weread download,直接生成epub。仅用于技术研究。

weread.user.js

// ==UserScript==
// @name         微信读书下载
// @namespace    http://tampermonkey.net/
// @version      0.3
// @description  下载微信读书的书籍资源
// @author       tang
// @match        https://weread.qq.com/web/reader/*
// @grant unsafeWindow
// @grant GM_setValue
// @grant GM_getValue
// @grant GM_xmlhttpRequest
// @run-at document-idle
// @connect res.weread.qq.com
// @connect tencent-cloud.com
// @connect myqcloud.com
// @require         https://cdn.bootcss.com/jszip/3.2.2/jszip.js
// @require         https://cdn.bootcss.com/jquery/3.4.1/jquery.min.js
// @require         https://unpkg.com/art-template/lib/template-web.js

// ==/UserScript==

(function () {
    'use strict';



    class Ebook {
        constructor(id, title, author, intro, publisher, publishTime, maxLevel) {
            this.id = id;
            this.title = title;
            this.author = author;
            this.intro = intro;
            this.publisher = publisher;
            this.publishTime = publishTime;
            this.maxLevel = maxLevel;
            this.images = [];
        };
        setCpid(cpid) {
            this.cpid = cpid;
        }
        setIsbn(isbn) {
            this.isbn = isbn;
        }
        setChapterList(chapterList) {
            this.chapterList = chapterList;
        }
        setImages(images){
            this.images = images;
        }
    }

    class Chapter {
        constructor(uid, path, title, level, playOrder) {
            this.uid = uid;
            this.path = path;
            this.title = title;
            this.level = level;
            this.playOrder = playOrder;
            this.subChapter = [];
        };

        addSubChapter(chapter) {
            this.subChapter.push(chapter)
        }
        getLastSubChapter() {
            return this.subChapter[this.subChapter.length - 1]
        }
    }

    const buildEbook = book => {
        var maxLevel = 1
        var chapterList = []
        var prveFirstLevelChapter

        book.chapterInfos.forEach((element, i) => {
            var chapter = new Chapter(element.chapterIdx, element.chapterIdx + ".html", element.title, element.level, i + 1)
            if (chapter.level > maxLevel) {
                maxLevel = chapter.level
            }
            if (chapter.level == 1) {
                chapterList.push(chapter)
                prveFirstLevelChapter = chapter
            } else if (chapter.level == 2) {
                prveFirstLevelChapter.addSubChapter(chapter)
            } else if (chapter.level == 3) {
                if (prveFirstLevelChapter.getLastSubChapter() == undefined) {
                    prveFirstLevelChapter.addSubChapter(chapter)
                } else {
                    prveFirstLevelChapter.getLastSubChapter().addSubChapter(chapter)
                }
            } else if (chapter.level == 4) {
                if (prveFirstLevelChapter.getLastSubChapter().getLastSubChapter() == undefined) {
                    prveFirstLevelChapter.getLastSubChapter().addSubChapter(chapter)
                } else {
                    prveFirstLevelChapter.getLastSubChapter().getLastSubChapter().addSubChapter(chapter)
                }
            } else {
                alert("暂不支持五级目录深度 " + chapter.level)
                return
            }
        });


        var ebook = new Ebook(
            book.bookInfo.bookId,
            book.bookInfo.title,
            book.bookInfo.author,
            book.bookInfo.intro,
            book.bookInfo.publisher,
            book.bookInfo.publishTime,
            maxLevel
        )
        ebook.setChapterList(chapterList)
        ebook.setIsbn(book.bookInfo.isbn)
        ebook.setCpid(book.bookInfo.cpid)
        ebook.setImages(bookImages)
        return ebook
    }

    const sleep = ms => {
        return new Promise(resolve =>
            setTimeout(resolve, ms)
        )
    }

    function get(url, headers, type) {
        return new Promise((resolve, reject) => {
            let requestObj = GM_xmlhttpRequest({
                method: "GET", url, headers,
                responseType: type || 'json',
                onload: (res) => {
                    if (res.status === 204) {
                        requestObj.abort();
                    }
                    if (type === 'blob') {
                        resolve(res.response);
                    } else {
                        resolve(res.response || res.responseText);
                    }
                },
                onerror: (err) => {
                    reject(err);
                },
            });
        });
    }

    function createAndDownloadFile(fileName, content) {
        var aTag = document.createElement('a');
        aTag.download = fileName;
        aTag.href = URL.createObjectURL(content);
        aTag.click();
        URL.revokeObjectURL(content);
    }

    const imageUrlToBlob = url => get(url, {}, 'blob')

    //var $ = unsafeWindow.$
    var vue = $("div.readerContent.routerView")[0]

    // Your code here...
    //book = unsafeWindow.$("div.readerContent.routerView").__vue__

    const parseCss = cssText => cssText.replace(/\.readerChapterContent/g, "")

    const fixBody = failBody => failBody.replace("</body>", "</div>")

    const buildHead = book => `<title>${book.currentChapter.title}</title>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"/>`

    const buildHtml = (head, body, css) => {
        var html = `<?xml version='1.0' encoding='utf-8'?><html xmlns="http://www.w3.org/1999/xhtml" xml:lang="zh-CN"><head>${head}</head><body><style>${css}</style>${body}</body></html>`
        return html
    }

    function cleanAttr(element) {
        element.removeAttr("data-wr-co")
        element.removeAttr("data-wr-bd")
        element.removeAttr("data-wr-id")
        element.removeAttr("data-ratio")
        element.removeAttr("data-w")
        element.removeAttr("data-w-new")

        element.removeData("wr-co")
        element.removeData("wr-bd")
        element.removeData("wr-id")
        element.removeData("ratio")
        element.removeData("w")
        element.removeData("w-new")
    }

    function cleanTag(element) {
        cleanAttr(element)
        element.html(element.text())
    }

    const bookImages = []

    const log = str => {
        $('.readerMemberCardTips').attr("style", "")
        $('.readerMemberCardTips > .text').html(str)
    }

    const replaceImages = (doc, zip) => {
        doc.find("img")
            .each(function () {
                var img = $(this)
                var url = img.attr("data-src");
                console.log("处理图片  " + url)
                if (url.indexOf("http") == -1) return

                var imageName = url.substr(url.lastIndexOf("/") + 1)
                if (imageName.indexOf(".") == -1) imageName += ".jpg"
                zip.file("img/" + imageName, imageUrlToBlob(url))
                img.attr("src", "../img/" + imageName)
                img.removeAttr("data-src")
                bookImages.push("img/" + imageName)
            })
        return doc.html()
    }

    const cleanHtml = doc => {
        doc.find("div").each(function () {
            cleanAttr($(this))
        })

        doc.find("img").each(function () {
            cleanAttr($(this))
        })

        doc.find("h1").each(function () {
            cleanTag($(this))
        })
        doc.find("h2").each(function () {
            cleanTag($(this))
        })
        doc.find("h3").each(function () {
            cleanTag($(this))
        })
        doc.find("p").each(function () {
            if ($(this).find("img").length > 0) {
                cleanAttr($(this))
            } else {
                cleanTag($(this))
            }
        })
    }

    var tocncx = ['<?xml version="1.0" encoding="UTF-8"?>',
        '<!DOCTYPE ncx PUBLIC "-//NISO//DTD ncx 2005-1//EN" "http://www.daisy.org/z3986/2005/ncx-2005-1.dtd">',
        '<ncx xmlns="http://www.daisy.org/z3986/2005/ncx/" version="2005-1" xml:lang="zh-CN">',
        '    <head>',
        '        <meta name="dtb:uid" content="{{book.id}}"/>',
        '        <meta name="dtb:depth" content="{{book.maxLevel}}"/>',
        '        <meta name="dtb:totalPageCount" content="0"/>',
        '        <meta name="dtb:maxPageNumber" content="0"/>',
        '    </head>',
        '    <docTitle>',
        '        <text>{{ book.title }}</text>',
        '    </docTitle>',
        '    <docAuthor>',
        '        <text>{{ book.author }}</text>',
        '    </docAuthor>',
        '    <navMap>',
        '        <navPoint class="toc" id="toc" playOrder="1">',
        '            <navLabel>',
        '                <text>Table of Contents</text>',
        '            </navLabel>',
        '            <content src="text/toc.html"/>',
        '        </navPoint>',
        '        {{each book.chapterList}}',
        '        <navPoint class="chapter" id="chapter_{{$value.uid}}" playOrder="{{$value.playOrder}}">',
        '            <navLabel>',
        '                <text>{{$value.title}}</text>',
        '            </navLabel>',
        '            <content src="text/{{$value.path}}"/>',
        '            {{each $value.subChapter}}',
        '            <navPoint class="chapter" id="chapter_{{$value.uid}}" playOrder="{{$value.playOrder}}">',
        '                <navLabel>',
        '                    <text>{{$value.title}}</text>',
        '                </navLabel>',
        '                <content src="text/{{$value.path}}"/>',
        '                {{each $value.subChapter}}',
        '                <navPoint class="chapter" id="chapter_{{$value.uid}}" playOrder="{{$value.playOrder}}">',
        '                    <navLabel>',
        '                        <text>{{$value.title}}</text>',
        '                    </navLabel>',
        '                    <content src="text/{{$value.path}}"/>',
        '                    {{each $value.subChapter}}',
        '                    <navPoint class="chapter" id="chapter_{{$value.uid}}" playOrder="{{$value.playOrder}}">',
        '                        <navLabel>',
        '                            <text>{{$value.title}}</text>',
        '                        </navLabel>',
        '                        <content src="text/{{$value.path}}"/>',
        '                    </navPoint>',
        '                    {{/each}}',
        '                </navPoint>',
        '                {{/each}}',
        '            </navPoint>',
        '            {{/each}}',
        '        </navPoint>',
        '        {{/each}}',
        '    </navMap>',
        '</ncx>'].join("\n");

    var tochtml = ['<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">',
        '<html xmlns="http://www.w3.org/1999/xhtml">',
        '<head>',
        '    <title>Table of Contents</title>',
        '    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"/>',
        '</head>',
        '<body>',
        '<h1><b>TABLE OF CONTENTS</b></h1>',
        '<br/>',
        '{{each book.chapterList}}',
        '<h3><b><a href="{{$value.path}}">{{$value.title}}</a></b></h3>',
        '<ul>',
        '    {{each $value.subChapter}}',
        '    <li><a href="{{$value.path}}">{{$value.title}}</a></li>',
        '    {{if $value.subChapter}}',
        '    <ul>',
        '        {{each $value.subChapter}}',
        '        <li><a href="{{$value.path}}">{{$value.title}}</a></li>',
        '        {{if $value.subChapter}}',
        '        <ul>',
        '            {{each $value.subChapter}}',
        '            <li><a href="{{$value.path}}">{{$value.title}}</a></li>',
        '            {{/each}}',
        '        </ul>',
        '        {{/if}}',
        '        {{/each}}',
        '    </ul>',
        '    {{/if}}',
        '    {{/each}}',
        '</ul>',
        '{{/each}}',
        '</body>',
        '</html>',
    ].join("\n");

    var opf_tmp = ['<?xml version="1.0" encoding="utf-8"?>',
        '<package xmlns="http://www.idpf.org/2007/opf" version="2.0" unique-identifier="BookId">',
        '    <metadata xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:dcterms="http://purl.org/dc/terms/" xmlns:opf="http://www.idpf.org/2007/opf" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">',
        '        <dc:title>{{ book.title }}</dc:title>',
        '        <dc:language>zh-cn</dc:language>',
        '        <dc:creator>{{ book.author }}</dc:creator>',
        '        {{if book.intro}}',
        '        <dc:description>&lt;div&gt;',
        '            &lt;p&gt;{{book.intro}}&lt;/p&gt;&lt;/div&gt;',
        '        </dc:description>',
        '        {{/if}}',
        '        {{if book.publisher}}',
        '        <dc:publisher>{{book.publisher}}</dc:publisher>',
        '        {{/if}}',
        '        {{if book.publishTime}}',
        '        <dc:date>{{book.publishTime}}</dc:date>',
        '        {{/if}}',
        '        {{if book.isbn}}',
        '        <dc:identifier opf:scheme="ISBN">{{book.isbn}}</dc:identifier>',
        '        {{/if}}',
        '        {{if book.cpid}}',
        '        <dc:identifier opf:scheme="CPID">{{book.cpid}}</dc:identifier>',
        '        {{/if}}',
        '        <meta name="cover" content="cover_image"/>',
        '    </metadata>',
        '    <manifest>',
        '        <item id="cover_image" href="cover.jpg" media-type="image/jpeg"/>',
        '        <item id="toc" media-type="application/x-dtbncx+xml" href="toc.ncx"/>',
        '        <item id="toc_html" media-type="application/xhtml+xml" href="toc.html"/>',
        '        {{each book.chapterList}}',
        '        <item id="chapter_{{$value.uid}}" media-type="application/xhtml+xml" href="text/{{$value.path}}"/>',
        '        {{each $value.subChapter}}',
        '        <item id="chapter_{{$value.uid}}" media-type="application/xhtml+xml" href="text/{{$value.path}}"/>',
        '        {{each $value.subChapter}}',
        '        <item id="chapter_{{$value.uid}}" media-type="application/xhtml+xml" href="text/{{$value.path}}"/>',
        '        {{each $value.subChapter}}',
        '        <item id="chapter_{{$value.uid}}" media-type="application/xhtml+xml" href="text/{{$value.path}}"/>',
        '        {{/each}}',
        '        {{/each}}',
        '        {{/each}}',
        '        {{/each}}',
        '        {{each book.images}}',
        '        <item id="image_{{$index}}" media-type="image/jpeg" href="{{$value}}"/>',
        '       {{/each}}',
        '    </manifest>',
        '    <spine toc="toc">',
        '        <itemref idref="toc_html"/>',
        '        {{each book.chapterList}}',
        '        <itemref idref="chapter_{{$value.uid}}"/>',
        '        {{each $value.subChapter}}',
        '        <itemref idref="chapter_{{$value.uid}}"/>',
        '        {{each $value.subChapter}}',
        '        <itemref idref="chapter_{{$value.uid}}"/>',
        '        {{each $value.subChapter}}',
        '        <itemref idref="chapter_{{$value.uid}}"/>',
        '        {{/each}}',
        '        {{/each}}',
        '        {{/each}}',
        '        {{/each}}',
        '    </spine>',
        '    <guide>',
        '    </guide>',
        '</package>',
    ].join("\n");

    var addToc = (book, zip) => {
        var toc = book.bookInfo.title
        book.chapterInfos.forEach(element => {
            var levelStr = "#".repeat(element.level)
            toc += "\n" + levelStr + " " + element.title
        });
        log("addToc")
        console.log(toc)
        //zip.file("toc.md", toc);

        var ebook = buildEbook(book)
        zip.file("toc.ncx", template.render(tocncx, { "book": ebook }));
        zip.file("text/toc.html", template.render(tochtml, { "book": ebook }));
        zip.file("content.opf", template.render(opf_tmp, { "book": ebook }));
        zip.file("mimetype", "application/epub+zip");


       var containerStr = ['<?xml version="1.0" encoding="UTF-8"?>',
        '<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">',
        '    <rootfiles>',
        '        <rootfile full-path="content.opf" media-type="application/oebps-package+xml"/>',
        '   </rootfiles>',
        '</container>'].join("\n");

        zip.file("META-INF/container.xml",containerStr)
    }

    var addInfo = (book, zip) => {
        log("addInfo")
        book.bookInfo.cover = $('img.wr_bookCover_img').attr("src")
        zip.file("bookInfo.json", JSON.stringify(book.bookInfo));
        zip.file("chapterInfos.json", JSON.stringify(book.chapterInfos));
        //zip.file("readme.txt", "使用kindlegen生成电子书，执行命令：\nkindlegen -dont_append_source " + book.bookInfo.title + ".opf");
    }

    var addCover = (book, zip) => {
        log("addCover")
        book.bookInfo.cover = $('img.wr_bookCover_img').attr("src")
        zip.file("cover.jpg", imageUrlToBlob(book.bookInfo.cover));
    }

    var count = 0
    var addChapter = (book, zip) => {
        log("正在下载数据  " + (count + 1) + "/" + book.chapterInfos.length + " : " + book.currentChapter.title)
        var head = buildHead(book)
        var rawBody = $(fixBody(book.chapterContentForEPub.join('')))
        cleanHtml(rawBody)
        var body = replaceImages(rawBody, zip)
        var newHtml = buildHtml(head, body, parseCss(book.chapterContentStyles));
        zip.file("text/" + book.currentChapter.chapterIdx + ".html", newHtml);
        count++
    }

    var download = (book, zip) => {
        if (count >= book.chapterInfos.length) {

            addToc(book, zip)
            console.log("生成epub文件")
            // if (count >= 4) {
            zip.generateAsync({ type: "blob" })
                .then(function (content) {
                    unsafeWindow.rawBook = content

                    log('已获取全部数据，点击<a href="javascript:" title="下载" class="click_download">下载</a>')
                    $(".click_download").click(function () {
                        if (unsafeWindow.rawBook) {
                            createAndDownloadFile(book.bookInfo.title + ".epub", unsafeWindow.rawBook);
                        } else {
                            log("缺失文件，请重新下载")
                        }
                    })
                    $(".click_download").click()
                });

            return
        }

        sleep(3000).then(() => {
            book.handleNextChapter().then(() => {
                addChapter(book, zip)
                download(book, zip)
            });
        })
    }


    sleep(5000).then(() => {

        var book = vue.__vue__
        unsafeWindow.book = book
        var downloadBtn = '<button title="下载" class="readerControls_item download1"><span class="icon" style="background-image: url(data:image/png;base64,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);"></span></button>';
        $('button.catalog').after(downloadBtn);


        $(".download1").click(function () {

            if (!book.isEPub) {
                alert("该书源非EPUB，暂不支持下载！")
            }
            var zip = new JSZip();
            unsafeWindow.$zip = zip
            // addInfo(book, zip)
            addCover(book, zip)
            book.changeChapter({ 'chapterUid': book.chapterInfos[0]['chapterUid'] }).then(() => {
                addChapter(book, zip)
                download(book, zip)
            })
        })
        console.log("微信读书下载插件已加载！")
        console.log(buildEbook(book))
    })



})();