EPUB Novel Reader

Problem

Reading EPUBs locally can feel limited when you need an offline web reader, chapter-level search, or audio playback. Manually unpacking EPUB files, converting chapters, building HTML, and wiring a TTS endpoint is fragmented and prone to cache, port, and runtime issues.

Core workflow

The skill separates the job into three scripts:

• extract_epub.py parses an EPUB file and outputs chapter JSON for retrieval and rendering.
• build_reader.py generates an HTML reader with a table of contents, search, paging, themes, and reading progress memory.
• tts_server.py runs a local Edge TTS proxy with voice discovery, speech synthesis, rate control, and .tts_cache/ caching.

In practice, start the TTS server first, then start the reader service and open the local reader in a browser. The TTS layer supports Mandarin, Cantonese, Taiwanese, and dialect voices, which makes it useful for long Chinese novels.

Boundaries and notes

It is best treated as a local Chinese reading and listening pipeline, not a general-purpose document editor or complex layout engine. Watch the Python version, edge-tts dependency, first-run network latency, and make sure the frontend TTS server address matches the running service. Multiple Python processes can cause port conflicts.