But keep support for deserializing from str, like
swh-storage does.

Code manipulating a Content object may want to access a hash
of configurable name; this method allows it to do that
without using getattr directly.

to the benefit of MultiHash api

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.

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

Related T421

Related T421

Related T1152

... and add an option to explicitly not follow them, if needed.

Closes T1135

with explicit option to hide it

Closes T1133

This commit changes the swh pids representation returned by the
parse_persistent_identifier function from a dict to a named tuple.
Also, the 'object_type' attribute of that named tuple now contains
a full name instead of an abbreviated one.

Closes T1112
Closes T1125

Close T1104
Close D346

Related T1104

As the client of this api might not be aware of what's the right id
key for a given object.

Related T1104

Related T1104

Related T1104

Related T1104