
to record repeated action that will get executed after all single actions are consumed.Parameters: action – Action to be executed.
See
mockify.actions
for details.
-
will_repeatedly
(action)¶ Attach repeated action to be executed when this expectation is called.
This method is used to record one action that gets executed each time this expectation object is called. By default, when repeated action is recorded, expectation can be called any number of times (including zero).
After setting repeated action, you can also set expected call count using
times()
.Parameters: action – Action to be executed.
See
mockify.actions
for details.
- call – Instance of
-
class
mockify.engine.
Registry
(expectation_class=None)¶ Bases:
object
Acts like a database for
Expectation
objects.This class is used as a backend for higher level mocking utilities (a.k.a. frontends), like
mockify.mock.function.Function
mocking class. It provides methods to record, lookup and verifying of expectations.There can be many instances of registry classes, or one that can be shared between various mock frontends. For example, you can create one registry in setup code, then create various mocks inside your tests, to finally trigger
assert_satisfied()
of that single registry in test’s teardown code. Or you can just use frontends with their defaults. It is completely up to you.Parameters: expectation_class – This is optional.
Used to give custom subclass of
Expectation
to be used inside this registry.-
__call__
(call)¶ Call a mock.
When this method is called, registry performs a lookup of matching unsatisfied expectations and calls first expectation found. If there are no matching expectation, then
mockify.exc.UninterestedCall
exception is raised. If there are matching expectations but all are satisfied, then last is called (making it oversaturated).Parameters: call – Instance of mockify.engine.Call
class representing mock being called
-
assert_satisfied
(name=None)¶ Assert that all expectations are satisfied.
If there is at least one unsatisfied expectation, then this method will raise
mockify.exc.Unsatisfied
exception containing list of failed expectations.This method can be called as many times as you want.
Parameters: name – This is optional.
If this is given, then method performs a lookup of expectations having expected calls matching
name
and verifies only found expectations. When called withname
, this method may pass for somename
’s, but can still fail if called without parameter.
-
expect_call
(call, filename, lineno)¶ Register expectation.
Returns instance of
expectation_class
(usuallyExpectation
) representing newly created expectation.Parameters: - call – Instance of
mockify.engine.Call
class representing exact mock call or a pattern (if created with matchers) that is expected to be executed - filename – Path to file were expectation is created
- lineno – Line number (inside
filename
) where expectation is created
- call – Instance of
-
mockify.actions
- Actions for use with will_once or will_repeatedly¶
Module containing predefined actions that can be used as argument for
Expectation.will_once()
or Expectation.will_repeatedly()
.
Basically, any class containing following methods is considered an action:
__str__(self)
Returning string representation of an action.
This is used for error reporting.
__call__(self, *args, **kwargs)
Method that is called when mock is called.
Entire action logic goes in here.
-
class
mockify.actions.
Invoke
(func)¶ Bases:
object
Makes mock invoking given function when called.
When mock is called, all arguments (if there are any) are passed to the
func
and its return value is returned as mock’s return value.Parameters: func – Function to be executed
mockify.exc
- Exceptions¶
-
exception
mockify.exc.
OversaturatedCall
(expectation, call)¶ Bases:
TypeError
Raised when mock is called more times than expected.
This exception will be thrown only if mock has actions defined as it does not know what to do next if all expected actions were already executed.
Parameters: - expectation – Instance of
mockify.engine.Expectation
class representing expectation that was oversaturated - call – Instance of
mockify.engine.Call
class representing mock call that oversaturatedexpectation
-
call
¶ Instance of
mockify.engine.Call
passed toOversaturatedCall
constructor.
-
expectation
¶ Instance of
mockify.engine.Expectation
passed toOversaturatedCall
constructor.
- expectation – Instance of
-
exception
mockify.exc.
UninterestedCall
(call)¶ Bases:
TypeError
Raised when uninterested mock is called.
Mockify requires each mock call to have matching expectation recorded. If none is found during call, then this exception is raised, terminating the test.
Parameters: call – Instance of mockify.engine.Call
class representing uinterested mock call-
call
¶ Instance of
mockify.engine.Call
passed toUninterestedCall
constructor.
-
-
exception
mockify.exc.
Unsatisfied
(expectations)¶ Bases:
AssertionError
Raised by
mockify.engine.Registry.assert_satisfied()
method when there is at least one unsatisfied expectation.This exception displays explanatory information to the user:
- file location where unsatisfied expectation was recorded
- expected call pattern
- expected call count
- actual call count
- next action to be executed (if any)
Parameters: expectations – List of mockify.engine.Expectation
instances representing all unsatisfied expectations-
expectations
¶ Instance of
mockify.engine.Expectation
passed toUnsatisfied
constructor.
mockify.helpers
- Various helper utilities¶
-
mockify.helpers.
assert_satisfied
(*subjects)¶ Context manager for verifying multiple subjects at once.
Each passed subject must have
assert_satisfied
method defined, so it can be used withmockify.mock.function.Function
ormockify.engine.Registry
instances for example.Basically, the role of this helper is to ensure that all subjects become satisfied after leaving wrapped context. For example:
>>> from mockify.mock.function import Function >>> foo = Function('foo') >>> bar = Function('bar') >>> foo.expect_call() <mockify.Expectation: foo()> >>> bar.expect_call().times(0) <mockify.Expectation: bar()> >>> with assert_satisfied(foo, bar): ... foo()
And that’s it - you don’t have to explicitly check if
foo
andbar
are satisfied, because the helper will do it for you. And also it emphasizes part of code that actually uses given mocks.
mockify.matchers
- Matchers for use with expect_call¶
Module containing predefined matchers.
A matcher is every class that inherits from Matcher
and implements
following methods:
__repr__(self)
- Return matcher’s text representation.
__eq__(self, other)
Check if
self
is equal toother
.Here we use standard Python
__eq__
operator as it will be automatically executed by Python no matter where the matcher is used. But equality definition is completely up to the matcher implementation.
-
class
mockify.matchers.
Any
¶ Bases:
mockify.matchers.Matcher
Matcher that matches any value.
It is available also as
_
(underscore) single instance that can be imported from this module.For example, you can record expectation that mock must be called with one positional argument of any value but exactly 3 times:
>>> from mockify.matchers import _ >>> from mockify.mock.function import Function >>> foo = Function('foo') >>> foo.expect_call(_).times(3) <mockify.Expectation: foo(_)> >>> for i in range(3): ... foo(i) >>> foo.assert_satisfied()
-
class
mockify.matchers.
SaveArg
¶ Bases:
mockify.matchers.Matcher
Matcher that matches any value and keeps ordered track of unique values.
This can be used as a replacement for
Any
in case that you need to ensure that mock was called in specified order.For example:
>>> from mockify.mock.function import Function >>> arg = SaveArg() >>> foo = Function('foo') >>> foo.expect_call(arg).times(3) <mockify.Expectation: foo(SaveArg)> >>> for i in range(3): ... foo(i) >>> foo.assert_satisfied() >>> arg.called_with == [0, 1, 2] True
-
called_with
¶ List of ordered unique values that this matcher was called with.
-
mockify.times
- Classes for use with times method¶
Module containing set of classes to be used with
mockify.engine.Expectation.times()
method.
You can also create your own classes to be used with that method. The only thing required from such class is to implement following interface:
is_satisfied(self, actual)
Return
True
ifactual
call count is satisfied byself
, orFalse
otherwise.Here,
actual
is absolute call count expectation received so far. It is completely implementation-specific of which values ofactual
are said to be satisfied and which are not. For example,Exactly
will compareactual
with fixed value (given via constructor) and returnTrue
only if those two are equal.adjust_by(self, minimal)
Adjust
self
by currentminimal
expected call count and return new instance oftype(self)
.In some complex expectation there could be a situation in which expectation must be computed again. This is not visible for library user, but must be done behind the scenes to properly process expectations. Such situation can be presented in this example:
>>> from mockify.actions import Return >>> from mockify.mock.function import Function >>> foo = Function('foo') >>> foo.expect_call(1, 2).will_once(Return(1)).will_repeatedly(Return(2)).times(2) <mockify.Expectation: foo(1, 2)> >>> foo(1, 2) 1 >>> foo(1, 2) 2 >>> foo(1, 2) 2 >>> foo.assert_satisfied()In example above we’ve used
times(2)
to tell that last repeated action is expected to be called twice, but real expected call count is 3 times, aswill_once
is used. Behind the scenes, this is recalculated using this metho.format_expected(self)
Return textual representation of expected call count.
This is used by
mockify.exc.Unsatisfied
exception when error message is being rendered.minimal(sefl)
(property)Property containing minimal call count that is considered valid for given instance.
For example, for
AtLeast
orExactly
it would be just its constructor argument, for :class`AtMost` it will be 0, forBetween
it will be itsminimal
argument.
-
class
mockify.times.
AtLeast
(minimal)¶ Bases:
object
Used to set minimal expected call count.
If this is used, then expectation is said to be satisfied if actual call count is not less that
minimal
.Parameters: minimal – Integer value representing minimal expected call count
-
class
mockify.times.
AtMost
(maximal)¶ Bases:
object
Used to set maximal expected call count.
If this is used, then expectation is said to be satisfied if actual call count is not greater than
maximal
.Parameters: maximal – Integer value representing maximal expected call count
-
class
mockify.times.
Between
(minimal, maximal)¶ Bases:
object
Used to set a range of valid call counts.
If this is used, then expectation is said to be satisfied if actual call count is not less than
minimal
and not greater thanmaximal
.Parameters: - minimal – Integer value representing minimal expected call count
- maximal – Integer value representing maximal expected call count
-
class
mockify.times.
Exactly
(expected)¶ Bases:
object
Used to expect fixed call count to be made.
You do not have to use this class explicitly as its instances are automatically created when you call
times
method with integer value as argument.Parameters: expected – Integer value representing expected call count
License¶
Mockify is released under the terms of the MIT license.
Copyright (C) 2018 Maciej Wiatrzyk
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.