refactor: split site adapters and site page modules

2026-10-11 07:46:37 +08:00 · 2026-06-07 18:55:51 +08:00
parent df02fd1166
commit a5d061afa6
10 changed files with 1866 additions and 1544 deletions
@@ -0,0 +1,167 @@
+// Package service — Discuz site adapter.
+package service
+
+import (
+	"context"
+	"fmt"
+	"net/http"
+	"net/url"
+	"regexp"
+	"strconv"
+	"strings"
+	"time"
+)
+
+// ─── Discuz 适配器 ───────────────────────────────────────────────────────────
+
+// DiscuzAdapter 基于 Discuz! X 的站点适配器。
+type DiscuzAdapter struct {
+	client *http.Client
+}
+
+// NewDiscuzAdapter 创建 Discuz 适配器。
+func NewDiscuzAdapter() *DiscuzAdapter {
+	return &DiscuzAdapter{
+		client: &http.Client{Timeout: 30 * time.Second},
+	}
+}
+
+func (a *DiscuzAdapter) Authenticate(ctx context.Context, cfg SiteConfig) error {
+	u := cfg.URL + "/home.php?mod=space"
+	data, status, err := doRequest(ctx, a.client, "GET", u, cfg, nil)
+	if err != nil {
+		return fmt.Errorf("authenticate: %w", err)
+	}
+	if status == http.StatusFound {
+		return fmt.Errorf("authentication failed: redirected to login")
+	}
+	if status != http.StatusOK {
+		return fmt.Errorf("authenticate failed: status %d", status)
+	}
+	body := string(data)
+	if strings.Contains(body, "login") && !strings.Contains(body, "我的空间") {
+		return fmt.Errorf("authentication failed: not logged in")
+	}
+	return nil
+}
+
+func (a *DiscuzAdapter) Search(ctx context.Context, cfg SiteConfig, keyword string, page int) (*SiteSearchResult, error) {
+	params := url.Values{}
+	params.Set("mod", "forum")
+	params.Set("srchtxt", keyword)
+	params.Set("searchsubmit", "true")
+	params.Set("page", strconv.Itoa(page))
+
+	u := cfg.URL + "/search.php?" + params.Encode()
+	data, status, err := doRequest(ctx, a.client, "GET", u, cfg, nil)
+	if err != nil {
+		return nil, fmt.Errorf("search: %w", err)
+	}
+	if status != http.StatusOK {
+		return nil, fmt.Errorf("search failed: status %d", status)
+	}
+
+	return parseDiscuzHTML(string(data), cfg.Name, cfg.URL)
+}
+
+func (a *DiscuzAdapter) Browse(ctx context.Context, cfg SiteConfig, category string, page int) (*SiteSearchResult, error) {
+	params := url.Values{}
+	if category != "" {
+		params.Set("fid", category)
+	}
+	params.Set("page", strconv.Itoa(page))
+
+	u := cfg.URL + "/forum.php?" + params.Encode()
+	data, status, err := doRequest(ctx, a.client, "GET", u, cfg, nil)
+	if err != nil {
+		return nil, fmt.Errorf("browse: %w", err)
+	}
+	if status != http.StatusOK {
+		return nil, fmt.Errorf("browse failed: status %d", status)
+	}
+
+	return parseDiscuzHTML(string(data), cfg.Name, cfg.URL)
+}
+
+func (a *DiscuzAdapter) GetDetail(ctx context.Context, cfg SiteConfig, id string) (*TorrentDetail, error) {
+	u := cfg.URL + "/forum.php?mod=viewthread&tid=" + id
+	data, status, err := doRequest(ctx, a.client, "GET", u, cfg, nil)
+	if err != nil {
+		return nil, fmt.Errorf("detail: %w", err)
+	}
+	if status != http.StatusOK {
+		return nil, fmt.Errorf("detail failed: status %d", status)
+	}
+
+	html := string(data)
+	detail := &TorrentDetail{
+		ID:        id,
+		DetailURL: cfg.URL + "/forum.php?mod=viewthread&tid=" + id,
+	}
+
+	// Title
+	titleRegex := regexp.MustCompile(`<span[^>]*id="thread_subject"[^>]*>([^<]+)</span>`)
+	if m := titleRegex.FindStringSubmatch(html); len(m) >= 2 {
+		detail.Title = strings.TrimSpace(m[1])
+	}
+
+	// Extract magnet/torrent links
+	magnetRegex := regexp.MustCompile(`magnet:\?[^\s"'<>]+`)
+	if m := magnetRegex.FindString(html); m != "" {
+		detail.DownloadURL = m
+	}
+	torrentRegex := regexp.MustCompile(`(attachment\.php\?aid=\d+)`)
+	if m := torrentRegex.FindString(html); m != "" && detail.DownloadURL == "" {
+		detail.DownloadURL = cfg.URL + "/" + m
+	}
+
+	// Description
+	descRegex := regexp.MustCompile(`<div[^>]*class="t_fsz"[^>]*>(.*?)</div>`)
+	if m := descRegex.FindStringSubmatch(html); len(m) >= 2 {
+		detail.Description = stripHTML(m[1])
+	}
+
+	return detail, nil
+}
+
+func (a *DiscuzAdapter) GetDownloadURL(ctx context.Context, cfg SiteConfig, id string) (string, error) {
+	return cfg.URL + "/forum.php?mod=viewthread&tid=" + id, nil
+}
+
+// parseDiscuzHTML 解析 Discuz HTML 响应。
+func parseDiscuzHTML(html, siteName, baseURL string) (*SiteSearchResult, error) {
+	result := &SiteSearchResult{
+		SiteName: siteName,
+		Items:    []TorrentItem{},
+		Page:     1,
+	}
+
+	// Extract thread links
+	threadRegex := regexp.MustCompile(`<a[^>]*href="(?:forum\.php\?mod=viewthread&tid=|thread-(\d+)-1-1)\.html"[^>]*>([^<]+)</a>`)
+	matches := threadRegex.FindAllStringSubmatch(html, -1)
+
+	for _, m := range matches {
+		item := TorrentItem{}
+		if m[1] != "" {
+			item.ID = m[1]
+		} else {
+			// Extract tid from URL
+			tidRegex := regexp.MustCompile(`tid=(\d+)`)
+			if tidM := tidRegex.FindStringSubmatch(m[0]); len(tidM) >= 2 {
+				item.ID = tidM[1]
+			}
+		}
+		if item.ID == "" {
+			continue
+		}
+
+		item.Title = strings.TrimSpace(m[2])
+		item.DetailURL = baseURL + "/forum.php?mod=viewthread&tid=" + item.ID
+		item.UploadTime = time.Now()
+
+		result.Items = append(result.Items, item)
+	}
+
+	result.Total = len(result.Items)
+	return result, nil
+}