reading-companion

Convert an EPUB into a self-contained, annotatable HTML reading surface using standard Python tools, enabling durable in-flow highlights and notes for active reading workflows.

Durable annotation persistence across regenerations is guaranteed by a stable content model and TextQuote anchor-based re-attach, using a deterministic source_hash to maintain highlights.

Yes, export highlights-in-context as part of an active-reading plan that includes a structure map and adversarial questions to enhance reading comprehension and analysis.

No external dependencies are required; the conversion uses stdlib Python tools only to produce deterministic content with a stable source_hash for reliable annotation re-attachment.

An active-reading plan provides a structure map of the EPUB, generates adversarial questions, and enables highlights exports with context to guide deeper engagement with the text.

Annotations persist across regenerations through a stable content model and anchor-based re-attach, ensuring highlights stored as sidecar data remain attached to the correct text.