The native logging utilities allow you to control whisper.cpp’s native log output and subscribe to log messages for debugging and monitoring.
Functions
toggleNativeLog()
Enable or disable native logging from whisper.cpp.
import { toggleNativeLog } from 'whisper.rn'
await toggleNativeLog(enabled)
Parameters
Whether to enable (true) or disable (false) native logging
Returns
Resolves when logging state is updated
Example
// Enable native logging
await toggleNativeLog(true)
// Disable native logging
await toggleNativeLog(false)
Notes
- Logging is automatically initialized with a default callback on first use
- If the first call is
toggleNativeLog(false), it will be skipped to avoid unnecessary initialization
- JSI bindings are automatically installed if needed
addNativeLogListener()
Add a listener to receive native log messages from whisper.cpp.
import { addNativeLogListener } from 'whisper.rn'
const subscription = addNativeLogListener((level, text) => {
console.log(`[${level}] ${text}`)
})
Parameters
listener
(level: string, text: string) => void
required
Callback function that receives log messagesParameters:
level - Log level (e.g., “info”, “warn”, “error”)text - Log message content
Returns
subscription
{ remove: () => void }
required
Subscription object with a remove() method to unsubscribe
Example
const subscription = addNativeLogListener((level, text) => {
if (level === 'error') {
console.error('Native error:', text)
} else {
console.log(`[${level}] ${text}`)
}
})
// Later, remove the listener
subscription.remove()
Notes
- Multiple listeners can be added simultaneously
- Each listener receives all log messages
- Remember to call
remove() to prevent memory leaks
Complete Example
Basic Logging
import { toggleNativeLog, addNativeLogListener } from 'whisper.rn'
// Add listener before enabling logs
const subscription = addNativeLogListener((level, text) => {
console.log(`[WHISPER ${level.toUpperCase()}] ${text}`)
})
// Enable native logging
await toggleNativeLog(true)
// ... use whisper.rn ...
// Clean up
subscription.remove()
await toggleNativeLog(false)
Filtering Log Levels
import { addNativeLogListener } from 'whisper.rn'
const subscription = addNativeLogListener((level, text) => {
switch (level) {
case 'error':
console.error('Whisper error:', text)
break
case 'warn':
console.warn('Whisper warning:', text)
break
case 'info':
console.info('Whisper info:', text)
break
default:
console.log(`Whisper [${level}]:`, text)
}
})
Collecting Logs for Debugging
import { toggleNativeLog, addNativeLogListener } from 'whisper.rn'
const logs: Array<{ level: string; text: string; timestamp: number }> = []
const subscription = addNativeLogListener((level, text) => {
logs.push({
level,
text,
timestamp: Date.now()
})
// Keep only last 100 logs
if (logs.length > 100) {
logs.shift()
}
})
await toggleNativeLog(true)
// ... use whisper.rn ...
// Export logs for debugging
const exportLogs = () => {
return logs.map(log =>
`[${new Date(log.timestamp).toISOString()}] [${log.level}] ${log.text}`
).join('\n')
}
console.log(exportLogs())
Multiple Listeners
import { addNativeLogListener } from 'whisper.rn'
// Listener 1: Console output
const consoleListener = addNativeLogListener((level, text) => {
console.log(`[${level}] ${text}`)
})
// Listener 2: Error tracking
const errorListener = addNativeLogListener((level, text) => {
if (level === 'error') {
// Send to error tracking service
trackError(text)
}
})
// Listener 3: Performance monitoring
const perfListener = addNativeLogListener((level, text) => {
if (text.includes('processing time')) {
// Extract and track performance metrics
trackPerformance(text)
}
})
// Clean up all listeners
const cleanup = () => {
consoleListener.remove()
errorListener.remove()
perfListener.remove()
}
Use Cases
Development Debugging
Enable logging during development to see what’s happening inside whisper.cpp:
if (__DEV__) {
const sub = addNativeLogListener((level, text) => {
console.log(`[WHISPER] ${text}`)
})
await toggleNativeLog(true)
}
Production Error Tracking
Capture only errors in production:
import { addNativeLogListener } from 'whisper.rn'
import * as Sentry from '@sentry/react-native'
const subscription = addNativeLogListener((level, text) => {
if (level === 'error') {
Sentry.captureMessage(`Whisper error: ${text}`, 'error')
}
})
Track transcription performance:
const subscription = addNativeLogListener((level, text) => {
// whisper.cpp logs processing times
const timeMatch = text.match(/(\d+\.\d+)\s*ms/)
if (timeMatch) {
const processingTime = parseFloat(timeMatch[1])
analytics.track('transcription_time', { ms: processingTime })
}
})
- iOS: Full support
- Android: Full support
Constants
libVersion
The current version of the whisper.cpp library.
import { libVersion } from 'whisper.rn'
console.log('whisper.cpp version:', libVersion)
The whisper.cpp library version string
isUseCoreML
Boolean indicating whether Core ML models are being used on iOS.
import { isUseCoreML } from 'whisper.rn'
if (isUseCoreML) {
console.log('Core ML is enabled')
}
true if Core ML models are configured and available, false otherwise
This constant is determined at build time based on the native configuration.
isCoreMLAllowFallback
Boolean indicating whether fallback to CPU is allowed if Core ML model loading fails.
import { isCoreMLAllowFallback } from 'whisper.rn'
console.log('Core ML fallback allowed:', isCoreMLAllowFallback)
true if fallback to CPU is allowed when Core ML fails, false otherwise
This constant is determined at build time based on the native configuration.
See Also