init

.github/workflows/go.yml

@@ -0,0 +1,28 @@
+# This workflow will build a golang project
+# For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-go
+
+name: Go
+
+on:
+ push:
+ branches: [ "master" ]
+ pull_request:
+ branches: [ "master" ]
+
+jobs:
+
+ build:
+ runs-on: ubuntu-latest
+ steps:
+ - uses: actions/checkout@v3
+
+ - name: Set up Go
+ uses: actions/setup-go@v3
+ with:
+ go-version: 1.23
+
+ - name: Build
+ run: go build -v ./...
+#
+# - name: Test
+# run: go test -v ./...

.gitignore

@@ -0,0 +1,9 @@
+# Default ignored files
+/target/
+/dist/
+*.exe
+/.idea/
+.fleet/
+*.xml
+*.7z
+*.bak

README.md

@@ -0,0 +1,36 @@
+# 簡介
+
+bookget 数字古籍图书下载工具，已支持约 50+ 个数字图书馆。 截止 2024/01/16 日，已停止更新。
+
+### 使用説明
+1. 打开 [https://github.com/deweizhu/bookget/releases](https://github.com/deweizhu/bookget/releases/latest) 下载最新版。
+1. [必读]使用手册wiki https://github.com/deweizhu/bookget/wiki
+1. 此项目代码仅供学习研究使用，欢迎有能力的朋友git clone 代码二次开发维护您自己的版本。
+
+#### 源碼編譯
+从源码构建，仅对计算机程序员参考。普通用户可直接跳过阅读。 
+阅读 [golang 官方文档](https://golang.google.cn/doc/install) ，给您的电脑安装 golang 开发环境。
+```shell
+git clone --depth=1 https://github.com/deweizhu/bookget.git
+cd bookget
+go build .
+```
+
+- For Win可用环境：windows 10 x64 / windows 11 （自2024/01/12日起，引入bookget-gui只适用于Win10/11 x64系统）。
+
+# 新書推薦
+
+本書即以清代避諱為例，試探如何運用融合數位與傳統的e考據之法，重新定義避諱學2.0該有的研究模式。這將是治清代的文史工作者案頭不可或缺的一本專著。[^1]
+- [清代避諱研究](https://gpi.culture.tw/books/1011300273)
+- 作者: 黃一農 / 黄一农
+- 出版社: 台灣清華大學出版社
+- 副标题: e考據的學術實踐
+- 出版年: 2024-4
+- ISBN: 9786269724987
+[^1]:[https://gpi.culture.tw/books/1011300273](https://gpi.culture.tw/books/1011300273)
+
+### 第三方社區/網站
+- 书格 https://www.shuge.org
+- Arch Linux AUR https://aur.archlinux.org/packages/bookget-git
+
+

app/berkeley.go

@@ -0,0 +1,134 @@
+package app
+
+import (
+ "bookget/config"
+ "bookget/lib/gohttp"
+ "bookget/lib/util"
+ "context"
+ "errors"
+ "fmt"
+ "log"
+ "net/http/cookiejar"
+ "net/url"
+ "path/filepath"
+ "regexp"
+)
+
+type Berkeley struct {
+ dt *DownloadTask
+}
+
+func (r *Berkeley) Init(iTask int, sUrl string) (msg string, err error) {
+ r.dt = new(DownloadTask)
+ r.dt.UrlParsed, err = url.Parse(sUrl)
+ r.dt.Url = sUrl
+ r.dt.Index = iTask
+ r.dt.BookId = r.getBookId(r.dt.Url)
+ if r.dt.BookId == "" {
+ return "requested URL was not found.", err
+ }
+ r.dt.Jar, _ = cookiejar.New(nil)
+ return r.download()
+}
+
+func (r *Berkeley) getBookId(sUrl string) (bookId string) {
+ m := regexp.MustCompile(`(?i)record/([A-z0-9_-]+)`).FindStringSubmatch(sUrl)
+ if m != nil {
+ bookId = m[1]
+ }
+ return bookId
+}
+
+func (r *Berkeley) download() (msg string, err error) {
+ name := util.GenNumberSorted(r.dt.Index)
+ log.Printf("Get %s %s\n", name, r.dt.Url)
+
+ r.dt.SavePath = CreateDirectory(r.dt.UrlParsed.Host, r.dt.BookId, "")
+ canvases, err := r.getCanvases(r.dt.Url, r.dt.Jar)
+ if err != nil || canvases == nil {
+ return "requested URL was not found.", err
+ }
+ log.Printf(" %d files \n", len(canvases))
+ r.do(canvases)
+ return "", nil
+}
+
+func (r *Berkeley) do(canvases []string) (msg string, err error) {
+ if canvases == nil {
+ return
+ }
+ fmt.Println()
+ referer := r.dt.Url
+ size := len(canvases)
+ for i, dUrl := range canvases {
+ if dUrl == "" || !config.PageRange(i, size) {
+ continue
+ }
+ sortId := util.GenNumberSorted(i + 1)
+ ext := filepath.Ext(dUrl)
+ filename := sortId + ext
+ dest := r.dt.SavePath + filename
+ if FileExist(dest) {
+ continue
+ }
+ log.Printf("Get %d/%d, URL: %s\n", i+1, size, dUrl)
+ ctx := context.Background()
+ opts := gohttp.Options{
+ DestFile: dest,
+ Overwrite: false,
+ Concurrency: 1,
+ CookieFile: config.Conf.CookieFile,
+ CookieJar: r.dt.Jar,
+ Headers: map[string]interface{}{
+ "User-Agent": config.Conf.UserAgent,
+ "Referer": referer,
+ },
+ }
+ gohttp.FastGet(ctx, dUrl, opts)
+ fmt.Println()
+ }
+ return "", err
+}
+
+func (r *Berkeley) getVolumes(sUrl string, jar *cookiejar.Jar) (volumes []string, err error) {
+ //TODO implement me
+ panic("implement me")
+}
+
+func (r *Berkeley) getCanvases(sUrl string, jar *cookiejar.Jar) (canvases []string, err error) {
+ bs, err := r.getBody(sUrl, jar)
+ if err != nil {
+ return
+ }
+ matches := regexp.MustCompile(`value="https://([^"]+)\.pdf"`).FindAllSubmatch(bs, -1)
+ if matches == nil {
+ return nil, errors.New("not match")
+ }
+ for _, match := range matches {
+ pdfUrl := "https://" + string(match[1]) + ".pdf"
+ canvases = append(canvases, pdfUrl)
+ }
+ return
+}
+
+func (r *Berkeley) getBody(apiUrl string, jar *cookiejar.Jar) ([]byte, error) {
+ referer := url.QueryEscape(apiUrl)
+ ctx := context.Background()
+ cli := gohttp.NewClient(ctx, gohttp.Options{
+ CookieFile: config.Conf.CookieFile,
+ CookieJar: jar,
+ Headers: map[string]interface{}{
+ "User-Agent": config.Conf.UserAgent,
+ "Referer": referer,
+ },
+ })
+ resp, err := cli.Get(apiUrl)
+ if err != nil {
+ return nil, err
+ }
+ bs, _ := resp.GetBody()
+ if resp.GetStatusCode() != 200 || bs == nil {
+ return nil, errors.New(fmt.Sprintf("ErrCode:%d, %s", resp.GetStatusCode(), resp.GetReasonPhrase()))
+ }
+ return bs, nil
+}