📖 EBook-RS: Multi-Format Rust EBook Parser and Reader Engine (v0.15.0)

ebook-rs is a high-performance, 100% pure Rust parser and reader engine for EPUB 2, EPUB 3, Amazon KFX (Kindle Format 10), MOBI, AZW3 (KF8), FB2 (FictionBook 2), KEPUB (Kobo EPUB), LIT (Microsoft Reader), CBZ (Comic Book ZIP), PDF, ODT, TXT, and MD formats, designed for full feature parity with epub.js and foliate-js, equipped with Model Context Protocol (MCP 2024-11-05) AI assistant support, native AI/RAG document chunking with Okapi BM25 scoring, C FFI multi-language bindings, and Web Component support.

⚡ Feature Parity Matrix

📂 Format Support

🧭 Navigation & Rendering

🔍 Search, Analytics & AI RAG

🌐 Interoperability & Web Component

⚡ Format Conversion Benchmark (ebook-rs vs Calibre ebook-convert)

Empirically measured conversion benchmark converting eBook formats to EPUB 3 on identical test hardware:

All conversions produce 100% W3C-validated EPUB 3 archives containing full metadata (dc:creator, dc:title, dc:language), Table of Contents (nav.xhtml), and image resources (OEBPS/images/).

🆕 What's New in v0.14.1

• Academic PDF Two-Column Spatial Reflowing Engine (src/pdf.rs):
  • Automatically detects multi-column line patterns and spatial column dividers in IEEE, ArXiv, and ACM academic paper PDFs.
  • Separates full-width paper headers/titles from column blocks, sorting Left-Column paragraphs top-to-bottom followed by Right-Column paragraphs top-to-bottom into continuous single-column EPUB sections & AI RAG Markdown chunks.
• Multi-Threaded Parallel ZIP Exporter (src/validator.rs):
  • Parallelizes HTML section document and image asset compression across Rayon worker threads (entries.par_iter()), significantly accelerating EPUB 3 export speeds for 100MB+ image-heavy books.
• CBZ Comic Reader Optimizations (src/cbz.rs):
  • Zero-Latency Page Pre-fetching: Added CbzBook::prefetch_page_images() and book.prefetch_comic_pages() to pre-load adjacent comic page image bytes into memory for instant page turns.
  • 2-Page Manga Spread View Mode: Added CbzBook::parse_manga() and CbzBook::enable_manga_mode() with Right-to-Left (direction: rtl) reading progression.
• Native Python / PyO3 Wheel Bindings (pip install ebook-rs):
  • Native Python extension bindings (PyBook, PySection) with automated PyPI publishing workflow (.github/workflows/pypi.yml).

🚀 Quick Start Examples

1. Basic Parsing & RAG Chunking (Rust)

use ebook_rs::{Book, RagChunkConfig};

fn main() -> Result<(), String> {
    let book = Book::from_file("book.epub")?;

    println!("Title: {}", book.metadata().title);
    println!("Language: {:?}", book.detect_language());

    // Generate AI / RAG document chunks with CFI citation anchors
    let config = RagChunkConfig {
        max_tokens: 512,
        overlap_tokens: 64,
        preserve_headings: true,
        ..Default::default()
    };
    let chunks = book.to_rag_chunks(&config);

    for chunk in chunks {
        println!("ID: {}", chunk.id);
        println!("CFI Anchor: {}", chunk.cfi);
        println!("Markdown: \n{}", chunk.markdown);
    }

    Ok(())
}

2. Multi-Format CLI Conversion & RAG Export

# Convert MOBI, PDF, FB2, or TXT into a valid EPUB 3 file
ebook-rs convert sample.mobi output.epub

# Export AI RAG chunks directly to JSON
ebook-rs convert book.epub chunks.json

# Stream JSON RAG chunks to stdout
ebook-rs rag book.epub 512

3. Native Python API (pip install ebook-rs)

import ebook_rs

# Open any eBook format (EPUB, MOBI, KFX, AZW3, FB2, LIT, CBZ, PDF, TXT, MD)
book = ebook_rs.Book.open("sample.mobi")

print(f"Title: {book.title}")
print(f"Authors: {book.authors}")
print(f"Section Count: {book.section_count}")

# Generate AI / RAG chunks with CFI citations directly in Python
rag_chunks_json = book.to_rag_chunks_json(max_tokens=512)

# Convert any format to EPUB 3 or KFX bytes
epub_bytes = book.export_epub3_bytes()

4. C FFI API Integration

#include "ebook_rs.h"

int main() {
    uint8_t* buffer = load_file("book.epub", &len);
    CBookHandle book = ebook_rs_book_from_bytes(buffer, len);

    char* json_meta = ebook_rs_get_metadata_json(book);
    printf("Metadata: %s\n", json_meta);
    ebook_rs_string_free(json_meta);

    char* json_rag = ebook_rs_to_rag_chunks_json(book, 512);
    printf("RAG Chunks: %s\n", json_rag);
    ebook_rs_string_free(json_rag);

    ebook_rs_book_free(book);
    return 0;
}

📖 Documentation & Wiki

Detailed API documentation and integration guides are available in docs/API.md, API.md, and WIKI.md.

📜 License

Licensed under the MIT License. Copyright (c) 2026 SV-Stark.