the concrete segmenter's options type.
OptionalContextContext user, required by segmenters that call MJ services (e.g. the LLM
boundary pass in SemanticTextSegmenter). Always pass it in server-side code.
OptionalCuesTimed transcript cues, when available (ASR output or MJ realtime capture).
OptionalDurationTotal duration of the source asset in milliseconds, for AV content.
OptionalMediaMedia pointer for the source asset, for image/audio/video content.
OptionalMimeMime type of the source asset — lets a segmenter pick a structure parser.
OptionalOptionsStrategy-specific options.
OptionalPagesPages of a paginated source, for page-aware segmentation.
OptionalTextExtracted text of the source content, when it has any.
Input to BaseSegmenter.Segment.