orchestrator

Documentation

Index

• type Config
• type Orchestrator
  • func NewOrchestrator(config Config) *Orchestrator
  • func (o *Orchestrator) Process(ctx context.Context) error
• type SegmentConfig
• type SegmentOrchestrator
  • func NewSegmentOrchestrator(config SegmentConfig, ttsGen *tts.SegmentTTSGenerator) *SegmentOrchestrator
  • func (o *SegmentOrchestrator) Process(ctx context.Context) error

Types

type Config

type Config struct {
	InputFile           string
	OutputFile          string
	WorkDir             string
	ElevenLabsAPIKey    string
	VoiceID             string
	Width               int
	Height              int
	FrameRate           int
	OutputIndividualDir string    // Directory for individual slide videos (Udemy)
	TransitionDuration  float64   // Duration of transitions between slides in seconds
	ScreenDevice        string    // Screen capture device (macOS, auto-detected if empty)
	AudioManifest       string    // Path to audio manifest file (from 'marp2video tts')
	ProgressWriter      io.Writer // Writer for progress output (nil to disable)
}

Config holds orchestrator configuration

type Orchestrator

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

Orchestrator coordinates the entire video generation process

func NewOrchestrator

func NewOrchestrator(config Config) *Orchestrator

NewOrchestrator creates a new orchestrator

func (*Orchestrator) Process

func (o *Orchestrator) Process(ctx context.Context) error

Process orchestrates the entire conversion process

type SegmentConfig

type SegmentConfig struct {
	// Source is the content source (transcript, config, etc.)
	Source source.ContentSource

	// OutputFile is the final video output path
	OutputFile string

	// WorkDir is where temporary files are stored
	WorkDir string

	// AudioOutputDir is where to save audio tracks (if set, audio is preserved)
	// Structure: {AudioOutputDir}/{language}/combined.mp3
	AudioOutputDir string

	// Languages is the list of languages to generate (uses default if empty)
	Languages []string

	// Width and Height are video dimensions
	Width  int
	Height int

	// FrameRate is the video frame rate
	FrameRate int

	// TransitionDuration is crossfade duration between segments (seconds)
	TransitionDuration float64

	// Headless runs browser in headless mode
	Headless bool

	// ProgressWriter receives progress updates
	ProgressWriter io.Writer

	// Subtitles enables simple subtitle generation from voiceover timing
	Subtitles bool

	// SubtitlesSTT enables word-level subtitles using speech-to-text
	SubtitlesSTT bool

	// SubtitlesBurn burns subtitles into video (permanent)
	SubtitlesBurn bool

	// SubtitleFormat is "srt" or "vtt" (default: srt)
	SubtitleFormat string

	// DeepgramAPIKey for STT subtitles
	DeepgramAPIKey string

	// NoAudio generates video without audio (TTS still used for timing/subtitles)
	NoAudio bool

	// Parallel is the number of concurrent segment recordings (default 1 = sequential)
	Parallel int

	// FastEncoding uses hardware acceleration for video encoding when available
	FastEncoding bool

	// SegmentLimit limits processing to first N segments (0 = no limit)
	SegmentLimit int

	// StepLimit limits browser segments to first N steps (0 = no limit)
	StepLimit int
}

SegmentConfig holds configuration for the segment-based orchestrator.

type SegmentOrchestrator

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

SegmentOrchestrator coordinates video generation using the segment abstraction. It works with any ContentSource (transcript JSON, config YAML) and any segment type (slides, browser demos).

func NewSegmentOrchestrator

func NewSegmentOrchestrator(config SegmentConfig, ttsGen *tts.SegmentTTSGenerator) *SegmentOrchestrator

NewSegmentOrchestrator creates a new segment-based orchestrator.

func (*SegmentOrchestrator) Process

func (o *SegmentOrchestrator) Process(ctx context.Context) error

Process runs the complete video generation pipeline.