ASCS Source Implementation

[zxzxwu]

• Implement ASE state transition for source
• Make ASE emit state changed events
• Replace pdu event with sink which is more reasonable and have a better type annotation
• Implement a LEA app
• Improve CIS handling in device and host

[zxzxwu]

SCO has the same issue. I think one solution could be - again, using microphone input from a web frontend.

[barbibulle]

SCO has the same issue. I think one solution could be - again, using microphone input from a web frontend.

You could read the input from an LC3 encoded audio file, like for the A2DP source example.

[zxzxwu]

Consider the difficulty of audio prcessing and state synchronization between Python and Web Frontend, I think it's better to host codec in Python instead of JS.

The liblc3.wasm file is built from liblc3 source. With wasmtime, it could be executed on all platforms (with CPython only?). Not sure whether there is a better way to pack liblc3?

[zxzxwu]

Not sure how to solve this...

thread 'tokio-runtime-worker' panicked at /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/pyo3-asyncio-0.18.0/src/testing.rs:285:39:
called `Result::unwrap()` on an `Err` value: PyErr { type: <class 'ModuleNotFoundError'>, value: ModuleNotFoundError("No module named 'cryptography'"), traceback: Some(<traceback object at 0x7bb4710b75c0>) }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Traceback (most recent call last):
  File "/usr/lib/python3.9/asyncio/base_events.py", line 642, in run_until_complete
    return future.result()
pyo3_asyncio.RustPanic: rust future panicked
thread 'main' panicked at pytests/pytests.rs:15:1:
called `Result::unwrap()` on an `Err` value: ()
error: test failed, to rerun pass `--test pytests`

[barbibulle]

Not sure how to solve this...

thread 'tokio-runtime-worker' panicked at /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/pyo3-asyncio-0.18.0/src/testing.rs:285:39:
called `Result::unwrap()` on an `Err` value: PyErr { type: <class 'ModuleNotFoundError'>, value: ModuleNotFoundError("No module named 'cryptography'"), traceback: Some(<traceback object at 0x7bb4710b75c0>) }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Traceback (most recent call last):
  File "/usr/lib/python3.9/asyncio/base_events.py", line 642, in run_until_complete
    return future.result()
pyo3_asyncio.RustPanic: rust future panicked
thread 'main' panicked at pytests/pytests.rs:15:1:
called `Result::unwrap()` on an `Err` value: ()
error: test failed, to rerun pass `--test pytests`

When do you get this error?

[zxzxwu]

It looks like an issue on my local machine. I tried again with GH workflow and then get

thread 'tokio-runtime-worker' panicked at /home/runner/.cargo/registry/src/index.crates.io-6f17d22bba15001f/pyo3-asyncio-0.18.0/src/testing.rs:285:39:
called `Result::unwrap()` on an `Err` value: PyErr { type: <class 'RuntimeError'>, value: RuntimeError("There is no current event loop in thread 'Dummy-1'."), traceback: Some(<traceback object at 0x7f421426b6c0>) }
stack backtrace:
   0: rust_begin_unwind
             at /rustc/07dca489ac2d933c78d3c51[58](https://github.com/zxzxwu/bumble/actions/runs/8964830670/job/24617243180?pr=5#step:10:59)e3f43beefeb02ce/library/std/src/panicking.rs:645:5
   1: core::panicking::panic_fmt
             at /rustc/07dca489ac2d933c78d3c5158e3f43beefeb02ce/library/core/src/panicking.rs:72:14
   2: core::result::unwrap_failed
             at /rustc/07dca489ac2d933c78d3c5158e3f43beefeb02ce/library/core/src/result.rs:1[64](https://github.com/zxzxwu/bumble/actions/runs/8964830670/job/24617243180?pr=5#step:10:65)9:5
   3: core::result::Result<T,E>::unwrap
   4: pyo3_asyncio::testing::test_harness::{{closure}}::{{closure}}::{{closure}}
   5: <futures_util::stream::futures_unordered::FuturesUnordered<Fut> as futures_core::stream::Stream>::poll_next
   6: futures_util::stream::stream::StreamExt::poll_next_unpin
   7: <futures_util::stream::stream::for_each_concurrent::ForEachConcurrent<St,Fut,F> as core::future::future::Future>::poll
   8: pyo3_asyncio::testing::test_harness::{{closure}}
   9: pyo3_asyncio::testing::main::{{closure}}

After some investigation, the root cause of this seems to be async lock?

[zxzxwu]

Since we started working on Broadcast, we may merge this PR first to avoid conflict.

[barbibulle]

Since we started working on Broadcast, we may merge this PR first to avoid conflict.

Yes. I wasn't sure if you were ready with this one yet or not. I'll do a quick re-review today.

[zxzxwu]

Yes, I think it's ready.

[barbibulle]

I recommend not sending the state in the event. The normal pattern is that if you notify of a change for a value that's already part of an object's properties, you don't need to repeat the value that changed, because the listener can access it from the object itself. For example, in device.py, when the encryption of the connection changes, we emit connection.emit('connection_encryption_change'). The listener can then look at connection.encryption to see what the new value is.

[zxzxwu]

Sounds reasonable. It is really a pain point to use eventemitter - Despite it benefits Bumble a lot, the missing of type hint is something unstable.