ePub 阅读/转换

复制import zipfile
import xml.etree.ElementTree as ET
from pathlib import Path

# 解析 ePub（本质是 ZIP 包），提取纯文本内容
epub_path = "example.epub"
with zipfile.ZipFile(epub_path, "r") as z:
    # 读取容器文件获取 OPF 路径
    container = ET.fromstring(z.read("META-INF/container.xml"))
    ns = {"c": "urn:oasis:names:tc:opendocument:xmlns:container"}
    opf_path = container.find(".//c:rootfile", ns).get("full-path")

    # 解析 OPF 获取所有 HTML 文件列表
    opf = ET.fromstring(z.read(opf_path))
    pkg_ns = {"p": "http://www.idpf.org/2007/opf"}
    manifest = opf.find("p:manifest", pkg_ns)
    spine = opf.find("p:spine", pkg_ns)
    id_to_href = {item.get("id"): item.get("href") for item in manifest.findall("p:item", pkg_ns)}
    base_dir = Path(opf_path).parent

    # 按 spine 顺序提取文本
    texts = []
    for itemref in spine.findall("p:itemref", pkg_ns):
        href = id_to_href[itemref.get("idref")]
        content = z.read(str(base_dir / href)).decode("utf-8")
        # 简单去除 HTML 标签
        import re
        text = re.sub(r"<[^>]+>", "", content)
        texts.append(text.strip())

    print("\n".join(texts)[:500])  # 输出前 500 字符

复制package main

import (
	"archive/zip"
	"encoding/xml"
	"fmt"
	"io"
	"path/filepath"
	"regexp"
	"strings"
)

// 解析 ePub 容器文件
func main() {
	r, _ := zip.OpenReader("example.epub")
	defer r.Close()

	// 读取 META-INF/container.xml 获取 OPF 路径
	var container struct {
		Rootfiles []struct {
			FullPath string `xml:"full-path,attr"`
		} `xml:"rootfiles>rootfile"`
	}
	data, _ := readFileInZip(r, "META-INF/container.xml")
	xml.Unmarshal(data, &container)
	opfPath := container.Rootfiles[0].FullPath

	// 读取 OPF 获取 spine 顺序
	opfData, _ := readFileInZip(r, opfPath)
	type Opf struct {
		Manifest []struct {
			ID   string `xml:"id,attr"`
			Href string `xml:"href,attr"`
		} `xml:"manifest>item"`
		Spine []struct {
			IDref string `xml:"idref,attr"`
		} `xml:"spine>itemref"`
	}
	var opf Opf
	xml.Unmarshal(opfData, &opf)

	idToHref := make(map[string]string)
	for _, item := range opf.Manifest {
		idToHref[item.ID] = item.Href
	}

	baseDir := filepath.Dir(opfPath)
	re := regexp.MustCompile(`<[^>]+>`)
	for _, ref := range opf.Spine {
		href := idToHref[ref.IDref]
		content, _ := readFileInZip(r, filepath.Join(baseDir, href))
		text := re.ReplaceAllString(string(content), "")
		fmt.Println(strings.TrimSpace(text)[:200])
	}
}

func readFileInZip(r *zip.ReadCloser, name string) ([]byte, error) {
	for _, f := range r.File {
		if f.Name == name {
			rc, _ := f.Open()
			defer rc.Close()
			return io.ReadAll(rc)
		}
	}
	return nil, fmt.Errorf("file not found: %s", name)
}

复制// 浏览器端用 JSZip 解析 ePub 并提取纯文本
const JSZip = require('jszip');
const fs = require('fs');

async function extractEpubText(filePath) {
  const data = fs.readFileSync(filePath);
  const zip = await JSZip.loadAsync(data);

  // 读取容器文件
  const containerXml = await zip.file('META-INF/container.xml').async('string');
  const opfPath = containerXml.match(/full-path="([^"]+)"/)[1];

  // 读取 OPF 获取 spine 顺序
  const opfXml = await zip.file(opfPath).async('string');
  const idRefs = [...opfXml.matchAll(/<itemref[^>]+idref="([^"]+)"/g)].map(m => m[1]);
  const items = [...opfXml.matchAll(/<item[^>]+id="([^"]+)"[^>]+href="([^"]+)"/g)];
  const idToHref = Object.fromEntries(items.map(m => [m[1], m[2]]));

  const baseDir = opfPath.substring(0, opfPath.lastIndexOf('/') + 1);
  const texts = [];

  for (const id of idRefs) {
    const href = idToHref[id];
    const html = await zip.file(baseDir + href).async('string');
    const text = html.replace(/<[^>]+>/g, '').trim();
    texts.push(text);
  }

  console.log(texts.join('\n').slice(0, 500));
}

extractEpubText('example.epub').catch(console.error);