🚸(backend) improve users similarity search and sort results

[sampaccoud]

Purpose

In some edge cases, the domain part the email addresse is longer than the name part. Users searches by email similarity then return a lot of unsorted results.

Proposal

• We can improve this by being more demanding on similarity when the query looks like an email.
• Sorting results by the similarity score is also an obvious improvement.

At the moment, we still think it is good to propose results with a weak similarity on the name part because we want to avoid as much as possible creating duplicate users by inviting one of is many emails, a user who is already in our database.

Fixes #399

[qbey]

Looks good to me, we may adapt is again later, because it may be confusing we return other "non related" email address even if the user requested a specific one (eg the michael.johnson@example.gouv.fr test).

It's a nice will to allow user to find "similar" emails, but I'm not sure it will be that easy, for instance

    response = client.get("/api/v1.0/users/?q=ajohnson@other.gouv.fr")

    assert response.status_code == 200
    user_ids = [user["id"] for user in response.json()["results"]]
    assert user_ids == []  # I would have Alice here ^^' but it's too complex

I'm just raising the fact we should consider the tradeoff "user confusion" vs. "don't duplicate users".