Remove dependency on mock and testfixtures

[mwouts]

That was suggested by @jsd-spif: mock is available as unittest.mock in Python 3, so we don't need a dependency on mock.

Similarly, testfixtures.compare could probably be replaced with assertDictEqual from unittest.

[mwouts]

Replacing mock with unitest.mock in Python 3 seems to be feasible.

Removing the dependency on testfixtures is a bit more complex. I am not sure we can use assertDictEqual directly because it is supposed to be used within unit tests. And using assert nb1 == nb2 is clearly not enough: we don't see the difference between notebooks well enough. See for instance this test:

import pytest
from nbformat.v4.nbbase import new_notebook, new_code_cell, new_markdown_cell


def notebook_metadata():
    return {
        "kernelspec": {
            "display_name": "Python 3",
            "language": "python",
            "name": "python3"
        },
        "language_info": {
            "codemirror_mode": {
                "name": "ipython",
                "version": 3
            },
            "file_extension": ".py",
            "mimetype": "text/x-python",
            "name": "python",
            "nbconvert_exporter": "python",
            "pygments_lexer": "ipython3",
            "version": "3.7.3"
        },
        "toc": {
            "base_numbering": 1,
            "nav_menu": {},
            "number_sections": True,
            "sideBar": True,
            "skip_h1_title": False,
            "title_cell": "Table of Contents",
            "title_sidebar": "Contents",
            "toc_cell": False,
            "toc_position": {},
            "toc_section_display": True,
            "toc_window_display": False
        }
    }


@pytest.fixture()
def notebook_1():
    return new_notebook(
        metadata=notebook_metadata(),
        cells=[new_markdown_cell('First markdown cell'),
               new_code_cell('1 + 1'),
               new_markdown_cell('Second markdown cell')])


@pytest.fixture()
def notebook_2():
    metadata = notebook_metadata()
    metadata['language_info']['version'] = '3.6.8'
    return new_notebook(
        metadata=metadata,
        cells=[new_markdown_cell('First markdown cell'),
               new_code_cell('1 + 1'),
               new_markdown_cell('Modified markdown cell')])


def test_notebooks_equal(notebook_1, notebook_2):
    assert notebook_1 == notebook_2

the result of which is

___________________________________________________________________________________________ test_notebooks_equal ____________________________________________________________________________________________

notebook_1 = {'nbformat_minor': 2, 'cells': [{'source': 'First markdown cell', 'cell_type':...nts_lexer': 'ipython3', 'codemirror_mode': {'version': 3, 'name': 'ipython'}}}}
notebook_2 = {'nbformat_minor': 2, 'cells': [{'source': 'First markdown cell', 'cell_type':...nts_lexer': 'ipython3', 'codemirror_mode': {'version': 3, 'name': 'ipython'}}}}

    def test_notebooks_equal(notebook_1, notebook_2):
>       assert notebook_1 == notebook_2
E       AssertionError: assert {'nbformat_mi... 'ipython'}}}} == {'nbformat_min... 'ipython'}}}}
E         Omitting 2 identical items, use -vv to show
E         Differing items:
E         {'cells': [{'source': 'First markdown cell', 'cell_type': 'markdown', 'metadata': {}}, {'source': '1 + 1', 'cell_type'... 'execution_count': None, 'outputs': []}, {'source': 'Second markdown cell', 'cell_type': 'markdown', 'metadata': {}}]} != {'cells': [{'source': 'First markdown cell', 'cell_type': 'markdown', 'metadata': {}}, {'source': '1 + 1', 'cell_type'...execution_count': None, 'outputs': []}, {'source': 'Modified markdown cell', 'cell_type': 'markdown', 'metadata': {}}]}
E         {'metadata': {'kernelspec': {'di...
E
E         ...Full output truncated (2 lines hidden), use '-vv' to show

test_compare_notebooks.py:61: AssertionError

which does not show clearly the difference between the two notebooks...

Hello @mwouts

You are right. It is clearly feasible to rely on unittest.mock to avoid depending on mock when running jupytext under python3, but we should not remove the dependency to testfixtures due to the usage you have of the compare method.

Anyway, thanks for the feedback, I just got time last week end to write changes for the mock package. I'll open the PR soon.

[mwouts]

Finally I was able to implement a replacement for testfixtures.compare using difflib, see this commit.

@jsd-spif , are you ok if I merge the corresponding PR now, or do you prefer to contribute your PR first?