API overview

The CLI is the shortest route for normal backups. The Python API exposes each stage separately for applications that need custom storage, rendering, OCR, or workflow orchestration.

Module	Purpose
glyphive.archive	Serialize a tree, list selected paths, and parse archive records
glyphive.codec	Resolve named printable codecs and use base16g-crc16-rs
glyphive.compression	Resolve whole-stream compression methods
glyphive.layout	Paginate encoded lines and validate protected pages
glyphive.render	Resolve and invoke text, PDF, and Word renderers
glyphive.restore	Decode a document and safely write archive records
glyphive.restore.ocr	Discover providers and OCR one or more images

Pipeline

archive.archive_tree(root)
  -> compression.get(name).compress(raw)
  -> codec.get("base16g-crc16-rs").encode(payload)
  -> layout.paginate(lines, metadata, lines_per_page=...)
  -> render.render(pages, output, format)

restore.decode_document(text_lines)
  -> restore.unarchive_bytes(raw, destination)

Registry names are serialized in the protected document header. Treat them as wire identifiers: adding an implementation under a new name is extensible; changing the behavior of an existing identifier breaks compatibility.

Optional renderer and OCR dependencies are checked only when the corresponding implementation is selected.