# Copyright 2024 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#      https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# -----------------------------------------------------------------------------
# Imports
# -----------------------------------------------------------------------------
from __future__ import annotations
import pprint
import asyncio
import contextlib
import logging
import wave
import itertools
import struct
from typing import cast, Any, AsyncGenerator, Coroutine, List
import itertools
import glob
import time
import threading

import numpy as np  # for audio down-mix
import samplerate
import os

import lc3  # type: ignore  # pylint: disable=E0401

from bumble.colors import color
from bumble import company_ids
from bumble import core
from bumble import gatt
from bumble import hci
from bumble.profiles import bap
from bumble.profiles import le_audio
from bumble.profiles import pbp
from bumble.profiles import bass
import bumble.device
import bumble.transport
import bumble.utils
from bumble.device import Host, AdvertisingChannelMap
from bumble.audio import io as audio_io

from auracast import auracast_config
from auracast.utils.read_lc3_file import read_lc3_file
from auracast.utils.network_audio_receiver import NetworkAudioReceiverUncoded
from auracast.utils.webrtc_audio_input import WebRTCAudioInput


# Patch sounddevice.InputStream globally to use low-latency settings
import alsaaudio
from collections import deque


class AlsaArecordAudioInput(audio_io.AudioInput):
    def __init__(self, device_name: str, pcm_format: audio_io.PcmFormat):
        self._device_name = device_name
        self._pcm_format = pcm_format
        self._proc: asyncio.subprocess.Process | None = None

    async def open(self) -> audio_io.PcmFormat:
        if self._proc is not None:
            return self._pcm_format

        args = [
            'arecord',
            '-D', self._device_name,
            '-q',
            '-t', 'raw',
            '-f', 'S16_LE',
            '-r', str(int(self._pcm_format.sample_rate)),
            '-c', str(int(self._pcm_format.channels)),
        ]

        logging.info(
            "Opening ALSA capture via arecord: device='%s' rate=%s ch=%s",
            self._device_name,
            self._pcm_format.sample_rate,
            self._pcm_format.channels,
        )

        self._proc = await asyncio.create_subprocess_exec(
            *args,
            stdout=asyncio.subprocess.PIPE,
            stderr=asyncio.subprocess.DEVNULL,
        )

        if self._proc.stdout is None:
            raise RuntimeError('arecord stdout pipe was not created')

        return self._pcm_format

    def frames(self, frame_size: int) -> AsyncGenerator[bytes]:
        async def _gen() -> AsyncGenerator[bytes]:
            if self._proc is None:
                await self.open()

            if self._proc is None or self._proc.stdout is None:
                return

            bytes_per_frame = frame_size * self._pcm_format.channels * self._pcm_format.bytes_per_sample

            while True:
                try:
                    data = await self._proc.stdout.readexactly(bytes_per_frame)
                except asyncio.IncompleteReadError:
                    return
                except Exception:
                    return
                yield data

        return _gen()

    async def aclose(self) -> None:
        if self._proc is None:
            return
        try:
            if self._proc.returncode is None:
                self._proc.terminate()
        except ProcessLookupError:
            pass
        except Exception:
            pass
        with contextlib.suppress(Exception):
            await asyncio.wait_for(self._proc.wait(), timeout=1.0)
        if self._proc.returncode is None:
            with contextlib.suppress(Exception):
                self._proc.kill()
            with contextlib.suppress(Exception):
                await asyncio.wait_for(self._proc.wait(), timeout=1.0)
        self._proc = None


class PyAlsaAudioInput(audio_io.ThreadedAudioInput):
    """PyALSA audio input with non-blocking reads - supports mono/stereo."""

    def __init__(self, device, pcm_format: audio_io.PcmFormat):
        super().__init__()
        logging.info("PyALSA: device = %s", device)
        self._device = str(device) if not isinstance(device, str) else device
        if self._device.isdigit():
            self._device = 'default' if self._device == '0' else f'hw:{self._device}'
        self._pcm_format = pcm_format
        self._pcm = None
        self._actual_channels = None
        self._periodsize = None
        self._hw_channels = None
        self._first_read = True
        self._resampler = None
        self._resampler_buffer = np.empty(0, dtype=np.float32)

    def _open(self) -> audio_io.PcmFormat:
        ALSA_PERIODSIZE = 240
        ALSA_PERIODS = 4
        ALSA_MODE = alsaaudio.PCM_NONBLOCK
        
        requested_rate = int(self._pcm_format.sample_rate)
        requested_channels = int(self._pcm_format.channels)
        self._periodsize = ALSA_PERIODSIZE
        
        self._pcm = alsaaudio.PCM(
            type=alsaaudio.PCM_CAPTURE,
            mode=ALSA_MODE,
            device=self._device,
            periods=ALSA_PERIODS,
        )
        
        self._pcm.setchannels(requested_channels)
        self._pcm.setformat(alsaaudio.PCM_FORMAT_S16_LE)
        actual_rate = self._pcm.setrate(requested_rate)
        self._pcm.setperiodsize(ALSA_PERIODSIZE)
        
        logging.info("PyALSA: device=%s rate=%d ch=%d periodsize=%d (%.1fms) periods=%d mode=%s", 
                     self._device, actual_rate, requested_channels, ALSA_PERIODSIZE, 
                     (ALSA_PERIODSIZE / actual_rate) * 1000, ALSA_PERIODS, ALSA_MODE)
        
        if actual_rate != requested_rate:
            logging.warning("PyALSA: Sample rate mismatch! requested=%d actual=%d", requested_rate, actual_rate)
        
        self._actual_channels = requested_channels
        self._resampler = samplerate.Resampler('sinc_fastest', channels=requested_channels)
        self._resampler_buffer = np.empty(0, dtype=np.float32)

        self._bang_bang = 0
        
        return audio_io.PcmFormat(
            audio_io.PcmFormat.Endianness.LITTLE,
            audio_io.PcmFormat.SampleType.INT16,
            actual_rate,
            requested_channels,
        )

    def _read(self, frame_size: int) -> bytes:
        try:
            avail = self._pcm.avail()
            logging.debug("PyALSA: avail before read: %d", avail)
            length, data = self._pcm.read_sw(frame_size + self._bang_bang)
            avail = self._pcm.avail()
            SETPOINT = 120 
            TOLERANCE = 80 
            if avail < SETPOINT - TOLERANCE:
                self._bang_bang = -1 
            elif avail > SETPOINT + TOLERANCE:
                self._bang_bang = 1
            else:
                self._bang_bang = 0
                
                
            logging.debug("PyALSA: read length=%d, data length=%d, avail=%d, bang_bang=%d", length, len(data), avail, self._bang_bang)

            if length > 0:
                if self._first_read:
                    expected_mono = self._periodsize * 2
                    expected_stereo = self._periodsize * 2 * 2
                    # self._hw_channels = 2 if len(data) == expected_stereo else 1
                    self._hw_channels = self._actual_channels
                    logging.info("PyALSA first read: bytes=%d detected_hw_channels=%d requested_channels=%d", 
                                len(data), self._hw_channels, self._actual_channels)
                    self._first_read = False
                
                if self._hw_channels == 2 and self._actual_channels == 1:
                    pcm_stereo = np.frombuffer(data, dtype=np.int16)
                    pcm_mono = pcm_stereo[::2]
                    data = pcm_mono.tobytes()
                
                actual_samples = len(data) // (2 * self._actual_channels)
                ratio = frame_size / actual_samples
                pcm_f32 = np.frombuffer(data, dtype=np.int16).astype(np.float32) / 32768.0
                if self._actual_channels > 1:
                    pcm_f32 = pcm_f32.reshape(-1, self._actual_channels)
                resampled = self._resampler.process(pcm_f32, ratio, end_of_input=False)
                if self._actual_channels > 1:
                    resampled = resampled.reshape(-1)
                self._resampler_buffer = np.concatenate([self._resampler_buffer, resampled])
            else:
                logging.warning("PyALSA: No data read from ALSA")
                self._resampler_buffer = np.concatenate([
                    self._resampler_buffer,
                    np.zeros(frame_size * self._actual_channels, dtype=np.float32),
                ])
        except alsaaudio.ALSAAudioError as e:
            logging.error("PyALSA: ALSA read error: %s", e)
            self._resampler_buffer = np.concatenate([
                self._resampler_buffer,
                np.zeros(frame_size * self._actual_channels, dtype=np.float32),
            ])
        except Exception as e:
            logging.error("PyALSA: Unexpected error in _read: %s", e, exc_info=True)
            self._resampler_buffer = np.concatenate([
                self._resampler_buffer,
                np.zeros(frame_size * self._actual_channels, dtype=np.float32),
            ])

        needed = frame_size * self._actual_channels
        if len(self._resampler_buffer) < needed:
            pad = np.zeros(needed - len(self._resampler_buffer), dtype=np.float32)
            self._resampler_buffer = np.concatenate([self._resampler_buffer, pad])
            logging.debug("PyALSA: padded buffer with %d samples", needed - len(self._resampler_buffer))

        output = self._resampler_buffer[:needed]
        self._resampler_buffer = self._resampler_buffer[needed:]

        logging.debug("PyALSA: resampler_buffer remaining=%d", len(self._resampler_buffer))
        return np.clip(output * 32767.0, -32768, 32767).astype(np.int16).tobytes()

  

    def _close(self) -> None:
        if self._pcm:
            self._pcm.close()
            self._pcm = None

audio_io.SoundDeviceAudioInput = PyAlsaAudioInput

# modified from bumble
class ModWaveAudioInput(audio_io.ThreadedAudioInput):
    """Audio input that reads PCM samples from a .wav file."""

    def __init__(self, filename: str) -> None:
        super().__init__()
        self._filename = filename
        self._wav: wave.Wave_read | None = None
        self._bytes_read = 0
        self.rewind=True

    def _open(self) -> audio_io.PcmFormat:
        self._wav = wave.open(self._filename, 'rb')
        if self._wav.getsampwidth() != 2:
            raise ValueError('sample width not supported')
        return audio_io.PcmFormat(
            audio_io.PcmFormat.Endianness.LITTLE,
            audio_io.PcmFormat.SampleType.INT16,
            self._wav.getframerate(),
            self._wav.getnchannels(),
        )

    def _read(self, frame_size: int) -> bytes:
        if not self._wav:
            return b''

        pcm_samples = self._wav.readframes(frame_size)
        if not pcm_samples and self._bytes_read:
            if not self.rewind:
                return None
            # Loop around.
            self._wav.rewind()
            self._bytes_read = 0
            pcm_samples = self._wav.readframes(frame_size)

        self._bytes_read += len(pcm_samples)
        return pcm_samples

    def _close(self) -> None:
        if self._wav:
            self._wav.close()

audio_io.WaveAudioInput = ModWaveAudioInput


def broadcast_code_bytes(broadcast_code: str) -> bytes:
    """
    Convert a broadcast code string to a 16-byte value.

    If `broadcast_code` is `0x` followed by 32 hex characters, it is interpreted as a
    raw 16-byte raw broadcast code in big-endian byte order.
    Otherwise, `broadcast_code` is converted to a 16-byte value as specified in
    BLUETOOTH CORE SPECIFICATION Version 6.0 | Vol 3, Part C , section 3.2.6.3
    """
    if broadcast_code.startswith("0x") and len(broadcast_code) == 34:
        return bytes.fromhex(broadcast_code[2:])[::-1]

    broadcast_code_utf8 = broadcast_code.encode("utf-8")
    if len(broadcast_code_utf8) > 16:
        raise ValueError("broadcast code must be <= 16 bytes in utf-8 encoding")
    padding = bytes(16 - len(broadcast_code_utf8))
    return broadcast_code_utf8 + padding

# -----------------------------------------------------------------------------
# Logging
# -----------------------------------------------------------------------------
logger = logging.getLogger(__name__)

@contextlib.asynccontextmanager
async def create_device(config: auracast_config.AuracastGlobalConfig) -> AsyncGenerator[bumble.device.Device, Any]:
    async with await bumble.transport.open_transport(config.transport) as (
        hci_source,
        hci_sink,
    ):
        device_config = bumble.device.DeviceConfiguration(
            name=config.device_name,
            address= hci.Address(config.auracast_device_address),
            keystore='JsonKeyStore',
            #le_simultaneous_enabled=True #TODO: What is this doing ?
        )

        device = bumble.device.Device.from_config_with_hci(
            device_config,
            hci_source,
            hci_sink,
        )
        await device.power_on()

        yield device


def run_async(async_command: Coroutine) -> None:
    try:
        asyncio.run(async_command)
    except core.ProtocolError as error:
        if error.error_namespace == 'att' and error.error_code in list(
            bass.ApplicationError
        ):
            message = bass.ApplicationError(error.error_code).name
        else:
            message = str(error)

        print(
            color('!!! An error occurred while executing the command:', 'red'), message
        )

def _build_bis_list(num_bis: int) -> list:
    """Build BIS list for BasicAudioAnnouncement based on num_bis (1=mono, 2=stereo)."""
    locations = [bap.AudioLocation.FRONT_LEFT, bap.AudioLocation.FRONT_RIGHT]
    return [
        bap.BasicAudioAnnouncement.BIS(
            index=idx + 1,
            codec_specific_configuration=bap.CodecSpecificConfiguration(
                audio_channel_allocation=locations[idx]
            ),
        )
        for idx in range(num_bis)
    ]


async def init_broadcast(
        device,
        global_config : auracast_config.AuracastGlobalConfig,
        big_config: List[auracast_config.AuracastBigConfig]
        ) -> dict:

    bap_sampling_freq = getattr(bap.SamplingFrequency, f"FREQ_{global_config.auracast_sampling_rate_hz}")
    bigs = {}
    for i, conf in enumerate(big_config):
        metadata=le_audio.Metadata(
            [
                le_audio.Metadata.Entry(
                    tag=le_audio.Metadata.Tag.LANGUAGE, data=conf.language.encode()
                ),
                le_audio.Metadata.Entry(
                    tag=le_audio.Metadata.Tag.PROGRAM_INFO,
                    data=conf.program_info.encode('latin-1')
                ),
                le_audio.Metadata.Entry(
                    tag=le_audio.Metadata.Tag.BROADCAST_NAME, data=conf.name.encode()
                ),
            ]
            + (
                [
                    # Broadcast Audio Immediate Rendering flag (type 0x09), zero-length value
                    le_audio.Metadata.Entry(tag = le_audio.Metadata.Tag.BROADCAST_AUDIO_IMMEDIATE_RENDERING_FLAG, data=b"")
                ]
                if global_config.immediate_rendering
                else []
            )
            + (
                [
                    # Assisted Listening Stream tag expects a 1-octet value. Use 0x01 to indicate enabled.
                    le_audio.Metadata.Entry(tag = le_audio.Metadata.Tag.ASSISTED_LISTENING_STREAM, data=b"\x01")
                ]
                if global_config.assisted_listening_stream
                else []
            )
        )
        try:
            logging.info(metadata.pretty_print("\n"))
        except UnicodeDecodeError:
            logging.info("Metadata: (contains non-UTF-8 bytes)")
        bigs[f'big{i}'] = {}
        # Config advertising set
        bigs[f'big{i}']['basic_audio_announcement'] = bap.BasicAudioAnnouncement(
            presentation_delay=global_config.presentation_delay_us,
            subgroups=[
                bap.BasicAudioAnnouncement.Subgroup(
                    codec_id=hci.CodingFormat(codec_id=hci.CodecID.LC3),
                    codec_specific_configuration=bap.CodecSpecificConfiguration(
                        sampling_frequency=bap_sampling_freq,
                        frame_duration=bap.FrameDuration.DURATION_7500_US if global_config.frame_duration_us == 7500 else bap.FrameDuration.DURATION_10000_US,
                        octets_per_codec_frame=global_config.octets_per_frame,
                    ),
                    metadata=metadata,
                    bis=_build_bis_list(conf.num_bis),
                )
            ],
        )
        logger.info('Setup Advertising')
        advertising_manufacturer_data = (
        b''
        if global_config.manufacturer_data == (None, None)
        else bytes(
            core.AdvertisingData(
                [
                    (
                        core.AdvertisingData.MANUFACTURER_SPECIFIC_DATA,
                        struct.pack('<H', global_config.manufacturer_data[0])
                        + global_config.manufacturer_data[1],
                    )
                ]
            )
        )
    )
        bigs[f'big{i}']['broadcast_audio_announcement'] = bap.BroadcastAudioAnnouncement(conf.id)

        # Build advertising data types list
        advertising_data_types = [
            (core.AdvertisingData.BROADCAST_NAME, conf.name.encode()),
        ]

        # [PBP] Add Public Broadcast Profile Service Data (UUID 0x1856)
        # Required for PTS Qualification (PBP/PBS/STR)
        # Dynamically calculate PBP features based on stream configuration
        pbp_features = 0x00

        # Bit 0: Encryption (set if broadcast_code is configured)
        if conf.code is not None:
            pbp_features |= 0x01

        # Bit 1 vs Bit 2: Quality based on sample rate
        if global_config.auracast_sampling_rate_hz in [16000, 24000]:
            pbp_features |= 0x02  # Standard Quality
        elif global_config.auracast_sampling_rate_hz == 48000:
            pbp_features |= 0x04  # High Quality

        # Build PBP service data with Program_Info metadata (LTV format: Length, Type=0x03, Value)
        # LTV: Length = 1 (type) + len(value), Type = 0x03 (Program_Info)
        program_info_bytes = conf.program_info.encode('latin-1')
        pbp_metadata_ltv = bytes([len(program_info_bytes) + 1, 0x03]) + program_info_bytes
        pbp_service_data = struct.pack('<H', 0x1856) + bytes([pbp_features, len(pbp_metadata_ltv)]) + pbp_metadata_ltv
        advertising_data_types.append(
            (core.AdvertisingData.SERVICE_DATA_16_BIT_UUID, pbp_service_data)
        )

        advertising_set = await device.create_advertising_set(
            random_address=hci.Address(conf.random_address),
            advertising_parameters=bumble.device.AdvertisingParameters(
                advertising_event_properties=bumble.device.AdvertisingEventProperties(
                    is_connectable=False
                ),
                primary_advertising_interval_min=100,
                primary_advertising_interval_max=200,
                advertising_sid=i,
                primary_advertising_phy=hci.Phy.LE_1M, # 2m phy config throws error - because for primary advertising channels, 1mbit is only supported
                secondary_advertising_phy=hci.Phy.LE_1M, # this is the secondary advertising beeing send on non advertising channels (extendend advertising)
                #advertising_tx_power= # tx power in dbm (max 20)
                #secondary_advertising_max_skip=10,
            ),
            advertising_data=(
                bigs[f'big{i}']['broadcast_audio_announcement'].get_advertising_data()
                + bytes(core.AdvertisingData(advertising_data_types))
                + advertising_manufacturer_data
            ),
            periodic_advertising_parameters=bumble.device.PeriodicAdvertisingParameters(
                periodic_advertising_interval_min=80,
                periodic_advertising_interval_max=160,
            ),
            periodic_advertising_data=bigs[f'big{i}']['basic_audio_announcement'].get_advertising_data(),
            auto_restart=True,
            auto_start=True,
        )
        bigs[f'big{i}']['advertising_set'] = advertising_set

        logging.info('Start Periodic Advertising')
        await advertising_set.start_periodic()

        logging.info('Setup BIG')
        if global_config.qos_config.iso_int_multiple_10ms == 1:
            frame_enable = 0
        else:
            frame_enable = 1

        big = await device.create_big(
            bigs[f'big{i}']['advertising_set'],
            parameters=bumble.device.BigParameters(
                num_bis=conf.num_bis,
                sdu_interval=global_config.qos_config.iso_int_multiple_10ms*10000, # Is the same as iso interval
                max_sdu=global_config.octets_per_frame,
                max_transport_latency=global_config.qos_config.max_transport_latency_ms,
                rtn=global_config.qos_config.number_of_retransmissions,
                broadcast_code=(
                    broadcast_code_bytes(conf.code) if conf.code else None
                ),
                framing=frame_enable # needed if iso interval is not frame interval of codedc
            ),
        )
        bigs[f'big{i}']['big'] = big

        for bis_link in big.bis_links:
            await bis_link.setup_data_path(
                direction=bis_link.Direction.HOST_TO_CONTROLLER
            )

        # Create ISO queue(s) - one per BIS
        iso_queues = [
            bumble.device.IsoPacketStream(link, conf.iso_que_len)
            for link in big.bis_links
        ]

        logging.info('Setup ISO Data Path')

        bigs[f'big{i}']['iso_queues'] = iso_queues
        bigs[f'big{i}']['num_bis'] = conf.num_bis
        # Keep backward compat: iso_queue points to first queue
        bigs[f'big{i}']['iso_queue'] = iso_queues[0]

        if global_config.debug:
            logging.info(f'big{i} parameters are:')
            logging.info('%s', pprint.pformat(vars(big)))
        logging.info(f'Finished setup of big{i}.')

    await asyncio.sleep(i+1) # Wait for advertising to set up

    def on_flow():
        data_packet_queue = iso_queue.data_packet_queue
        logging.info(
            f'\rPACKETS: pending={data_packet_queue.pending}, '
            f'queued={data_packet_queue.queued}, '
            f'completed={data_packet_queue.completed}',
            end='',
        )

    if global_config.debug:
        bigs[f'big{0}']['iso_queue'].data_packet_queue.on('flow', on_flow)

    return bigs


class Streamer():
    """
    Streamer class that supports multiple input formats. See bumble for streaming from wav or device
    Added functionallity on top of bumble:
        - loop parameter 
            - if True the audio_source will be looped for ever
        - precode wav files
        - lc3 coded files
            - just use a .lc3 file as audio_source
        - lc3 coded from ram
            - use a bytestring b'' as audio_source
    """

    def __init__(
            self,
            bigs,
            global_config : auracast_config.AuracastGlobalConfig,
            big_config: List[auracast_config.AuracastBigConfig]
            ):
        self.task = None
        self.is_streaming = False
        self.bigs = bigs
        self.global_config = global_config
        self.big_config = big_config

    def start_streaming(self):
        if not self.is_streaming:
            self.task = asyncio.create_task(self.stream())
        else:
            logging.warning('Streamer is already running')

    async def stop_streaming(self):
        """Gracefully stop streaming and release audio devices."""
        if not self.is_streaming and self.task is None:
            return

        # Ask the streaming loop to finish
        self.is_streaming = False
        if self.task is not None:
            self.task.cancel()

            # Let cancellation propagate to the stream() coroutine
            await asyncio.sleep(0.01)

            self.task = None

        # Close audio inputs (await to ensure ALSA devices are released)
        async_closers = []
        sync_closers = []
        for big in self.bigs.values():
            ai = big.get("audio_input")
            if not ai:
                continue
            # First close any frames generator backed by the input to stop reads
            frames_gen = big.get("frames_gen")
            if frames_gen and hasattr(frames_gen, "aclose"):
                try:
                    await frames_gen.aclose()
                except Exception:
                    pass
                big.pop("frames_gen", None)
            if hasattr(ai, "aclose") and callable(getattr(ai, "aclose")):
                async_closers.append(ai.aclose())
            elif hasattr(ai, "close") and callable(getattr(ai, "close")):
                sync_closers.append(ai.close)
            # Remove reference so a fresh one is created next time
            big.pop("audio_input", None)

        if async_closers:
            await asyncio.gather(*async_closers, return_exceptions=True)
        for fn in sync_closers:
            try:
                fn()
            except Exception:
                pass

        # Reset PortAudio to drop lingering PipeWire capture nodes
        try:
            import sounddevice as _sd
            if hasattr(_sd, "_terminate"):
                _sd._terminate()
            await asyncio.sleep(0.05)
            if hasattr(_sd, "_initialize"):
                _sd._initialize()
        except Exception:
            pass

    def get_audio_levels(self) -> list[float]:
        """Return current RMS audio levels (0.0-1.0) for each BIG."""
        if not self.bigs:
            return []
        return [big.get('_audio_level_rms', 0.0) for big in self.bigs.values()]

    async def stream(self):

        bigs = self.bigs
        big_config = self.big_config
        global_config = self.global_config
        for i, big in enumerate(bigs.values()):
            audio_source = big_config[i].audio_source
            input_format = big_config[i].input_format

            # --- New: network_uncoded mode using NetworkAudioReceiver ---
            if isinstance(audio_source, NetworkAudioReceiverUncoded):
                # Start the UDP receiver coroutine so packets are actually received
                asyncio.create_task(audio_source.receive())
                encoder = lc3.Encoder(
                    frame_duration_us=global_config.frame_duration_us,
                    sample_rate_hz=global_config.auracast_sampling_rate_hz,
                    num_channels=1,
                    input_sample_rate_hz=audio_source.samplerate,
                )
                lc3_frame_samples = encoder.get_frame_samples()
                big['pcm_bit_depth'] = 16
                big['lc3_frame_samples'] = lc3_frame_samples
                big['lc3_bytes_per_frame'] = global_config.octets_per_frame
                big['audio_input'] = audio_source
                big['encoder'] = encoder
                big['precoded'] = False
                # Prepare frames generator for graceful shutdown
                big['frames_gen'] = big['audio_input'].frames(lc3_frame_samples)

            elif audio_source == 'webrtc':
                big['audio_input'] = WebRTCAudioInput()
                encoder = lc3.Encoder(
                    frame_duration_us=global_config.frame_duration_us,
                    sample_rate_hz=global_config.auracast_sampling_rate_hz,
                    num_channels=1,
                    input_sample_rate_hz=48000, # TODO: get samplerate from webrtc
                )
                lc3_frame_samples = encoder.get_frame_samples()
                big['pcm_bit_depth'] = 16
                big['lc3_frame_samples'] = lc3_frame_samples
                big['lc3_bytes_per_frame'] = global_config.octets_per_frame
                big['encoder'] = encoder
                big['precoded'] = False
                # Prepare frames generator for graceful shutdown
                big['frames_gen'] = big['audio_input'].frames(lc3_frame_samples)

            # precoded lc3 from ram
            elif isinstance(big_config[i].audio_source, bytes):
                big['precoded'] = True
                big['lc3_bytes_per_frame'] = global_config.octets_per_frame

                lc3_frames = iter(big_config[i].audio_source)

                if big_config[i].loop:
                    lc3_frames = itertools.cycle(lc3_frames)
                big['lc3_frames'] = lc3_frames

            # precoded lc3 file
            elif big_config[i].audio_source.endswith('.lc3'):
                big['precoded'] = True
                big['lc3_bytes_per_frame'] = global_config.octets_per_frame
                filename = big_config[i].audio_source.replace('file:', '')

                lc3_bytes = read_lc3_file(filename)
                lc3_frames = iter(lc3_bytes)

                if big_config[i].loop:
                    lc3_frames = itertools.cycle(lc3_frames)
                big['lc3_frames'] = lc3_frames

            # use wav files and code them entirely before streaming
            elif big_config[i].precode_wav and big_config[i].audio_source.endswith('.wav'):
                logging.info('Precoding wav file: %s, this may take a while', big_config[i].audio_source)
                big['precoded'] = True
                big['lc3_bytes_per_frame'] = global_config.octets_per_frame

                audio_input = await audio_io.create_audio_input(audio_source, input_format)
                audio_input.rewind = False
                pcm_format = await audio_input.open()

                if pcm_format.channels != 1:
                    logging.error("Only 1 channels PCM configurations are supported")
                    return
                if pcm_format.sample_type == audio_io.PcmFormat.SampleType.INT16:
                    pcm_bit_depth = 16
                elif pcm_format.sample_type == audio_io.PcmFormat.SampleType.FLOAT32:
                    pcm_bit_depth = None
                else:
                    logging.error("Only INT16 and FLOAT32 sample types are supported")
                    return
                encoder = lc3.Encoder(
                    frame_duration_us=global_config.frame_duration_us,
                    sample_rate_hz=global_config.auracast_sampling_rate_hz,
                    num_channels=1,
                    input_sample_rate_hz=pcm_format.sample_rate,
                )
                lc3_frame_samples = encoder.get_frame_samples() # number of the pcm samples per lc3 frame

                lc3_bytes = b''
                async for pcm_frame in audio_input.frames(lc3_frame_samples):
                    lc3_bytes += encoder.encode(
                            pcm_frame, num_bytes=global_config.octets_per_frame, bit_depth=pcm_bit_depth
                        )
                lc3_frames = iter(lc3_bytes)

                # have a look at itertools.islice
                if big_config[i].loop:
                    lc3_frames = itertools.cycle(lc3_frames)
                big['lc3_frames'] = lc3_frames

            # anything else, e.g. realtime stream from device (bumble)
            else:
                if isinstance(audio_source, str) and audio_source.startswith('alsa:'):
                    if input_format == 'auto':
                        raise ValueError('input format details required for alsa input')
                    pcm = audio_io.PcmFormat.from_str(input_format)
                    device_name = audio_source[5:]
                    if device_name.startswith('dante_'):
                        audio_input = PyAlsaAudioInput(device_name, pcm)
                    else:
                        audio_input = AlsaArecordAudioInput(device_name, pcm)
                else:
                    audio_input = await audio_io.create_audio_input(audio_source, input_format)
                # Store early so stop_streaming can close even if open() fails
                big['audio_input'] = audio_input
                # SoundDeviceAudioInput (used for `mic:<device>` captures) has no `.rewind`.
                if hasattr(audio_input, "rewind"):
                    audio_input.rewind = big_config[i].loop

                pcm_format = await audio_input.open()

                num_bis = big.get('num_bis', 1)
                if num_bis == 2 and pcm_format.channels < 2:
                    logging.error("Stereo (num_bis=2) requires at least 2 input channels, got %d", pcm_format.channels)
                    return
                if pcm_format.channels != num_bis:
                    if num_bis == 1:
                        logging.info("Input device provides %d channels – will down-mix to mono for LC3", pcm_format.channels)
                    else:
                        logging.info("Input device provides %d channels – using first %d for stereo", pcm_format.channels, num_bis)

                if pcm_format.sample_type == audio_io.PcmFormat.SampleType.INT16:
                    pcm_bit_depth = 16
                elif pcm_format.sample_type == audio_io.PcmFormat.SampleType.FLOAT32:
                    pcm_bit_depth = None
                else:
                    logging.error("Only INT16 and FLOAT32 sample types are supported")
                    return
                
                # Create one encoder per BIS (mono: 1 encoder, stereo: 2 encoders)
                encoders = [
                    lc3.Encoder(
                        frame_duration_us=global_config.frame_duration_us,
                        sample_rate_hz=global_config.auracast_sampling_rate_hz,
                        num_channels=1,
                        input_sample_rate_hz=pcm_format.sample_rate,
                    )
                    for _ in range(num_bis)
                ]
                
                lc3_frame_samples = encoders[0].get_frame_samples() # number of the pcm samples per lc3 frame
                big['pcm_bit_depth'] = pcm_bit_depth
                big['channels'] = pcm_format.channels
                big['lc3_frame_samples'] = lc3_frame_samples
                big['lc3_bytes_per_frame'] = global_config.octets_per_frame
                big['audio_input'] = audio_input
                big['encoders'] = encoders
                # Keep backward compat
                big['encoder'] = encoders[0]
                big['precoded'] = False

        logging.info("Streaming audio...")
        bigs = self.bigs
        self.is_streaming = True
    
        frame_count = 0

        # One streamer fits all
        while self.is_streaming:
            stream_finished = [False for _ in range(len(bigs))]
            for i, big in enumerate(bigs.values()):
                if big['precoded']:  # everything was already lc3 coded beforehand
                    lc3_frame = bytes(
                        itertools.islice(big['lc3_frames'], big['lc3_bytes_per_frame'])
                    )

                    if lc3_frame == b'':  # Not all streams may stop at the same time
                        stream_finished[i] = True
                        continue
                else:  # code lc3 on the fly with perf counters
                    # Ensure frames generator exists (so we can aclose() on stop)
                    frames_gen = big.get('frames_gen')
                    if frames_gen is None:
                        # For stereo, request frame_samples per channel (interleaved input)
                        frames_gen = big['audio_input'].frames(big['lc3_frame_samples'])
                        big['frames_gen'] = frames_gen

                    # Initialize perf tracking bucket per BIG
                    perf = big.setdefault('_perf', {
                        'n': 0,
                        'samples_sum': 0.0, 'samples_max': 0.0,
                        'enc_sum': 0.0, 'enc_max': 0.0,
                        'write_sum': 0.0, 'write_max': 0.0,
                        'loop_sum': 0.0, 'loop_max': 0.0,
                    })

                    # Total loop duration timer (sample + encode + write)
                    t_loop0 = time.perf_counter()

                    # Measure time to get a sample from the buffer
                    t0 = time.perf_counter()
                    pcm_frame = await anext(frames_gen, None)
                    dt_sample = time.perf_counter() - t0

                    if pcm_frame is None:  # Not all streams may stop at the same time
                        stream_finished[i] = True
                        continue

                    # Compute RMS audio level (normalized 0.0-1.0) for level monitoring
                    pcm_samples = np.frombuffer(pcm_frame, dtype=np.int16).astype(np.float32)
                    rms = np.sqrt(np.mean(pcm_samples ** 2)) / 32768.0 if len(pcm_samples) > 0 else 0.0
                    big['_audio_level_rms'] = float(rms)

                    # Measure LC3 encoding time
                    t1 = time.perf_counter()
                    num_bis = big.get('num_bis', 1)
                    if num_bis == 1:
                        # Mono: single encoder, single queue
                        lc3_frame = big['encoder'].encode(
                            pcm_frame, num_bytes=big['lc3_bytes_per_frame'], bit_depth=big['pcm_bit_depth']
                        )
                        lc3_frames_out = [lc3_frame]
                    else:
                        # Stereo: split interleaved PCM into L/R, encode separately
                        pcm_array = np.frombuffer(pcm_frame, dtype=np.int16)
                        channels_in = big['channels']
                        lc3_frames_out = []
                        for ch_idx, encoder in enumerate(big['encoders']):
                            # Extract channel (interleaved: L,R,L,R,... or L,R,C,... for >2 ch)
                            ch_pcm = pcm_array[ch_idx::channels_in].tobytes()
                            lc3_frame = encoder.encode(
                                ch_pcm, num_bytes=big['lc3_bytes_per_frame'], bit_depth=big['pcm_bit_depth']
                            )
                            lc3_frames_out.append(lc3_frame)
                    dt_enc = time.perf_counter() - t1

                    # Measure write blocking time
                    t2 = time.perf_counter()
                    for q_idx, lc3_frame in enumerate(lc3_frames_out):
                        await big['iso_queues'][q_idx].write(lc3_frame)
                    dt_write = time.perf_counter() - t2

                    # Total loop duration
                    dt_loop = time.perf_counter() - t_loop0

                    # Update stats
                    perf['n'] += 1
                    perf['samples_sum'] += dt_sample
                    perf['enc_sum'] += dt_enc
                    perf['write_sum'] += dt_write
                    perf['loop_sum'] += dt_loop
                    perf['samples_max'] = max(perf['samples_max'], dt_sample)
                    perf['enc_max'] = max(perf['enc_max'], dt_enc)
                    perf['write_max'] = max(perf['write_max'], dt_write)
                    perf['loop_max'] = max(perf['loop_max'], dt_loop)

                    frame_count += 1

                    # Log every 500 frames for this BIG and reset accumulators
                    if perf['n'] >= 500:
                        n = perf['n']
                        logging.info(
                            "Perf(i=%d, last %d): sample mean=%.6fms max=%.6fms | encode mean=%.6fms max=%.6fms | write mean=%.6fms max=%.6fms | loop mean=%.6fms max=%.6fms",
                            i,
                            n,
                            (perf['samples_sum'] / n) * 1e3, perf['samples_max'] * 1e3,
                            (perf['enc_sum'] / n) * 1e3, perf['enc_max'] * 1e3,
                            (perf['write_sum'] / n) * 1e3, perf['write_max'] * 1e3,
                            (perf['loop_sum'] / n) * 1e3, perf['loop_max'] * 1e3,
                        )
                        perf.update({
                            'n': 0,
                            'samples_sum': 0.0, 'samples_max': 0.0,
                            'enc_sum': 0.0, 'enc_max': 0.0,
                            'write_sum': 0.0, 'write_max': 0.0,
                            'loop_sum': 0.0, 'loop_max': 0.0,
                        })

            if all(stream_finished): # Take into account that multiple files have different lengths
                logging.info('All streams finished, stopping streamer')
                self.is_streaming = False
                break


# -----------------------------------------------------------------------------
# Main
# -----------------------------------------------------------------------------

async def broadcast(global_conf: auracast_config.AuracastGlobalConfig, big_conf: List[auracast_config.AuracastBigConfig]):
    """Start a broadcast."""

    if global_conf.transport == 'auto':
        devices = glob.glob('/dev/serial/by-id/*')
        logging.info('Found serial devices: %s', devices)
        for device in devices:
            if 'usb-ZEPHYR_Zephyr_HCI_UART_sample' in device:
                logging.info('Using: %s', device)
                global_conf.transport = f'serial:{device},115200,rtscts'
                break

    # check again if transport is still auto
    if global_conf.transport == 'auto':
        raise AssertionError('No suitable transport found.')

    async with create_device(global_conf) as device:
        if not device.supports_le_periodic_advertising:
            logger.error(color('Periodic advertising not supported', 'red'))
            return

        bigs = await init_broadcast( # the bigs dictionary contains all the global configurations
            device,
            global_conf,
            big_conf
        )
        streamer = Streamer(bigs, global_conf, big_conf)
        streamer.start_streaming()

        await asyncio.wait([streamer.task])


# -----------------------------------------------------------------------------
if __name__ == "__main__":
    import os
    
    logging.basicConfig( #export LOG_LEVEL=DEBUG
        level=os.environ.get('LOG_LEVEL', logging.INFO),
        format='%(module)s.py:%(lineno)d %(levelname)s: %(message)s'
    )
    os.chdir(os.path.dirname(__file__))
        
    # Find ALSA host API
    alsa_hostapi = next(i for i, ha in enumerate(sd.query_hostapis()) 
                            if 'ALSA' in ha['name'])

    search_str='ch1'
    # Use ALSA devices
    from auracast.utils.sounddevice_utils import get_alsa_usb_inputs
    devices = get_alsa_usb_inputs()
    logging.info(f"Searching ALSA devices for first device with string {search_str}...")
    
    audio_dev = None
    for idx, dev in devices:
        logging.info(f"  ALSA device [{idx}]: {dev['name']} ({dev['max_input_channels']} ch)")
        if search_str in dev['name'].lower():
            audio_dev = idx
            logging.info(f"✓ Selected ALSA device {idx}: {dev['name']}")
            break

    
    if audio_dev is None:
        logging.error(f"Audio device {audio_dev} not found in {AUDIO_BACKEND} devices!")
        raise RuntimeError(f"Audio device not found for {AUDIO_BACKEND} backend")
    
    config = auracast_config.AuracastConfigGroup(
        bigs = [
            auracast_config.AuracastBigConfigDeu(),
            #auracast_config.AuracastBigConfigEng(),
            #auracast_config.AuracastBigConfigFra(),
            #auracast_config.AuracastBigConfigEs(),
            #auracast_config.AuracastBigConfigIt(),
        ]
    )

    # TODO: How can we use other iso interval than 10ms ?(medium or low rel) ? - nrf53audio receiver repports I2S tx underrun
    config.qos_config=auracast_config.AuracastQosRobust()

    #config.transport='serial:/dev/serial/by-id/usb-ZEPHYR_Zephyr_HCI_UART_sample_81BD14B8D71B5662-if00,1000000,rtscts' # transport for nrf52 dongle
    #config.transport='serial:/dev/serial/by-id/usb-SEGGER_J-Link_001050076061-if02,1000000,rtscts' # transport for nrf53dk
    #config.transport='serial:/dev/serial/by-id/usb-SEGGER_J-Link_001057705357-if02,1000000,rtscts' # transport for nrf54l15dk
    #config.transport='serial:/dev/serial/by-id/usb-ZEPHYR_Zephyr_HCI_UART_sample_95A087EADB030B24-if00,115200,rtscts' #nrf52dongle hci_uart usb cdc
    #config.transport='usb:2fe3:000b' #nrf52dongle hci_usb # TODO: iso packet over usb not supported
    #config.transport= 'auto'
    config.transport='serial:/dev/ttyAMA3,1000000,rtscts' # transport for raspberry pi

    for big in config.bigs: 
        #big.code = 'abcd'
        #big.code = '78 e5 dc f1 34 ab 42 bf c1 92 ef dd 3a fd 67 ae'
        big.precode_wav = False
        #big.audio_source = big.audio_source.replace('.wav', '_10_16_32.lc3') #lc3 precoded files
        #big.audio_source = read_lc3_file(big.audio_source)  # load files in advance

        # --- Configure device (ALSA backend) ---
        if audio_dev is not None:
            big.audio_source = f'device:{audio_dev}' 
            big.input_format = 'int16le,48000,1'  # int16, 48kHz, mono
            logging.info(f"Configured BIG '{big.name}' with (device:{audio_dev}, 48kHz mono)")
        else:
            logging.warning(f"Shure device not found, BIG '{big.name}' will use default audio_source: {big.audio_source}")
        
        big.name='Broadcast0'
        big.iso_que_len=1
        # --- Network_uncoded mode using NetworkAudioReceiver ---
        #big.audio_source = NetworkAudioReceiverUncoded(port=50007, samplerate=16000, channels=1, chunk_size=1024)
    
    # 16kHz works reliably with 3 streams
    # 24kHz is only working with 2 streams - probably airtime constraint
    # TODO: with more than three broadcasters (16kHz) no advertising (no primary channels is present anymore)
    # TODO: find the bottleneck - probably airtime

    config.auracast_sampling_rate_hz = 24000
    config.octets_per_frame = 60 # 32kbps@16kHz
    #config.immediate_rendering = True
    #config.debug = True
    
    run_async(
        broadcast(
            config,
            config.bigs
        )
    )

    # TODO: possible inputs:
    # wav file locally
    # precoded lc3 file locally
    # realtime audio locally
    # realtime audio network lc3 coded
    # (realtime audio network uncoded)