run mypy on the whole swh/ directory

enforce running mypy on the whole swh package each time

In D2311#54287, @zack wrote:

In D2311#54278, @douardda wrote:

seems to work just fine.
[...]

Found 3 errors in 1 file (checked 1 source file)

I don't think this is good enough, in the sense that in this test you've introduced a type error in the file that is also the only one modified here, whereas modifying one file might introduce type errors in *other* files.
What we want here is to always run mypy swh as a pre-commit test, no matter which files have been modified.

add 'args: []' in mypy's pre-commit config and fix mypy.ini for missing stubs

In fact using:

- repo: https://github.com/pre-commit/mirrors-mypy
  rev: 'master'
  hooks:
  - id: mypy
    args: []

rebased

rebased

rebase

closed by 4399d6c7bc0b

wrong ard diff --update...

rebase

mercurial should not be in this lot

This has been fixed by cb42fea77070

remove sql migration file

In D2295#53881, @anlambert wrote:

Apart the sql upgrade file which should not be here, looks good to me.

This task is the result of this discussion https://octodon.social/@joeyh/103133965986721726

I'm not convinced this really is a good idea, it makes it difficult to document. Why not just replace get_xxx(cls, args): by get_xxx(cls, **args), then we can write this yaml as (with a little bit of manipulation on the yaml->arguments transformation):

storage:
  - filter
  - storage:
    - buffer
    - storage:
      - remote
      - url: http://swh-storage:5002/

This needs one or 2 tests.

what's the current status of this task?

update/add headers in test files

rebase

ok now I can get rid of the spurious if in list_origins_by_producer...

rebase (hopefully)

this version should address the points raised by vlorentz

Refactor the tests to prevent these ugly injected functions

rebase + better ci message + s/_w_/_with_/g

rebase + keep the test_storage.py file for now

use 'page_token' as (opaque) pagination argument

In D2222#52502, @douardda wrote:

In D2222#52362, @vlorentz wrote:

I would rather we move to opaque tokens for all pagination stuff.

That's why I put a 'Note' in the Diff's description ;-)

In D2222#52362, @vlorentz wrote:

I would rather we move to opaque tokens for all pagination stuff.

Any chance we can close this now?

This might be "solved" by T2063

A few notes

This task should be completed once https://forge.softwareheritage.org/source/swh-docker/ is in a decent state.

rebase + small fix

reformat for style consistency in the function