feat(): add assert_match_with_ignore

[LuisFros]

Description

This is a complementary method to allow asserting snapshots while ignoring certain fields.

• Based on Feature - assert_match() with ignored keys #102 and solving Please add ignored fields #21 Add ignore fields parameter #68 Ignore fields parameter in assert methods #32
• Our team at https://github.com/reversso has been using this as a monkey patch of the library and we wanted to share (and hopefully contribute to this awesome library )
• I would like to give credit to @HeyHugo since we based our solution on his PR Feature - assert_match() with ignored keys #102

Why?

• The most classic use case for this method is autogenerated IDs that will change in between test batteries. (For example if databases are reused or random test order implies IDs are not unique)
• Also when using random generated data without seeds

Solution

It covers recursive:

• Lists
• Dictionaries
• Tuples

Additionally, it allows to exclude multiple values in the same "level".

 def assert_match_with_ignore(self, data, ignore_keys):
        """Extension of assert_match to ignore data.
        Args:
            data (dict,list,tuple): Data to be asserted
            ignored_keys (list): List of strings containing path to be ignored,
            special character "[.]" can be used to ignore multiple elements in list.
            (See Example 2)
        Returns:
            None: Asserts if the values are the same
        Raises:
            AssertionError: If the snapshot is different than the incoming data
        Examples:
            Test examples at: apps/tests/utils/test_asserts.py
            Example 1:
            >>> data={"dict1": {"dict2": {"dict3": {"id": "importantId", "other": "value"}}}}
            >>> ignore_keys=["dict1.dict2.dict3.id"]
            >>> assert_match_with_ignore(data,ignore_keys)
                # Will create the following snapshot
                snapshots['example_snapshot'] = {
                    'dict1': {
                        'dict2': {
                            'dict3': {
                                'id': None,
                                'other': 'value'
                            }
                        }
                    }
                }
            ---
            Example 2:
            >>> data=[
                    {
                    "name": "objectList",
                    "children": [
                        {"id": "random_string", "name": "child_1",},
                        {"id": "random_string2", "name": "child_2",},
                    ],
                    }
                ]
            >>> ignore_keys=["[0].children[.].id"]
            >>> assert_match_with_ignore(data,ignore_keys)
                # Will create the following snapshot
                snapshots['example2_snapshot'] = [
                    {
                    "name": "objectList",
                    "children": [
                        {"id": None, "name": "child_1",},
                        {"id": None, "name": "child_2",},
                    ],
                    }
                ]
        """

        self.assert_match(clear_ignore_keys(data, ignore_keys))

[syrusakbary]

This looks great, however I think we may not need a new assert_match_with_ignore.

We can probably just add the clear_ignore_keys method so people can use assert_match normally.

[LuisFros]

@syrusakbary thank you for your answer, sounds like a good proposal.
I am using pytest and one reason to add a new method was that it's more practical to use the fixture and then snapshot.assert_match_with_ignore instead of having to import a method from the library like from snapshot.utils import clear_ignore_keys every time we need to use it. What are your thoughts on this?