|
| 1 | +======= |
| 2 | +History |
| 3 | +======= |
| 4 | + |
| 5 | +3.12.0 (2022-05-11) |
| 6 | +------------------- |
| 7 | + |
| 8 | +* Support Python 3.11. |
| 9 | + |
| 10 | +3.11.0 (2022-01-10) |
| 11 | +------------------- |
| 12 | + |
| 13 | +* Drop Python 3.6 support. |
| 14 | + |
| 15 | +3.10.3 (2021-11-30) |
| 16 | +------------------- |
| 17 | + |
| 18 | +* Work on FIPS Python 3.9+, by declaring use of ``hashlib.md5()`` as not used for security. |
| 19 | + |
| 20 | + Thanks to dantebben for the report in `Issue #414 <https://github.com/pytest-dev/pytest-randomly/issues/414>`__. |
| 21 | + |
| 22 | +3.10.2 (2021-11-10) |
| 23 | +------------------- |
| 24 | + |
| 25 | +* Fix crash when pytest’s cacheprovider is disabled. |
| 26 | + |
| 27 | + Thanks to Mandeep Sandhu for the report in `Issue #408 |
| 28 | + <https://github.com/pytest-dev/pytest-randomly/issues/408>`__. |
| 29 | + |
| 30 | +* Improve group name in ``pytest --help``. |
| 31 | + |
| 32 | +3.10.1 (2021-08-13) |
| 33 | +------------------- |
| 34 | + |
| 35 | +* Fix new shuffling to work when one or more test in a class or module have the |
| 36 | + same test id. |
| 37 | + |
| 38 | + Thanks to Nikita Sobolev for the report in `Issue #378 |
| 39 | + <https://github.com/pytest-dev/pytest-randomly/issues/378>`__. |
| 40 | + |
| 41 | +3.10.0 (2021-08-13) |
| 42 | +------------------- |
| 43 | + |
| 44 | +* Rework shuffling algorithm to use hashing. This means that running a subset |
| 45 | + of tests with the same seed will now produce the same ordering as running the |
| 46 | + full set of tests. This allows narrowing down ordering-related failures. |
| 47 | + |
| 48 | + Thanks to Tom Grainger for the suggestion in `Issue #210 |
| 49 | + <https://github.com/pytest-dev/pytest-randomly/issues/210>`__. |
| 50 | + |
| 51 | +* Shuffle before other test collection hooks. This allows |
| 52 | + pytest’s `--stepwise flag |
| 53 | + <https://docs.pytest.org/en/latest/cache.html#stepwise>`__ to work, among |
| 54 | + other things. |
| 55 | + |
| 56 | + Thanks to Tom Grainger for the suggestion to try with ``--stepwise``. Fixes |
| 57 | + `Issue #376 |
| 58 | + <https://github.com/pytest-dev/pytest-randomly/issues/376>`__. |
| 59 | + |
| 60 | +3.9.0 (2021-08-12) |
| 61 | +------------------ |
| 62 | + |
| 63 | +* Add type hints. |
| 64 | + |
| 65 | +3.8.0 (2021-05-10) |
| 66 | +------------------ |
| 67 | + |
| 68 | +* Support Python 3.10. |
| 69 | + |
| 70 | +3.7.0 (2021-04-11) |
| 71 | +------------------ |
| 72 | + |
| 73 | +* Drop dependency on ``backports.entry-points-selectable`` by depending on |
| 74 | + ``importlib-metadata`` version 3.6.0+. |
| 75 | + |
| 76 | +3.6.0 (2021-04-01) |
| 77 | +------------------ |
| 78 | + |
| 79 | +* Fix deprecation warning from importlib-metadata 3.9.0+. |
| 80 | + |
| 81 | + Thanks to Dominic Davis-Foster for report in `Issue #333 |
| 82 | + <https://github.com/pytest-dev/pytest-randomly/issues/333>`__. |
| 83 | + |
| 84 | +* Stop distributing tests to reduce package size. Tests are not intended to be |
| 85 | + run outside of the tox setup in the repository. Repackagers can use GitHub's |
| 86 | + tarballs per tag. |
| 87 | + |
| 88 | +* Drop Python 3.5 support. |
| 89 | + |
| 90 | +3.5.0 (2020-11-16) |
| 91 | +------------------ |
| 92 | + |
| 93 | +* Support Python 3.9. |
| 94 | +* Move license from BSD to MIT License. |
| 95 | + |
| 96 | +3.4.1 (2020-07-10) |
| 97 | +------------------ |
| 98 | + |
| 99 | +* Fix numpy error ``ValueError: Seed must be between 0 and 2**32 - 1`` when |
| 100 | + passed a seed outside of this range. |
| 101 | + |
| 102 | +3.4.0 (2020-05-27) |
| 103 | +------------------ |
| 104 | + |
| 105 | +* Provide a ``faker_seed`` fixture to set the seed for tests using faker's |
| 106 | + pytest fixtures (as per its |
| 107 | + `docs <https://faker.readthedocs.io/en/master/pytest-fixtures.html#seeding-configuration>`__). |
| 108 | + |
| 109 | + Thanks to Romain Létendart for the change in `PR #261 |
| 110 | + <https://github.com/pytest-dev/pytest-randomly/pull/261>`__. |
| 111 | + |
| 112 | +3.3.1 (2020-04-15) |
| 113 | +------------------ |
| 114 | + |
| 115 | +* Fix to work when pytest-xdist is not installed or active |
| 116 | + (``PluginValidationError: unknown hook 'pytest_configure_node'``). |
| 117 | + |
| 118 | +3.3.0 (2020-04-15) |
| 119 | +------------------ |
| 120 | + |
| 121 | +* Add `pytest-xdist <https://pypi.org/project/pytest-xdist/>`__ support. |
| 122 | + Previously it only worked reliably when setting ``--randomly-seed`` |
| 123 | + explicitly. When not provided, the default seed generated in workers could |
| 124 | + differ and collection would fail. Now when it is not provided, all xdist |
| 125 | + worker processes shared the same default seed generated in the main |
| 126 | + process. |
| 127 | + |
| 128 | +3.2.1 (2020-01-13) |
| 129 | +------------------ |
| 130 | + |
| 131 | +* Update ``MANIFEST.in`` so tests are included in the sdist tarball again. |
| 132 | + |
| 133 | +3.2.0 (2019-12-19) |
| 134 | +------------------ |
| 135 | + |
| 136 | +* Converted setuptools metadata to configuration file. This meant removing the |
| 137 | + ``__version__`` attribute from the package. If you want to inspect the |
| 138 | + installed version, use |
| 139 | + ``importlib.metadata.version("pytest-randomly")`` |
| 140 | + (`docs <https://docs.python.org/3.8/library/importlib.metadata.html#distribution-versions>`__ / |
| 141 | + `backport <https://pypi.org/project/importlib-metadata/>`__). |
| 142 | +* Convert reading entrypoints to use ``importlib.metadata``. Depend on |
| 143 | + ``importlib-metadata`` on Python < 3.8. |
| 144 | +* Update Python support to 3.5-3.8. |
| 145 | + |
| 146 | +3.1.0 (2019-08-25) |
| 147 | +------------------ |
| 148 | + |
| 149 | +* Add plugins via entry points ``pytest_randomly.random_seeder`` to allow |
| 150 | + outside packages to register additional random generators to seed. This has |
| 151 | + added a dependency on the ``entrypoints`` package. |
| 152 | + |
| 153 | +3.0.0 (2019-04-05) |
| 154 | +------------------ |
| 155 | + |
| 156 | +* Update Python support to 3.5-3.7, as 3.4 has reached its end of life. |
| 157 | +* Handle ``CollectError``\s and ``ImportError``\s during collection when |
| 158 | + accessing ``item.module``. |
| 159 | + |
| 160 | +2.1.1 (2019-03-26) |
| 161 | +------------------ |
| 162 | + |
| 163 | +* Fix including tests in sdist after re-arrangement in 2.1.0. |
| 164 | + |
| 165 | +2.1.0 (2019-03-01) |
| 166 | +------------------ |
| 167 | + |
| 168 | +* Add the option ``--randomly-seed=last`` to reuse the last used value for the |
| 169 | + seed. |
| 170 | + |
| 171 | +2.0.0 (2019-02-28) |
| 172 | +------------------ |
| 173 | + |
| 174 | +* Drop Python 2 support, only Python 3.4+ is supported now. |
| 175 | + |
| 176 | +1.2.3 (2017-12-06) |
| 177 | +------------------ |
| 178 | + |
| 179 | +* Fix ``DeprecationWarning`` with recent versions of ``factory_boy``. |
| 180 | + |
| 181 | +1.2.2 (2017-11-03) |
| 182 | +------------------ |
| 183 | + |
| 184 | +* Fix collection to not sometimes crash when encoutering pytest ``Item``\s |
| 185 | + without a module. |
| 186 | + |
| 187 | +1.2.1 (2017-06-17) |
| 188 | +------------------ |
| 189 | + |
| 190 | +* Fix collection to be deterministically shuffled again, regression in 1.2.0. |
| 191 | + |
| 192 | +1.2.0 (2017-06-16) |
| 193 | +------------------ |
| 194 | + |
| 195 | +* Dropped Python 2.6 compatibility, as upstream dependency NumPy did. |
| 196 | +* Reset and output the seed at the start of the test run when |
| 197 | + ``--randomly-dont-reset-seed`` is set, to allow the reorganization of tests |
| 198 | + to be reproducible. |
| 199 | + |
| 200 | +1.1.2 (2016-10-27) |
| 201 | +------------------ |
| 202 | + |
| 203 | +* Reset the random state for NumPy too. |
| 204 | + |
| 205 | +1.1.1 (2016-09-16) |
| 206 | +------------------ |
| 207 | + |
| 208 | +* Add Python 2.6 compatibility |
| 209 | + |
| 210 | +1.1.0 (2016-09-12) |
| 211 | +------------------ |
| 212 | + |
| 213 | +* Offset the random seed during test setup and teardown. This is to avoid the |
| 214 | + awkward situation where test setup generates a random object, then the test |
| 215 | + generates a second one, but due to the re-seeding, they end up being always |
| 216 | + the same object. Thanks @rouge8 for the report. |
| 217 | + |
| 218 | +1.0.0 (2016-04-15) |
| 219 | +------------------ |
| 220 | + |
| 221 | +* First release on PyPI. |
0 commit comments