📖 EBook-RS: Multi-Format Rust EBook Parser and Reader Engine (v0.13.6)

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 native AI/RAG document chunking, 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 v0.13.6

• Ultra-Fast Multi-Format Conversion Engine (ebook-rs convert):
  • Convert any supported eBook format (MOBI, AZW3, FB2, LIT, KFX, PDF, CBZ, ODT, TXT, MD) directly into W3C-valid EPUB 3 archives or KFX outputs in < 1.1s (up to 9.7× faster than Calibre).
• 100% Embedded Image Asset Extraction:
  • Decodes and extracts all embedded JPEG, PNG, GIF, and WebP images from MOBI/AZW3 PalmDOC records, FB2 Base64 <binary> tags, Microsoft Reader LIT container streams, and Amazon KFX payload streams into OEBPS/images/.
• Amazon KFX (Kindle Format 10) & LIT Chapter Sectioning:
  • Intelligent paragraph and heading chunking for KFX and LIT formats, generating proper chapter sections (sec_0.xhtml .. sec_N.xhtml) instead of single-line micro-files or monolith documents.
• Lazy MOBI Image Base64 Inlining:
  • Deferred heavy Base64 image inlining to render-time (before_display_hooks), eliminating conversion hangs on large image-heavy MOBI/AZW3 books.
• Complete Metadata Parity:
  • Extracts and populates OPF <dc:creator>, <dc:title>, <dc:language>, <dc:publisher>, and <dc:identifier> tags across all formats.

🚀 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. C / Python 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.