stt

type AlignedTranscript string

type AvailabilityChangedEvent struct {
	STT       STT
	Available bool
}

type Base struct {
	// contains filtered or unexported fields
}

type BaseStream struct {
	// contains filtered or unexported fields
}

type Capabilities struct {
	Streaming         bool
	InterimResults    bool
	AlignedTranscript AlignedTranscript
	Diarization       bool
	Keyterms          bool
	ChatContext       bool
}

type ErrorEvent struct {
	Timestamp   time.Time
	Label       string
	Err         error
	Recoverable bool
}

type FallbackAdapter struct {
	// contains filtered or unexported fields
}

type FallbackOptions struct {
	STTs []STT
	// STTInstances is the agents-js-compatible spelling. Set only one of
	// STTs or STTInstances.
	STTInstances     []STT
	VAD              vad.VAD
	AttemptTimeout   time.Duration
	MaxRetriesPerSTT int
	// MaxRetryPerSTT is the agents-js-compatible spelling. Non-zero values
	// override MaxRetriesPerSTT.
	MaxRetryPerSTT int
	RetryInterval  time.Duration
	StreamCapacity int

	// KeepProvidersOpen leaves the original STT instances caller-owned. By
	// default Close shuts them down after all adapter work has stopped.
	KeepProvidersOpen bool
}

type FallbackStatus struct {
	Available  bool
	Recovering bool
}

type RecognitionUsage struct {
	AudioDuration time.Duration
	InputTokens   int64
	OutputTokens  int64
}

type RecognizeOptions struct {
	ConnectOptions agents.APIConnectOptions
	Language       agents.LanguageCode
}

type STT interface {
	Label() string
	Provider() string
	Model() string
	Capabilities() Capabilities
	Recognize(context.Context, []agents.AudioFrame, RecognizeOptions) (SpeechEvent, error)
	Stream(context.Context, StreamOptions) (SpeechStream, error)
	Close(context.Context) error
	OnMetrics(func(metrics.STT)) func()
	OnError(func(ErrorEvent)) func()
}

type SessionKeytermUpdater interface {
	UpdateSessionKeyterms(keyterms []string) error
}

type SpeechData struct {
	Language        agents.LanguageCode
	Text            string
	StartTime       time.Duration
	EndTime         time.Duration
	Confidence      float64
	Words           []agents.TimedString
	SpeakerID       string
	SourceLanguages []agents.LanguageCode
	SourceTexts     []string
	TargetLanguages []agents.LanguageCode
	TargetTexts     []string
	Metadata        map[string]any
}

type SpeechEvent struct {
	Type             SpeechEventType
	Alternatives     []SpeechData
	RequestID        string
	RecognitionUsage *RecognitionUsage
}

type SpeechEventType uint8

type SpeechStream interface {
	stream.Reader[SpeechEvent]
	Push(context.Context, agents.AudioFrame) error
	Flush(context.Context) error
	EndInput() error
	Close() error
}

type StreamAdapter struct {
	// contains filtered or unexported fields
}

type StreamAdapterOptions struct {
	StreamCapacity int
}

type StreamAdapterWrapper struct {
	*BaseStream
	// contains filtered or unexported fields
}

type StreamInput struct {
	Frame *agents.AudioFrame
	Flush bool
}

type StreamOptions struct {
	ConnectOptions agents.APIConnectOptions
	Language       agents.LanguageCode
}