Our `stop_sender`/`stop_receiver` watch channel is used to signal to
background tasks that they need to stop as we're about to shut down.

Previously, we stored both sides in our `Node` object. However, when
restarting the node, this would lead to us cloning a stale
`stop_receiver`, i.e., one that still had the change signal set, messing
with our shutdown logic on restarts.

Here, we instead drop the stored receiver and just create all
receivers by `subscribe`ing to the sender where necessary, which fixes
the prior bug.

While we previously took care of persisting peer details when a channel
is opened and if the user signals to `connect` to have the connection
persisted, our background task would still only try to reconnect to
channel peers. Here, we fix this omission and add a regression test
ensuring we'll see the proper behavior going forward.

…connection

Reconnect all persisted peers on restart

.. unfortunately the `windows-sys` dependency recently seems to have
broken their doc builds on rustc 1.63.0. As tests and builds still work
without issue, we fix CI by just not building docs on
`windows-latest`/`1.63.0`

…-doc-build

Fix CI by not building docs Windows on 1.63.0

The fee as returned from `PaymentSent` event generated by LDK and is
  saved in `PaymentSuccessful` event.

.. which will tell Rust not to treat it as an integration test module.

.. no need to use a `Result` here

Minor follow-up to lightningdevkit#271

.. we replace the simple `is_running` with a more verbose `status`
method returning a `NodeStatus` struct, giving more information on
syncing states etc.

…atus

Introduce `status` method allowing to query the `Node`'s status

…-for-msrv

Pin `reqwest` to fix MSRV builds in CI

.. switching to `dyn KVStore + Send + Sync`

... now that we don't have any generic to hide, we can just use the
`Node` type directly.

.. now that we can.

.. just a minor cleanup to further modularize the codebase. Also, we'll
be reusing these methods in `Event::ConnectionNeeded` soon.

.. we should consider dropping `Deref` and instead just commiting to
store a `Arc<L>` everwhere, as it gets tedious to maintain.

However, this is barely scraping by and the least invasive change here.

.. which is a bit more readable and in-line what we do other places,
plus it allows us to drop the `futures` dependency.

... we check that we can successfully issue concurrent connection
attempts, which all succeed.

Previously, concurrent calls to
`do_connect_peer`/`connect_peer_if_necessary` could result in multiple
connections being opened, just to be closed as redundant shortly after.
Parallel connection attempt were therefore prone to fail.

Here, we introduce a `ConnectionManager` that implements a pub/sub
pattern: upon the initial call, the task responsible for polling the
connection future to completion registers that a connection is
in-flight. Any calls following will check this and register a `oneshot`
channel to be notified of the `Result`.

.. which declutters our top-level docs a bit.

.. as we changed quite a bit and moved required fields, we add a test
here that adds the old `PaymentDetails` as a test struct and ensures
we're able to parse them just fine

.. allowing to configure the per-channel emergency reserve as well as
some trusted peers for which we won't maintain any reserve.

When Anchor outputs need to be spent LDK will generate
`BumpTransactionEvent`s. Here, we add the corresponding event-handling
and PSBT-signing support.

.. because they will be the new default.

Note the upcoming CLN 24.02 release will make Anchors default, too, but
for now we have to set the `experimental-anchors` config option.

.. which we somehow so far ommitted exposing in the API.

We now introduce a `force` flag to `close_channel` and broadcast if the
counterparty is not trusted.

…merge-anchor-support-from-upstream

# Conflicts:
#	.github/workflows/rust.yml
#	bindings/ldk_node.udl
#	src/event.rs
#	src/lib.rs
#	src/payment_store.rs
#	tests/common/mod.rs