@siteed/expo-audio-stream 1.1.2 → 1.1.3

package/src/useAudioRecorder.tsx

@@ -1,420 +0,0 @@
-// src/useAudioRecorder.ts
-import { Platform, Subscription } from 'expo-modules-core'
-import { useCallback, useEffect, useReducer, useRef } from 'react'
-
-import { AudioAnalysis } from './AudioAnalysis/AudioAnalysis.types'
-import {
-    AudioDataEvent,
-    AudioRecording,
-    AudioStreamStatus,
-    RecordingConfig,
-    StartRecordingResult,
-} from './ExpoAudioStream.types'
-import ExpoAudioStreamModule from './ExpoAudioStreamModule'
-import {
-    addAudioAnalysisListener,
-    addAudioEventListener,
-    AudioEventPayload,
-} from './events'
-import { getLogger } from './logger'
-
-const TAG = 'useAudioRecorder'
-const logger = getLogger(TAG)
-export interface UseAudioRecorderProps {
-    debug?: boolean
-    audioWorkletUrl?: string
-    featuresExtratorUrl?: string
-}
-
-export interface UseAudioRecorderState {
-    startRecording: (_: RecordingConfig) => Promise<StartRecordingResult>
-    stopRecording: () => Promise<AudioRecording>
-    pauseRecording: () => void
-    resumeRecording: () => void
-    isRecording: boolean
-    isPaused: boolean
-    durationMs: number // Duration of the recording
-    size: number // Size in bytes of the recorded audio
-    analysisData?: AudioAnalysis
-}
-
-interface RecorderReducerState {
-    isRecording: boolean
-    isPaused: boolean
-    durationMs: number
-    size: number
-    analysisData?: AudioAnalysis
-}
-
-type RecorderAction =
-    | { type: 'START' | 'STOP' | 'PAUSE' | 'RESUME' }
-    | { type: 'UPDATE_STATUS'; payload: { durationMs: number; size: number } }
-    | { type: 'UPDATE_ANALYSIS'; payload: AudioAnalysis }
-
-const defaultAnalysis: AudioAnalysis = {
-    pointsPerSecond: 10,
-    bitDepth: 32,
-    numberOfChannels: 1,
-    durationMs: 0,
-    sampleRate: 44100,
-    samples: 0,
-    dataPoints: [],
-    amplitudeAlgorithm: 'rms',
-    speakerChanges: [],
-    amplitudeRange: {
-        min: Number.POSITIVE_INFINITY,
-        max: Number.NEGATIVE_INFINITY,
-    },
-}
-
-function audioRecorderReducer(
-    state: RecorderReducerState,
-    action: RecorderAction
-): RecorderReducerState {
-    switch (action.type) {
-        case 'START':
-            return {
-                ...state,
-                isRecording: true,
-                isPaused: false,
-                durationMs: 0,
-                size: 0,
-                analysisData: defaultAnalysis, // Reset analysis data
-            }
-        case 'STOP':
-            return { ...state, isRecording: false, isPaused: false }
-        case 'PAUSE':
-            return { ...state, isPaused: true, isRecording: false }
-        case 'RESUME':
-            return { ...state, isPaused: false, isRecording: true }
-        case 'UPDATE_STATUS':
-            return {
-                ...state,
-                durationMs: action.payload.durationMs,
-                size: action.payload.size,
-            }
-        case 'UPDATE_ANALYSIS':
-            return {
-                ...state,
-                analysisData: action.payload,
-            }
-        default:
-            return state
-    }
-}
-
-interface HandleAudioAnalysisProps {
-    analysis: AudioAnalysis
-    visualizationDuration: number
-}
-
-export function useAudioRecorder({
-    debug = false,
-    audioWorkletUrl,
-    featuresExtratorUrl,
-}: UseAudioRecorderProps = {}): UseAudioRecorderState {
-    const [state, dispatch] = useReducer(audioRecorderReducer, {
-        isRecording: false,
-        isPaused: false,
-        durationMs: 0,
-        size: 0,
-        analysisData: undefined,
-    })
-
-    const analysisListenerRef = useRef<Subscription | null>(null)
-    const analysisRef = useRef<AudioAnalysis>({ ...defaultAnalysis })
-
-    // Instantiate the module for web with URLs
-    const ExpoAudioStream =
-        Platform.OS === 'web'
-            ? ExpoAudioStreamModule({ audioWorkletUrl, featuresExtratorUrl })
-            : ExpoAudioStreamModule
-
-    const onAudioStreamRef = useRef<
-        ((_: AudioDataEvent) => Promise<void>) | null
-    >(null)
-
-    const handleAudioAnalysis = useCallback(
-        async ({
-            analysis,
-            visualizationDuration,
-        }: HandleAudioAnalysisProps) => {
-            const savedAnalysisData = analysisRef.current || {
-                ...defaultAnalysis,
-            }
-
-            const maxDuration = visualizationDuration
-
-            logger.debug(
-                `[handleAudioAnalysis] Received audio analysis: maxDuration=${maxDuration} analysis.dataPoints=${analysis.dataPoints.length} analysisData.dataPoints=${savedAnalysisData.dataPoints.length}`,
-                analysis
-            )
-
-            // Combine data points
-            const combinedDataPoints = [
-                ...savedAnalysisData.dataPoints,
-                ...analysis.dataPoints,
-            ]
-
-            // Calculate the new duration
-            const pointsPerSecond =
-                analysis.pointsPerSecond || savedAnalysisData.pointsPerSecond
-            const maxDataPoints =
-                (pointsPerSecond * visualizationDuration) / 1000
-
-            logger.debug(
-                `[handleAudioAnalysis] Combined data points before trimming: pointsPerSecond=${pointsPerSecond} visualizationDuration=${visualizationDuration} combinedDataPointsLength=${combinedDataPoints.length} vs maxDataPoints=${maxDataPoints}`
-            )
-
-            // Trim data points to keep within the maximum number of data points
-            if (combinedDataPoints.length > maxDataPoints) {
-                combinedDataPoints.splice(
-                    0,
-                    combinedDataPoints.length - maxDataPoints
-                )
-            }
-
-            savedAnalysisData.dataPoints = combinedDataPoints
-            savedAnalysisData.bitDepth =
-                analysis.bitDepth || savedAnalysisData.bitDepth
-            savedAnalysisData.durationMs =
-                combinedDataPoints.length * (1000 / pointsPerSecond)
-
-            // Update amplitude range
-            const newMin = Math.min(
-                savedAnalysisData.amplitudeRange.min,
-                analysis.amplitudeRange.min
-            )
-            const newMax = Math.max(
-                savedAnalysisData.amplitudeRange.max,
-                analysis.amplitudeRange.max
-            )
-
-            savedAnalysisData.amplitudeRange = {
-                min: newMin,
-                max: newMax,
-            }
-
-            logger.debug(
-                `[handleAudioAnalysis] Updated analysis data: durationMs=${savedAnalysisData.durationMs}`,
-                savedAnalysisData
-            )
-
-            // Update the ref
-            analysisRef.current = savedAnalysisData
-
-            // Dispatch the updated analysis data to state to trigger re-render
-            // need to use spread operator otherwise it doesnt trigger update.
-            dispatch({
-                type: 'UPDATE_ANALYSIS',
-                payload: { ...savedAnalysisData },
-            })
-        },
-        [dispatch]
-    )
-
-    const handleAudioEvent = useCallback(
-        async (eventData: AudioEventPayload) => {
-            const {
-                fileUri,
-                deltaSize,
-                totalSize,
-                lastEmittedSize,
-                position,
-                streamUuid,
-                encoded,
-                mimeType,
-                buffer,
-            } = eventData
-            logger.debug(`[handleAudioEvent] Received audio event:`, {
-                fileUri,
-                deltaSize,
-                totalSize,
-                position,
-                mimeType,
-                lastEmittedSize,
-                streamUuid,
-                encodedLength: encoded?.length,
-            })
-            if (deltaSize === 0) {
-                // Ignore packet with no data
-                return
-            }
-            try {
-                // Coming from native ( ios / android ) otherwise buffer is set
-                if (Platform.OS !== 'web') {
-                    // Read the audio file as a base64 string for comparison
-                    if (!encoded) {
-                        console.error(`${TAG} Encoded audio data is missing`)
-                        throw new Error('Encoded audio data is missing')
-                    }
-                    onAudioStreamRef.current?.({
-                        data: encoded,
-                        position,
-                        fileUri,
-                        eventDataSize: deltaSize,
-                        totalSize,
-                    })
-                } else if (buffer) {
-                    // Coming from web
-                    const webEvent: AudioDataEvent = {
-                        data: buffer,
-                        position,
-                        fileUri,
-                        eventDataSize: deltaSize,
-                        totalSize,
-                    }
-                    onAudioStreamRef.current?.(webEvent)
-                    logger.debug(
-                        `[handleAudioEvent] Audio data sent to onAudioStream`,
-                        webEvent
-                    )
-                }
-            } catch (error) {
-                console.error(`${TAG} Error processing audio event:`, error)
-            }
-        },
-        []
-    )
-
-    const checkStatus = useCallback(async () => {
-        try {
-            if (!state.isRecording) {
-                logger.debug(`Not recording, exiting status check.`)
-                return
-            }
-
-            const status: AudioStreamStatus = ExpoAudioStream.status()
-            if (debug) {
-                logger.debug(`Status:`, status)
-            }
-
-            dispatch({
-                type: 'UPDATE_STATUS',
-                payload: { durationMs: status.durationMs, size: status.size },
-            })
-        } catch (error) {
-            console.error(`${TAG} Error getting status:`, error)
-        }
-    }, [state.isRecording])
-
-    const startRecording = useCallback(
-        async (recordingOptions: RecordingConfig) => {
-            logger.debug(`start recoding`, recordingOptions)
-
-            analysisRef.current = { ...defaultAnalysis } // Reset analysis data
-
-            const { onAudioStream, ...options } = recordingOptions
-            const { enableProcessing } = options
-
-            const maxRecentDataDuration = 10000 // TODO compute maxRecentDataDuration based on screen dimensions
-            if (typeof onAudioStream === 'function') {
-                onAudioStreamRef.current = onAudioStream
-            } else {
-                console.warn(
-                    `${TAG} onAudioStream is not a function`,
-                    onAudioStream
-                )
-                onAudioStreamRef.current = null
-            }
-            const startResult: StartRecordingResult =
-                await ExpoAudioStream.startRecording(options)
-            dispatch({ type: 'START' })
-
-            if (enableProcessing) {
-                logger.debug(`Enabling audio analysis listener`)
-                const listener = addAudioAnalysisListener(
-                    async (analysisData) => {
-                        try {
-                            await handleAudioAnalysis({
-                                analysis: analysisData,
-                                visualizationDuration: maxRecentDataDuration,
-                            })
-                        } catch (error) {
-                            console.warn(
-                                `${TAG} Error processing audio analysis:`,
-                                error
-                            )
-                        }
-                    }
-                )
-
-                analysisListenerRef.current = listener
-            }
-
-            return startResult
-        },
-        [handleAudioAnalysis, dispatch]
-    )
-
-    const stopRecording = useCallback(async () => {
-        logger.debug(`stoping recording`)
-
-        const stopResult: AudioRecording = await ExpoAudioStream.stopRecording()
-        stopResult.analysisData = analysisRef.current
-
-        if (analysisListenerRef.current) {
-            analysisListenerRef.current.remove()
-            analysisListenerRef.current = null
-        }
-        onAudioStreamRef.current = null
-        logger.debug(`recording stopped`, stopResult)
-        dispatch({ type: 'STOP' })
-        return stopResult
-    }, [dispatch])
-
-    const pauseRecording = useCallback(async () => {
-        logger.debug(`pause recording`)
-        const pauseResult = await ExpoAudioStream.pauseRecording()
-        dispatch({ type: 'PAUSE' })
-        return pauseResult
-    }, [dispatch])
-
-    const resumeRecording = useCallback(async () => {
-        logger.debug(`resume recording`)
-        const resumeResult = await ExpoAudioStream.resumeRecording()
-        dispatch({ type: 'RESUME' })
-        return resumeResult
-    }, [dispatch])
-
-    useEffect(() => {
-        let interval: ReturnType<typeof setTimeout>
-        if (state.isRecording) {
-            interval = setInterval(checkStatus, 1000)
-        }
-        return () => {
-            if (interval) {
-                clearInterval(interval)
-            }
-        }
-    }, [checkStatus, state.isRecording])
-
-    useEffect(() => {
-        logger.debug(`Registering audio event listener`)
-        const subscribeAudio = addAudioEventListener(handleAudioEvent)
-
-        logger.debug(
-            `Subscribed to audio event listener and analysis listener`,
-            {
-                subscribeAudio,
-            }
-        )
-
-        return () => {
-            logger.debug(`Removing audio event listener`)
-            subscribeAudio.remove()
-        }
-    }, [handleAudioEvent, handleAudioAnalysis])
-
-    return {
-        startRecording,
-        stopRecording,
-        pauseRecording,
-        resumeRecording,
-        isPaused: state.isPaused,
-        isRecording: state.isRecording,
-        durationMs: state.durationMs,
-        size: state.size,
-        analysisData: state.analysisData,
-    }
-}

package/src/utils/convertPCMToFloat32.ts

@@ -1,75 +0,0 @@
-import { getWavFileInfo, WavFileInfo } from './getWavFileInfo'
-import { getLogger } from '../logger'
-
-export const WAV_HEADER_SIZE = 44
-
-const logger = getLogger('convertPCMToFloat32')
-
-const convertSample = (
-    dataView: DataView,
-    offset: number,
-    bitDepth: number
-): number => {
-    switch (bitDepth) {
-        case 8:
-            return (dataView.getUint8(offset) - 128) / 128
-        case 16:
-            return dataView.getInt16(offset, true) / 32768
-        case 24:
-            return (
-                ((dataView.getUint8(offset) |
-                    (dataView.getUint8(offset + 1) << 8) |
-                    (dataView.getUint8(offset + 2) << 16)) /
-                    8388608) *
-                    2 -
-                1
-            )
-        case 32:
-            return dataView.getFloat32(offset, true)
-        default:
-            throw new Error(`Unsupported bit depth: ${bitDepth}`)
-    }
-}
-
-export const convertPCMToFloat32 = async ({
-    bitDepth,
-    buffer,
-    skipWavHeader = false,
-}: {
-    buffer: ArrayBuffer
-    bitDepth: number
-    skipWavHeader?: boolean
-}): Promise<{ pcmValues: Float32Array; min: number; max: number }> => {
-    try {
-        logger.debug(
-            `Converting PCM to Float32: bitDepth: ${bitDepth}, buffer.byteLength: ${buffer.byteLength}`
-        )
-        const dataView = new DataView(buffer)
-        let headerOffset = 0
-        if (skipWavHeader) {
-            const wavFileInfo: WavFileInfo = await getWavFileInfo(buffer)
-            headerOffset = wavFileInfo.dataChunkOffset
-        }
-
-        const dataLength = buffer.byteLength - headerOffset
-        const sampleLength = Math.floor(dataLength / (bitDepth / 8))
-        const float32Array = new Float32Array(sampleLength)
-        let min = Infinity
-        let max = -Infinity
-
-        for (let i = 0; i < sampleLength; i++) {
-            const offset = headerOffset + i * (bitDepth / 8)
-            const value = convertSample(dataView, offset, bitDepth)
-
-            if (value < min) min = value
-            if (value > max) max = value
-
-            float32Array[i] = value
-        }
-
-        return { pcmValues: float32Array, min, max }
-    } catch (error: unknown) {
-        logger.error(`Error converting PCM to Float32`, error)
-        return { pcmValues: new Float32Array(), min: 0, max: 0 }
-    }
-}

package/src/utils/encodingToBitDepth.ts

@@ -1,18 +0,0 @@
-import { BitDepth, EncodingType } from '../ExpoAudioStream.types'
-
-export const encodingToBitDepth = ({
-    encoding,
-}: {
-    encoding: EncodingType
-}): BitDepth => {
-    switch (encoding) {
-        case 'pcm_32bit':
-            return 32
-        case 'pcm_16bit':
-            return 16
-        case 'pcm_8bit':
-            return 8
-        default:
-            throw new Error(`Unsupported encoding type: ${encoding}`)
-    }
-}

package/src/utils/getWavFileInfo.ts

@@ -1,132 +0,0 @@
-// packages/expo-audio-stream/src/utils/getWavFileInfo.ts
-
-import { BitDepth, SampleRate } from '../ExpoAudioStream.types'
-import {
-    DATA_CHUNK_ID,
-    DEFAULT_BIT_DEPTH,
-    DEFAULT_SAMPLE_RATE,
-    FMT_CHUNK_ID,
-    INFO_CHUNK_ID,
-    RIFF_HEADER,
-    WAVE_HEADER,
-} from '../constants'
-
-// Audio format descriptions
-const AUDIO_FORMATS: { [key: number]: string } = {
-    1: 'PCM',
-    3: 'IEEE float',
-    6: '8-bit ITU-T G.711 A-law',
-    7: '8-bit ITU-T G.711 µ-law',
-    65534: 'WAVE_FORMAT_EXTENSIBLE',
-}
-
-/**
- * Interface representing the metadata of a WAV file.
- */
-export interface WavFileInfo {
-    sampleRate: SampleRate
-    numChannels: number
-    bitDepth: BitDepth
-    size: number // in bytes
-    durationMs: number // in ms
-    audioFormatDescription: string // Description of the audio format
-    byteRate: number // Average bytes per second
-    blockAlign: number // Number of bytes for one sample including all channels
-    creationDateTime?: string // Optional creation date and time
-    comments?: string // Optional comments or tags
-    compressionType?: string // Optional compression type
-    dataChunkOffset: number // Position of the first data chunk
-}
-
-/**
- * Extracts metadata from a WAV buffer.
- *
- * @param arrayBuffer - The array buffer containing the WAV data.
- * @returns A promise that resolves to the extracted metadata.
- */
-export const getWavFileInfo = async (
-    arrayBuffer: ArrayBuffer
-): Promise<WavFileInfo> => {
-    const view = new DataView(arrayBuffer)
-
-    // Check if the file is a valid RIFF/WAVE file
-    const riffHeader = view.getUint32(0, false)
-    const waveHeader = view.getUint32(8, false)
-    if (riffHeader !== RIFF_HEADER || waveHeader !== WAVE_HEADER) {
-        throw new Error('Invalid WAV file')
-    }
-
-    // Initialize variables for the metadata
-    let fmtChunkOffset = 12
-    let sampleRate: SampleRate = DEFAULT_SAMPLE_RATE
-    let numChannels = 0
-    let bitDepth: BitDepth = DEFAULT_BIT_DEPTH
-    let dataChunkSize = 0
-    let audioFormat = 0
-    let byteRate = 0
-    let blockAlign = 0
-    let creationDateTime = ''
-    let comments = ''
-    let dataChunkOffset = 0
-
-    // Parse chunks to find the "fmt " and "data" chunks
-    while (fmtChunkOffset < view.byteLength) {
-        const chunkId = view.getUint32(fmtChunkOffset, false)
-        const chunkSize = view.getUint32(fmtChunkOffset + 4, true)
-        if (chunkId === FMT_CHUNK_ID) {
-            // "fmt "
-            audioFormat = view.getUint16(fmtChunkOffset + 8, true)
-            if (!AUDIO_FORMATS[audioFormat]) {
-                throw new Error('Unsupported WAV file format')
-            }
-            numChannels = view.getUint16(fmtChunkOffset + 10, true)
-            sampleRate = view.getUint32(fmtChunkOffset + 12, true) as SampleRate
-            byteRate = view.getUint32(fmtChunkOffset + 16, true)
-            blockAlign = view.getUint16(fmtChunkOffset + 20, true)
-            bitDepth = view.getUint16(fmtChunkOffset + 22, true) as BitDepth
-        } else if (chunkId === DATA_CHUNK_ID) {
-            // "data"
-            dataChunkSize = chunkSize
-            dataChunkOffset = fmtChunkOffset + 8 // Position after chunk header
-            break
-        } else if (chunkId === INFO_CHUNK_ID) {
-            // "INFO"
-            // Read INFO chunk (assuming it contains a text-based creation date/time and comments)
-            const infoStart = fmtChunkOffset + 8
-            const infoText = new TextDecoder().decode(
-                new Uint8Array(
-                    arrayBuffer.slice(infoStart, infoStart + chunkSize)
-                )
-            )
-            const infoParts = infoText.split('\0')
-            creationDateTime = infoParts[0]
-            comments = infoParts[1]
-        }
-        fmtChunkOffset += 8 + chunkSize
-    }
-
-    if (!sampleRate || !numChannels || !bitDepth || !dataChunkSize) {
-        throw new Error('Incomplete WAV file information')
-    }
-
-    // Calculate duration
-    const bytesPerSample = bitDepth / 8
-    const numSamples = dataChunkSize / (numChannels * bytesPerSample)
-    const durationMs = (numSamples / sampleRate) * 1000
-
-    return {
-        sampleRate,
-        numChannels,
-        bitDepth,
-        size: arrayBuffer.byteLength,
-        durationMs,
-        audioFormatDescription: AUDIO_FORMATS[audioFormat],
-        byteRate,
-        blockAlign,
-        creationDateTime: creationDateTime || undefined,
-        comments: comments || undefined,
-        compressionType:
-            audioFormat === 1 ? 'None' : AUDIO_FORMATS[audioFormat],
-        dataChunkOffset,
-    }
-}

package/src/utils/writeWavHeader.ts

@@ -1,61 +0,0 @@
-// packages/expo-audio-stream/src/utils/writeWavHeader.ts
-interface WavHeaderOptions {
-    buffer: ArrayBuffer
-    sampleRate: number
-    numChannels: number
-    bitDepth: number
-}
-
-export const writeWavHeader = ({
-    buffer,
-    sampleRate,
-    numChannels,
-    bitDepth,
-}: WavHeaderOptions): ArrayBuffer => {
-    const bytesPerSample = bitDepth / 8
-    const numSamples = buffer.byteLength / (numChannels * bytesPerSample)
-    const view = new DataView(buffer)
-    const blockAlign = numChannels * bytesPerSample
-    const byteRate = sampleRate * blockAlign
-
-    // Function to write a string to the DataView
-    const writeString = (view: DataView, offset: number, string: string) => {
-        for (let i = 0; i < string.length; i++) {
-            view.setUint8(offset + i, string.charCodeAt(i))
-        }
-    }
-
-    // Check if the buffer already has a WAV header by looking for "RIFF" at the start
-    const existingHeader = view.getUint32(0, false) === 0x52494646 // "RIFF" in ASCII
-
-    if (!existingHeader) {
-        // Ensure the buffer is large enough for the WAV header
-        if (buffer.byteLength < 44) {
-            throw new Error('Buffer is too small to contain a valid WAV header')
-        }
-
-        // Write the WAV header
-        writeString(view, 0, 'RIFF') // ChunkID
-        view.setUint32(4, 36 + numSamples * blockAlign, true) // ChunkSize
-        writeString(view, 8, 'WAVE') // Format
-        writeString(view, 12, 'fmt ') // Subchunk1ID
-        view.setUint32(16, 16, true) // Subchunk1Size (16 for PCM)
-        view.setUint16(20, bitDepth === 32 ? 3 : 1, true) // AudioFormat (3 for float, 1 for PCM)
-        view.setUint16(22, numChannels, true) // NumChannels
-        view.setUint32(24, sampleRate, true) // SampleRate
-        view.setUint32(28, byteRate, true) // ByteRate
-        view.setUint16(32, blockAlign, true) // BlockAlign
-        view.setUint16(34, bitDepth, true) // BitsPerSample
-        writeString(view, 36, 'data') // Subchunk2ID
-        view.setUint32(40, numSamples * blockAlign, true) // Subchunk2Size
-    } else {
-        // Update the existing WAV header if necessary
-        view.setUint32(4, 36 + numSamples * blockAlign, true) // Update ChunkSize
-        view.setUint32(24, sampleRate, true) // Update SampleRate
-        view.setUint32(28, byteRate, true) // Update ByteRate
-        view.setUint16(32, blockAlign, true) // Update BlockAlign
-        view.setUint32(40, numSamples * blockAlign, true) // Update Subchunk2Size
-    }
-
-    return buffer
-}