phase3: implement fansubscat, fuzzydoodle, galleryadults, gattsu, goda

2026-05-11 06:45:48 +07:00
parent 1e6d72b046
commit 224266ffe3
6 changed files with 1178 additions and 5 deletions
@@ -0,0 +1,314 @@
+// Package goda implements the GoDa manga base.
+// Popular: GET {base}/hots/page/{n}; Chapter list via {base}/manga/get?mid={id}&mode=all.
+package goda
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"strings"
+
+	"github.com/PuerkitoBio/goquery"
+
+	"goyomi/internal/httpclient"
+	"goyomi/internal/source"
+)
+
+type Config struct {
+	Name    string
+	BaseURL string
+	Lang    string
+}
+
+type Source struct {
+	cfg    Config
+	client *httpclient.Client
+	id     int64
+}
+
+func New(cfg Config) *Source {
+	c := httpclient.NewClient(httpclient.WithRateLimit(1, 2))
+	return &Source{cfg: cfg, client: c, id: source.GenerateSourceID(cfg.Name, cfg.Lang)}
+}
+
+func (s *Source) ID() int64            { return s.id }
+func (s *Source) Name() string         { return s.cfg.Name }
+func (s *Source) Lang() string         { return s.cfg.Lang }
+func (s *Source) SupportsLatest() bool { return true }
+
+func (s *Source) base() string { return strings.TrimRight(s.cfg.BaseURL, "/") }
+
+func (s *Source) get(ctx context.Context, rawURL string) (*goquery.Document, error) {
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, rawURL, nil)
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("Referer", s.cfg.BaseURL+"/")
+	resp, err := s.client.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("goda: HTTP %d", resp.StatusCode)
+	}
+	return goquery.NewDocumentFromReader(resp.Body)
+}
+
+func (s *Source) getRaw(ctx context.Context, rawURL string) ([]byte, error) {
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, rawURL, nil)
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("Referer", s.cfg.BaseURL+"/")
+	resp, err := s.client.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+	return io.ReadAll(resp.Body)
+}
+
+// getKey extracts the manga key from a URL: strip /manga/ prefix and trailing /.
+func getKey(href string) string {
+	u, err := url.Parse(href)
+	if err != nil {
+		return href
+	}
+	path := u.Path
+	if idx := strings.Index(path, "/manga/"); idx >= 0 {
+		path = path[idx+len("/manga/"):]
+	}
+	return strings.TrimSuffix(path, "/")
+}
+
+// getMangaID returns the manga ID from the #mangachapters element.
+func getMangaID(doc *goquery.Document) string {
+	return doc.Find("#mangachapters").AttrOr("data-mid", "")
+}
+
+func (s *Source) parseMangaList(doc *goquery.Document) source.MangasPage {
+	var mangas []source.SManga
+	doc.Find(".container > .cardlist .pb-2 a").Each(func(_ int, el *goquery.Selection) {
+		m := source.SManga{}
+		href, _ := el.Attr("href")
+		m.URL = getKey(href)
+		m.Title = strings.TrimSpace(el.Find("h3").Text())
+		if img := el.Find("img").First(); img.Length() > 0 {
+			src := img.AttrOr("src", "")
+			// Some sites proxy images as ?url=...
+			if strings.Contains(src, "url=") {
+				if parsed, err := url.Parse(src); err == nil {
+					if u := parsed.Query().Get("url"); u != "" {
+						src = u
+					}
+				}
+			}
+			m.ThumbnailURL = src
+		}
+		if m.URL != "" && m.Title != "" {
+			mangas = append(mangas, m)
+		}
+	})
+	nextLabel := "NEXT"
+	if s.cfg.Lang == "zh" {
+		nextLabel = "下一頁"
+	}
+	hasNext := doc.Find(fmt.Sprintf("a[aria-label=%s] button", nextLabel)).Length() > 0
+	return source.MangasPage{Mangas: mangas, HasNextPage: hasNext}
+}
+
+func (s *Source) GetPopularManga(page int) (source.MangasPage, error) {
+	doc, err := s.get(context.Background(), fmt.Sprintf("%s/hots/page/%d", s.base(), page))
+	if err != nil {
+		return source.MangasPage{}, err
+	}
+	return s.parseMangaList(doc), nil
+}
+
+func (s *Source) GetLatestUpdates(page int) (source.MangasPage, error) {
+	doc, err := s.get(context.Background(), fmt.Sprintf("%s/newss/page/%d", s.base(), page))
+	if err != nil {
+		return source.MangasPage{}, err
+	}
+	return s.parseMangaList(doc), nil
+}
+
+func (s *Source) GetSearchManga(page int, query string, filters []source.Filter) (source.MangasPage, error) {
+	return s.GetPopularManga(page)
+}
+
+func (s *Source) GetMangaDetails(manga source.SManga) (source.SManga, error) {
+	doc, err := s.get(context.Background(), fmt.Sprintf("%s/manga/%s", s.base(), manga.URL))
+	if err != nil {
+		return manga, err
+	}
+	result := source.SManga{URL: manga.URL}
+	main := doc.Find("main").First()
+	titleEl := main.Find("h1").First()
+	result.Title = titleEl.Text()
+	if span := titleEl.Find("span").First(); span.Length() > 0 {
+		result.Title = strings.TrimSpace(strings.Replace(result.Title, span.Text(), "", 1))
+	}
+	if result.Title == "" {
+		result.Title = manga.Title
+	}
+
+	// Status badge is first child of h1
+	statusText := strings.TrimSpace(titleEl.Children().First().Text())
+	switch statusText {
+	case "連載中", "Ongoing":
+		result.Status = source.StatusOngoing
+	case "完結", "Completed":
+		result.Status = source.StatusCompleted
+	case "停止更新", "Cancelled":
+		result.Status = source.StatusCancelled
+	case "休刊", "On Hiatus":
+		result.Status = source.StatusHiatus
+	default:
+		result.Status = source.StatusUnknown
+	}
+
+	// Siblings structure: h1, then elements for author, genre1, tags, description
+	parent := titleEl.Parent()
+	children := parent.Children()
+	if children.Length() >= 2 {
+		var authorParts []string
+		children.Eq(1).Children().Each(func(i int, el *goquery.Selection) {
+			if i == 0 {
+				return // skip label
+			}
+			t := strings.TrimSuffix(strings.TrimSpace(el.Text()), " ,")
+			if t != "" {
+				authorParts = append(authorParts, t)
+			}
+		})
+		result.Author = strings.Join(authorParts, ", ")
+	}
+
+	var genres []string
+	if children.Length() >= 3 {
+		children.Eq(2).Children().Each(func(i int, el *goquery.Selection) {
+			if i == 0 {
+				return // skip label
+			}
+			t := strings.TrimSuffix(strings.TrimSpace(el.Text()), " ,")
+			if t != "" {
+				genres = append(genres, t)
+			}
+		})
+	}
+	if children.Length() >= 4 {
+		children.Eq(3).Children().Each(func(_ int, el *goquery.Selection) {
+			t := strings.TrimPrefix(strings.TrimSpace(el.Text()), "#")
+			if t != "" {
+				genres = append(genres, t)
+			}
+		})
+	}
+	result.Genre = strings.Join(genres, ", ")
+
+	mangaID := getMangaID(doc)
+	var descParts []string
+	if children.Length() >= 5 {
+		descParts = append(descParts, strings.TrimSpace(children.Eq(4).Text()))
+	}
+	if mangaID != "" {
+		descParts = append(descParts, "ID: "+mangaID)
+	}
+	result.Description = strings.Join(descParts, "\n\n")
+
+	main.Find("img.object-cover").First().Each(func(_ int, img *goquery.Selection) {
+		result.ThumbnailURL, _ = img.Attr("src")
+	})
+	return result, nil
+}
+
+func (s *Source) GetChapterList(manga source.SManga) ([]source.SChapter, error) {
+	// Extract manga ID from description or re-fetch.
+	mangaID := ""
+	if desc := manga.Description; desc != "" {
+		if idx := strings.LastIndex(desc, "ID: "); idx >= 0 {
+			mangaID = strings.TrimSpace(desc[idx+4:])
+		}
+	}
+	if mangaID == "" {
+		doc, err := s.get(context.Background(), fmt.Sprintf("%s/manga/%s", s.base(), manga.URL))
+		if err != nil {
+			return nil, err
+		}
+		mangaID = getMangaID(doc)
+	}
+	if mangaID == "" {
+		return nil, fmt.Errorf("goda: could not find manga ID for %s", manga.URL)
+	}
+
+	body, err := s.getRaw(context.Background(), fmt.Sprintf("%s/manga/get?mid=%s&mode=all", s.base(), mangaID))
+	if err != nil {
+		return nil, err
+	}
+	doc, err := goquery.NewDocumentFromReader(strings.NewReader(string(body)))
+	if err != nil {
+		return nil, err
+	}
+
+	var chapters []source.SChapter
+	doc.Find(".chapteritem").Each(func(_ int, el *goquery.Selection) {
+		a := el.Find("a").First()
+		href, _ := a.Attr("href")
+		key := getKey(href)
+		chKey := a.AttrOr("data-cs", "")
+		name := a.AttrOr("data-ct", "")
+		if key == "" {
+			return
+		}
+		chapters = append(chapters, source.SChapter{
+			URL:  key + "#" + mangaID + "/" + chKey,
+			Name: name,
+		})
+	})
+	// reverse: chapters come latest-first from API
+	for i, j := 0, len(chapters)-1; i < j; i, j = i+1, j-1 {
+		chapters[i], chapters[j] = chapters[j], chapters[i]
+	}
+	return chapters, nil
+}
+
+func (s *Source) GetPageList(chapter source.SChapter) ([]source.Page, error) {
+	// URL format: {key}#{mangaId}/{chapterId}
+	id := ""
+	if idx := strings.LastIndex(chapter.URL, "#"); idx >= 0 {
+		id = chapter.URL[idx+1:]
+	}
+	mangaID := ""
+	chapterID := ""
+	if slashIdx := strings.Index(id, "/"); slashIdx >= 0 {
+		mangaID = id[:slashIdx]
+		chapterID = id[slashIdx+1:]
+	}
+	if mangaID == "" || chapterID == "" {
+		return nil, fmt.Errorf("goda: invalid chapter URL: %s", chapter.URL)
+	}
+
+	doc, err := s.get(context.Background(),
+		fmt.Sprintf("%s/chapter/getcontent?m=%s&c=%s", s.base(), mangaID, chapterID))
+	if err != nil {
+		return nil, err
+	}
+	var pages []source.Page
+	doc.Find("#chapcontent > div > img").Each(func(i int, img *goquery.Selection) {
+		u := img.AttrOr("data-src", "")
+		if u == "" {
+			u, _ = img.Attr("src")
+		}
+		if u != "" {
+			pages = append(pages, source.Page{Index: i, ImageURL: u})
+		}
+	})
+	return pages, nil
+}
+
+func (s *Source) GetImageURL(page source.Page) (string, error) { return page.ImageURL, nil }
+func (s *Source) GetFilterList() []source.Filter               { return nil }