Loaders use it, but it is ignored by the pg storage.
However, as the Cassandra storage uses swh-model to validate
its input, it refuses this input from the loaders (and journal).

This is a fix to workaround pip's inability to correctly solve
extra requirements (swh-model depends on swh-core[], but if other
packages depend on swh-model and swh-core[http], the 'http' extra
does not always get installed).

pgsql does not support it.

There is more repetition, but it's easier to read and
'%timeit Revision.from_dict(d)' is 5 times faster.

to the benefit of MultiHash api

Summary: to make sphinx happy.

Test Plan: make -C docs

Reviewers: ardumont

Reviewed By: ardumont

Differential Revision: https://forge.softwareheritage.org/D571

so pytest does not think they are test classes.

to clear deprecation warnings.

related to T1261

Summary: Related to T1261

Test Plan: before and after: 138 tests passed (2 skipped)

Reviewers: #reviewers!

Differential Revision: https://forge.softwareheritage.org/D520

and ensure it is included in the source package.

Related to T1265.

This aligns the new api with the old one.

Remove unused track_length parameter

And revert to their original behavior (returning only digest as bytes)

This allows calls like:

- MultiHash.from_file(file_object).digest()
- MultiHash.from_path(b'foo').hexdigest()
- MultiHash.from_data(b'foo').bytehexdigest()

This is the first step to improve the hashutil module according to
"the" plan [1].

In this regards, the hashutil exposes the same function as before.
Internally though, they now uses a MultiHash instance.

Related D410

[1] D410#7952