From 10a3c43140a41f4f7f520598bc19a6a0aafc9bb2 Mon Sep 17 00:00:00 2001
From: Fan Yang <missdeer@gmail.com>
Date: Mon, 26 Feb 2024 22:43:10 +0800
Subject: [PATCH] add 69xinshu handler WIP

---
 handler/69xinshu.go | 99 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 99 insertions(+)
 create mode 100644 handler/69xinshu.go

diff --git a/handler/69xinshu.go b/handler/69xinshu.go
new file mode 100644
index 0000000..222ec30
--- /dev/null
+++ b/handler/69xinshu.go
@@ -0,0 +1,99 @@
+package handler
+
+import (
+	"bytes"
+	"fmt"
+	"log"
+	"regexp"
+	"strings"
+	"unicode/utf8"
+
+	"github.com/PuerkitoBio/goquery"
+	"github.com/missdeer/getnovel/config"
+	"github.com/missdeer/golib/ic"
+)
+
+func preprocess69xinshuChapterListURL(u string) string {
+	reg := regexp.MustCompile(`https://www\.69xinshu\.com/book/([0-9]+)\.htm`)
+	if reg.MatchString(u) {
+		ss := reg.FindAllStringSubmatch(u, -1)
+		s := ss[0]
+		return fmt.Sprintf("https://www.69xinshu.com/book/%s/", s[1])
+	}
+	return u
+}
+
+func extract69xinshuChapterList(u string, rawPageContent []byte) (title string, chapters []*config.NovelChapterInfo) {
+	c := ic.Convert("gbk", "utf-8", rawPageContent)
+	c = bytes.Replace(c, []byte("\r\n"), []byte(""), -1)
+	c = bytes.Replace(c, []byte("\r"), []byte(""), -1)
+	c = bytes.Replace(c, []byte("\n"), []byte(""), -1)
+
+	doc, err := goquery.NewDocumentFromReader(bytes.NewReader(c))
+	if err != nil {
+		log.Fatal(err)
+	}
+	doc.Find("div.bread a").Each(func(index int, item *goquery.Selection) {
+		title = item.Text()
+	})
+	if strings.HasSuffix(title, `章节列表`) {
+		for i := 0; i < 4; i++ {
+			_, size := utf8.DecodeLastRuneInString(title)
+			title = title[:len(title)-size]
+		}
+	}
+	doc.Find("#catalog li").Each(func(i int, s *goquery.Selection) {
+		if a := s.Find("a"); a != nil {
+			if href, exists := a.Attr("href"); exists {
+				chapters = append(chapters, &config.NovelChapterInfo{
+					Index: i + 1,
+					Title: a.Text(),
+					URL:   href,
+				})
+			}
+		}
+	})
+
+	return
+}
+
+func extract69xinshuChapterContent(rawPageContent []byte) (c []byte) {
+	c = ic.Convert("gbk", "utf-8", rawPageContent)
+	c = bytes.Replace(c, []byte("\r\n"), []byte(""), -1)
+	c = bytes.Replace(c, []byte("\r"), []byte(""), -1)
+	c = bytes.Replace(c, []byte("\n"), []byte(""), -1)
+
+	doc, err := goquery.NewDocumentFromReader(bytes.NewReader(c))
+	if err != nil {
+		log.Fatal(err)
+	}
+	c = []byte(doc.Find("div.txtnav").Text())
+	return
+}
+
+func init() {
+	registerNovelSiteHandler(&config.NovelSiteHandler{
+		Sites: []config.NovelSite{
+			{
+				Title: `69书吧`,
+				Urls:  []string{`https://www.69xinshu.com/`},
+			},
+		},
+		CanHandle: func(u string) bool {
+			patterns := []string{
+				`https://www\.69xinshu\.com/book/[0-9]+/`,
+				`^https://www\.69xinshu\.com/book/[0-9]+\.html?$`,
+			}
+			for _, pattern := range patterns {
+				reg := regexp.MustCompile(pattern)
+				if reg.MatchString(u) {
+					return true
+				}
+			}
+			return false
+		},
+		PreprocessChapterListURL: preprocess69xinshuChapterListURL,
+		ExtractChapterList:       extract69xinshuChapterList,
+		ExtractChapterContent:    extract69xinshuChapterContent,
+	})
+}