goyomi/sources/base/oceanwp/oceanwp.go

// Package oceanwp implements the OceanWP manga base.
// Blog-style WordPress site; each post is a single-chapter gallery; pages from entry-content images.
package oceanwp

import (
	"context"
	"fmt"
	"net/http"
	"strings"

	"github.com/PuerkitoBio/goquery"

	"goyomi/internal/httpclient"
	"goyomi/internal/source"
	"goyomi/sources/base/util"
)

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 false }

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("oceanwp: HTTP %d", resp.StatusCode)
	}
	return goquery.NewDocumentFromReader(resp.Body)
}

func (s *Source) parseMangaList(doc *goquery.Document) source.MangasPage {
	var mangas []source.SManga
	doc.Find("article.blog-entry").Each(func(_ int, el *goquery.Selection) {
		titleEl := el.Find("h2.blog-entry-title a").First()
		if titleEl.Length() == 0 {
			return
		}
		m := source.SManga{
			URL:   titleEl.AttrOr("href", ""),
			Title: strings.TrimSpace(titleEl.Text()),
		}
		if img := el.Find("img").First(); img.Length() > 0 {
			m.ThumbnailURL = util.AbsURL(s.cfg.BaseURL, img.AttrOr("src", img.AttrOr("data-src", "")))
		}
		if m.URL != "" && m.Title != "" {
			mangas = append(mangas, m)
		}
	})
	hasNext := doc.Find(".pagination .next, a[rel=next]").Length() > 0
	return source.MangasPage{Mangas: mangas, HasNextPage: hasNext}
}

func (s *Source) GetPopularManga(page int) (source.MangasPage, error) {
	var u string
	if page == 1 {
		u = s.base()
	} else {
		u = fmt.Sprintf("%s/page/%d/", s.base(), page)
	}
	doc, err := s.get(context.Background(), u)
	if err != nil {
		return source.MangasPage{}, err
	}
	return s.parseMangaList(doc), nil
}

func (s *Source) GetLatestUpdates(page int) (source.MangasPage, error) {
	return s.GetPopularManga(page)
}

func (s *Source) GetSearchManga(page int, query string, filters []source.Filter) (source.MangasPage, error) {
	u := fmt.Sprintf("%s/?s=%s&paged=%d", s.base(), query, page)
	doc, err := s.get(context.Background(), u)
	if err != nil {
		return source.MangasPage{}, err
	}
	return s.parseMangaList(doc), nil
}

func (s *Source) GetMangaDetails(manga source.SManga) (source.SManga, error) {
	doc, err := s.get(context.Background(), util.AbsURL(s.cfg.BaseURL, manga.URL))
	if err != nil {
		return manga, err
	}
	result := source.SManga{URL: manga.URL}
	result.Title = strings.TrimSpace(doc.Find("h1.entry-title, h1.page-title").First().Text())
	if result.Title == "" {
		result.Title = manga.Title
	}
	result.Description = strings.TrimSpace(doc.Find("div.entry-content p").First().Text())
	if img := doc.Find("div.entry-content img, img.wp-post-image").First(); img.Length() > 0 {
		result.ThumbnailURL = util.AbsURL(s.cfg.BaseURL, img.AttrOr("src", img.AttrOr("data-src", "")))
	}
	return result, nil
}

func (s *Source) GetChapterList(manga source.SManga) ([]source.SChapter, error) {
	return []source.SChapter{{
		URL:  manga.URL,
		Name: "Chapter 1",
	}}, nil
}

func (s *Source) GetPageList(chapter source.SChapter) ([]source.Page, error) {
	doc, err := s.get(context.Background(), util.AbsURL(s.cfg.BaseURL, chapter.URL))
	if err != nil {
		return nil, err
	}
	var pages []source.Page
	doc.Find("div.entry-content img").Each(func(i int, img *goquery.Selection) {
		u := img.AttrOr("src", img.AttrOr("data-src", ""))
		if u != "" {
			pages = append(pages, source.Page{Index: i, ImageURL: util.AbsURL(s.cfg.BaseURL, 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 }