Handle LND startup correctly

[fusion44]

On low power systems like RaspiBlitz it can take longer to start up LND. Refactor the startup sequence to take this properly into account.

LND message sequence on startup:

• LND refuses connections until a certain point
• LND returns a message "waiting to start, RPC services not available"
• LND returns a message "Wallet locked"

Possible solution: Add a _handle_lightning_ready function before checking for wallet locked. Each lightning implementation should have their own checking function.

[fusion44]

During startup the API will try to connect to Bitcoin Core and the Lightning Node. If it can't connect it will check every n seconds (currently 2s) and connect when available. A new SSE event called "system_startup_info" is introduced. This event contains all startup status information during the startup procedure.

The old wallet_locked event is no more, thus the Frontend will require a bit of a modification to account for the changes. This is much more robust and future proof than the hack we had before.

known issues:

• Warmup for new SSE connections don't yet work
• CLN implementation is missing

Event samples:

event: system_startup_info
data: {"bitcoin": "offline", "bitcoin_msg": "", "lightning": "offline", "lightning_msg": "Unable to connect to LND daemon, waiting..."}

event: system_startup_info
data: {"bitcoin": "done", "bitcoin_msg": "", "lightning": "offline", "lightning_msg": "Unable to connect to LND daemon, waiting..."}

event: system_startup_info
data: {"bitcoin": "done", "bitcoin_msg": "", "lightning": "locked", "lightning_msg": "Wallet locked, unlock it to enable full RPC access"}

event: system_startup_info
data: {"bitcoin": "done", "bitcoin_msg": "", "lightning": "done", "lightning_msg": ""}

event: system_startup_info
data: {"bitcoin": "done", "bitcoin_msg": "", "lightning": "bootstraping", "lightning_msg": "RPC not yet available"}