Addresses https://github.com/matrix-org/synapse-dinsic/issues/70

This PR causes `ProxyAgent` to attempt to extract credentials from an `HTTPS_PROXY` env var. If credentials are found, a `Proxy-Authorization` header ([details](https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Proxy-Authorization)) is sent to the proxy server to authenticate against it. The headers are *not* passed to the remote server.

Also added some type hints.

Currently federation catchup will send the last *local* event that we
failed to send to the remote. This can cause issues for large rooms
where lots of servers have sent events while the remote server was down,
as when it comes back up again it'll be flooded with events from various
points in the DAG.

Instead, let's make it so that all the servers send the most recent
events, even if its not theirs. The remote should deduplicate the
events, so there shouldn't be much overhead in doing this.
Alternatively, the servers could only send local events if they were
also extremities and hope that the other server will send the event
over, but that is a bit risky.

This bug was discovered by DINUM. We were modifying `serialized_event["content"]`, which - if you've got `USE_FROZEN_DICTS` turned on or are [using a third party rules module](https://github.com/matrix-org/synapse/blob/17cd48fe5171d50da4cb59db647b993168e7dfab/synapse/events/third_party_rules.py#L73-L76) - will raise a 500 if you try to a edit a reply to a message.

`serialized_event["content"]` could be set to the edit event's content, instead of a copy of it, which is bad as we attempt to modify it. Instead, we also end up modifying the original event's content. DINUM uses a third party rules module, which meant the event's content got frozen and thus an exception was raised.

To be clear, the problem is not that the event's content was frozen. In fact doing so helped us uncover the fact we weren't copying event content correctly.

* Adds B00 to ignored checks.
* Fixes remaining issues.

Allows limiting who can login using OIDC via the claims
made from the IdP.

Instead of if the user does not have a password hash. This allows a SSO
user to add a password to their account, but only if the local password
database is configured.

Fixes https://github.com/matrix-org/synapse/issues/9572

When a SSO user logs in for the first time, we create a local Matrix user for them. This goes through the register_user flow, which ends up triggering the spam checker. Spam checker modules don't currently have any way to differentiate between a user trying to sign up initially, versus an SSO user (whom has presumably already been approved elsewhere) trying to log in for the first time.

This PR passes `auth_provider_id` as an argument to the `check_registration_for_spam` function. This argument will contain an ID of an SSO provider (`"saml"`, `"cas"`, etc.) if one was used, else `None`.

The stable format uses different brand identifiers, so we need to support two
identifiers for each IdP.

Federation catch up mode is very inefficient if the number of events
that the remote server has missed is small, since handling gaps can be
very expensive, c.f. #9492.

Instead of going into catch up mode whenever we see an error, we instead
do so only if we've backed off from trying the remote for more than an
hour (the assumption being that in such a case it is more than a
transient failure).

This uses a simplified version of get_chain_cover_difference to calculate
auth chain of events.

Apple had to be special. They want a client secret which is generated from an EC key.

Fixes #9220. Also fixes #9212 while I'm here.

Type hint fixes due to Twisted 21.2.0 adding type hints.

Properly uses RGBA mode for 1- and 8-bit images with transparency
(instead of RBG mode).

Unfortunately this doesn't test re-joining the room since
that requires having another homeserver to query over
federation, which isn't easily doable in unit tests.

This great big stack of commits is a a whole load of hoop-jumping to make it easier to store additional values in login tokens, and then to actually store the SSO Identity Provider in the login token. (Making use of that data will follow in a subsequent PR.)

By consuming the response if the headers imply that the
content is too large.

### Changes proposed in this PR

- Add support for the `no_proxy` and `NO_PROXY` environment variables
  - Internally rely on urllib's [`proxy_bypass_environment`](https://github.com/python/cpython/blob/bdb941be423bde8b02a5695ccf51c303d6204bed/Lib/urllib/request.py#L2519)
- Extract env variables using urllib's `getproxies`/[`getproxies_environment`](https://github.com/python/cpython/blob/bdb941be423bde8b02a5695ccf51c303d6204bed/Lib/urllib/request.py#L2488

) which supports lowercase + uppercase, preferring lowercase, except for `HTTP_PROXY` in a CGI environment

This does contain behaviour changes for consumers so making sure these are called out:
- `no_proxy`/`NO_PROXY` is now respected
- lowercase `https_proxy` is now allowed and taken over `HTTPS_PROXY`

Related to #9306 which also uses `ProxyAgent`

Signed-off-by: Timothy Leung <tim95@hotmail.co.uk>

... otherwise, we don't get the cookie back.

* Split ShardedWorkerHandlingConfig

This is so that we have a type level understanding of when it is safe to
call `get_instance(..)` (as opposed to `should_handle(..)`).

* Remove special cases in ShardedWorkerHandlingConfig.

`ShardedWorkerHandlingConfig` tried to handle the various different ways
it was possible to configure federation senders and pushers. This led to
special cases that weren't hit during testing.

To fix this the handling of the different cases is moved from there and
`generic_worker` into the worker config class. This allows us to have
the logic in one place and allows the rest of the code to ignore the
different cases.

`uploads_path` was a thing that was never used; most of it was removed in #6628
but a few vestiges remained.

This PR remove the cache for the `get_shared_rooms_for_users` storage method (the db method driving the experimental "what rooms do I share with this user?" feature: [MSC2666](https://github.com/matrix-org/matrix-doc/pull/2666)). Currently subsequent requests to the endpoint will return the same result, even if your shared rooms with that user have changed.

The cache was added in https://github.com/matrix-org/synapse/pull/7785, but we forgot to ensure it was invalidated appropriately.

Upon attempting to invalidate it, I found that the cache had to be entirely invalidated whenever a user (remote or local) joined or left a room. This didn't make for a very useful cache, especially for a function that may or may not be called very often. Thus, I've opted to remove it instead of invalidating it.

This PR attempts to eliminate unnecessary presence sending work when your local server joins a room, or when a remote server joins a room your server is participating in by processing state deltas in chunks rather than individually.

---

When your server joins a room for the first time, it requests the historical state as well. This chunk of new state is passed to the presence handler which, after filtering that state down to only membership joins, will send presence updates to homeservers for each join processed.

It turns out that we were being a bit naive and processing each event individually, and sending out presence updates for every one of those joins. Even if many different joins were users on the same server (hello IRC bridges), we'd send presence to that same homeserver for every remote user join we saw.

This PR attempts to deduplicate all of that by processing the entire batch of state deltas at once, instead of only doing each join individually. We process the joins and note down which servers need which presence:

* If it was a local user join, send that user's latest presence to all servers in the room
* If it was a remote user join, send the presence for all local users in the room to that homeserver

We deduplicate by inserting all of those pending updates into a dictionary of the form:

```
{
  server_name1: {presence_update1, ...},
  server_name2: {presence_update1, presence_update2, ...}
}
```

Only after building this dict do we then start sending out presence updates.

This PR adds a homeserver config option, `user_directory.prefer_local_users`, that when enabled will show local users higher in user directory search results than remote users. This option is off by default.

Note that turning this on doesn't necessarily mean that remote users will always be put below local users, but they should be assuming all other ranking factors (search query match, profile information present etc) are identical.

This is useful for, say, University networks that are openly federating, but want to prioritise local students and staff in the user directory over other random users.

Add off-by-default configuration settings to:
- disable putting an invitee's profile info in invite events
- disable profile lookup via federation

Signed-off-by: Andrew Ferrazzutti <fair@miscworks.net>

This reduces the memory usage of previewing media files which
end up larger than the `max_spider_size` by avoiding buffering
content internally in treq.

It also checks the `Content-Length` header in additional places
instead of streaming the content to check the body length.

Apple want to POST the OIDC auth response back to us rather than using query-params; add the necessary support to make that work.

 - Update black version to the latest
 - Run black auto formatting over the codebase
    - Run autoformatting according to [`docs/code_style.md
`](https://github.com/matrix-org/synapse/blob/80d6dc9783aa80886a133756028984dbf8920168/docs/code_style.md)
 - Update `code_style.md` docs around installing black to use the correct version