ttsscript

type BatchConfig struct {
	// OutputDir is the directory for output files.
	OutputDir string

	// FilePrefix is added before each filename.
	FilePrefix string

	// FileSuffix is added after each filename (before extension).
	FileSuffix string

	// IncludeLanguageInFilename adds language code to filename.
	IncludeLanguageInFilename bool
}

type CompiledSegment struct {
	// SlideIndex is the 0-based slide index.
	SlideIndex int

	// SegmentIndex is the 0-based segment index within the slide.
	// For title segments, this is -1.
	SegmentIndex int

	// SlideTitle is the slide title (if any).
	SlideTitle string

	// IsTitleSegment indicates this segment was generated from a slide title.
	IsTitleSegment bool

	// IsSectionHeader indicates this segment belongs to a section header slide.
	IsSectionHeader bool

	// Text is the processed text with pronunciations applied.
	Text string

	// OriginalText is the text before pronunciation substitutions.
	OriginalText string

	// VoiceID is the voice to use for this segment.
	VoiceID string

	// Language is the language code.
	Language string

	// PauseBeforeMs is the pause before in milliseconds.
	PauseBeforeMs int

	// PauseAfterMs is the pause after in milliseconds.
	PauseAfterMs int

	// Emphasis is the emphasis level.
	Emphasis string

	// Rate is the speaking rate.
	Rate string

	// Pitch is the pitch adjustment.
	Pitch string
}

type Compiler struct {
	// AdditionalPronunciations are extra pronunciations to apply.
	AdditionalPronunciations map[string]map[string]string

	// DefaultPauseAfterSlide is the pause after each slide if not specified.
	DefaultPauseAfterSlide string

	// DefaultPauseAfterSegment is the pause after each segment if not specified.
	DefaultPauseAfterSegment string
}

type ElevenLabsFormatter struct {
	// UsePauseMarkers includes [pause:Xms] markers in text output.
	// When false, pauses are tracked separately for post-processing.
	UsePauseMarkers bool

	// PauseMarkerFormat is the format for pause markers (default: "[pause:%s]").
	PauseMarkerFormat string
}

type ElevenLabsSegment struct {
	// Text is the text to generate speech for.
	Text string

	// VoiceID is the ElevenLabs voice ID.
	VoiceID string

	// SlideIndex is the source slide index.
	SlideIndex int

	// SegmentIndex is the source segment index (-1 for title segments).
	SegmentIndex int

	// SlideTitle is the slide title for reference.
	SlideTitle string

	// IsTitleSegment indicates this segment was generated from a slide title.
	IsTitleSegment bool

	// IsSectionHeader indicates this segment belongs to a section header slide.
	IsSectionHeader bool

	// PauseBeforeMs is silence to add before this segment.
	PauseBeforeMs int

	// PauseAfterMs is silence to add after this segment.
	PauseAfterMs int

	// SuggestedFilename is a suggested output filename.
	SuggestedFilename string
}

type ManifestEntry struct {
	SlideIndex      int    `json:"slide_index"`
	SegmentIndex    int    `json:"segment_index"`
	SlideTitle      string `json:"slide_title,omitempty"`
	IsTitleSegment  bool   `json:"is_title_segment,omitempty"`
	IsSectionHeader bool   `json:"is_section_header,omitempty"`
	Text            string `json:"text"`
	VoiceID         string `json:"voice_id"`
	Language        string `json:"language"`
	OutputFile      string `json:"output_file"`
	PauseBeforeMs   int    `json:"pause_before_ms,omitempty"`
	PauseAfterMs    int    `json:"pause_after_ms,omitempty"`
}

type SSMLFormatter struct {
	// Version is the SSML version (default: "1.1").
	Version string

	// IncludeComments includes slide title comments in output.
	IncludeComments bool

	// IndentSpaces is the number of spaces for indentation.
	IndentSpaces int
}

type Script struct {
	// Title is the script title.
	Title string `json:"title,omitempty"`

	// Description is an optional description.
	Description string `json:"description,omitempty"`

	// DefaultLanguage is the primary language code (e.g., "en-US").
	DefaultLanguage string `json:"default_language,omitempty"`

	// DefaultVoices maps language codes to default voice IDs.
	DefaultVoices map[string]string `json:"default_voices,omitempty"`

	// Pronunciations maps terms to their pronunciation by language.
	// Example: {"ADK": {"en": "A D K", "es": "A D K"}}
	Pronunciations map[string]map[string]string `json:"pronunciations,omitempty"`

	// Slides contains the ordered list of slides/sections.
	Slides []Slide `json:"slides"`
}

type Segment struct {
	// Text contains the text content by language code.
	// Example: {"en": "Hello world", "es": "Hola mundo"}
	Text map[string]string `json:"text"`

	// Voice overrides the default voice for this segment by language.
	// Example: {"en": "voice-id-1", "es": "voice-id-2"}
	Voice map[string]string `json:"voice,omitempty"`

	// PauseBefore is the pause duration before this segment (e.g., "500ms", "1s").
	PauseBefore string `json:"pause_before,omitempty"`

	// PauseAfter is the pause duration after this segment (e.g., "500ms", "1s").
	PauseAfter string `json:"pause_after,omitempty"`

	// Emphasis indicates the emphasis level ("strong", "moderate", "reduced").
	Emphasis string `json:"emphasis,omitempty"`

	// Rate is the speaking rate ("slow", "medium", "fast", or percentage like "80%").
	Rate string `json:"rate,omitempty"`

	// Pitch adjusts the pitch ("low", "medium", "high", or percentage like "+10%").
	Pitch string `json:"pitch,omitempty"`

	// Pronunciations are segment-specific pronunciation overrides.
	Pronunciations map[string]map[string]string `json:"pronunciations,omitempty"`
}

type Slide struct {
	// Title is the slide title (optional).
	Title string `json:"title,omitempty"`

	// Notes are speaker notes or comments (not rendered to audio).
	Notes string `json:"notes,omitempty"`

	// IsSectionHeader marks this slide as the start of a new section.
	// Section headers can have their titles spoken and use longer transition pauses.
	IsSectionHeader bool `json:"is_section_header,omitempty"`

	// SpeakTitle causes the slide title to be spoken before the segments.
	// If true, the title is converted to a segment. Defaults to true for section headers.
	SpeakTitle *bool `json:"speak_title,omitempty"`

	// TitleVoice overrides the voice used for speaking the title, by language.
	// If not set, uses the segment voice or default voice.
	TitleVoice map[string]string `json:"title_voice,omitempty"`

	// TitlePauseAfter is the pause after the spoken title (e.g., "500ms").
	// Defaults to "500ms" for section headers, "300ms" for regular slides.
	TitlePauseAfter string `json:"title_pause_after,omitempty"`

	// Segments are the audio segments for this slide.
	Segments []Segment `json:"segments"`
}

type TTSRequest struct {
	VoiceID  string
	Text     string
	ModelID  string
	Segment  ElevenLabsSegment
	Language string
}