- pre-commit from 4.1.0 to 4.3.0,
- codespell from 2.2.1 to 2.2.2,
- black from 22.3.0 to 22.10.0 and
- flake8 from 4.0.1 to 5.0.4. Also freeze flake8 dependencies.

Also change flake8's repo config to github (the gitlab mirror
being outdated).

to more reasonable versions...

instead of creating it in the VaultBackend constructor: when configured
(with host and port), SMTP.connect() is immediately called, which makes
it mandatory to have the smtp server up and running to be able to create
the VaultBackend object (which makes it hard to run properly in an elastic
environment like docker or k8s).

This also removes the fallback to hardcoded 'localhost:25' smtp server;
if the smtp server is not configured or not reachable, the call to
_smtp_send() will fail logging the failure (both in logging and sentry).

Hash dicts are now prefered by swh-objstorage, in order to support
individual hash collisions.

because this later is not yet available on debian, and we do not use new
API features anyway in test_cookers.py

Should fix deb packaging.

'local' has been deprecated for a while.

This is not a runtime issue, because when the supplied mode is DentryPerms.symlink,
the target is only used passed to os.symlink, which does allow str too.

This can happen for various reasons, such as running the vault on
a mirror that is partially replayed

objstorage.add() will change/remove its return value in a future version,
and we do not use it in the vault anyway.

This would be more usefully implemented by generating a (temporary) URL
for direct download of vault bundles from the azure bucket.

This fixes test build [1]

[1] https://jenkins.softwareheritage.org/view/swh%20master%20(draft)/job/DVAU/job/tests/1575/console

Related to T4305

instead of (soon-to-be-deprecated) swh-core's postgresql_fact one.

to reduce test execution time a bit.
Keep test_git_bare_cooker using a postgresql backend, since it depends
on it (support for displayname).

Fix "more than one target found for cross-reference 'ObjectType'"
sphinx warning.

That hook can be frustrating as it can discard a long commit message
if it finds a typo in it so better removing it.

It was a class parameter from the swh.core.api.asynchronous.RPCServerApp which got
removed some time ago.

This is not defined in the swh.core.api.RPCServerApp.

Related to T3922

black is considered stable since release 22.1.0 and the version
we are currently using is quite outdated and not compatible with
click 8.1.0, so it is time to bump it to its latest stable release.

Please note that E501 pycodestyle warning related to line length
is replaced by B950 one from flake8-bugbear as recommended by black.
https://black.readthedocs.io/en/stable/the_black_code_style/current_style.html#line-length

Related to T3922

pytest-postgresql 3.1.3 and pytest-redis 2.4.0 added support for
pytest >= 7 so we can now drop the pytest pinning.

Due to test modules being copied in subdirectories of the
build directory by setuptools, it makes pytest fail by raising
ImportPathMismatchError exceptions when invoked from root
directory of the module.

So ignore the build folder to discover tests.

check_config converts the "stored" configuration (with toplevel keys for
most services) into a config suitable for the ``get_vault`` function.

Ensure all tests use the same codepath to make sure it is properly
exercized, and that the vault backend object can be properly initialized
in all cases.

This makes the server "fixtured" config use the production structure,
with storage, objstorage, cache etc. set at the top level instead of
within the vault configuration key.

The azure backend cannot be run in the same thread as an asyncio
event loop: https://forge.softwareheritage.org/T3981

The objstorage no longer implements it.

- add the `get_datastore` function in `swh.vault`
- add the `get_current_version` method in `VaultBackend`,
- remove dbversion management from sql init script
- update tests accordingly

otherwise XXX

To install the new hook:

  $ pre-commit install -t commit-msg

PyPI cannot render :ref: or domain roles

We need to use the original names for commit/tag content to match their hash;
otherwise git-fsck/git-clone would error and make cooked repositories
unusable.