2026-03-18 17:18:23 +08:00
|
|
|
"""Transcript chunker — split raw text into retrieval-ready chunks (skeleton)."""
|
|
|
|
|
|
|
|
|
|
|
2026-03-19 14:36:14 +08:00
|
|
|
def chunk_transcript(
|
|
|
|
|
text: str, *, max_tokens: int = 512, overlap: int = 64
|
|
|
|
|
) -> list[str]:
|
2026-03-18 17:18:23 +08:00
|
|
|
"""Split transcript text into overlapping chunks."""
|
|
|
|
|
raise NotImplementedError
|
