refactor: split modules and harden scraping workflows

2026-09-30 19:46:38 +08:00 · 2026-06-24 11:59:18 +08:00
parent efca3cbe69
commit 192f35d9fa
470 changed files with 47957 additions and 33839 deletions
@@ -21,6 +21,11 @@ const (
 	CtxUserTier     = "ctx_user_tier"
 	CtxTokenPurpose = "ctx_token_purpose"
 	CtxTokenMediaID = "ctx_token_media_id"
+
+	// AccessTokenCookieName carries the web access token for browser-managed
+	// resource requests such as <img>, which cannot attach Authorization.
+	AccessTokenCookieName = "msgo_access_token"
+	AccessTokenCookiePath = "/api"
 )

 // RequestLogger logs one structured line per request.
@@ -199,6 +204,7 @@ func AuthRequired(secret string) gin.HandlerFunc {
 			c.AbortWithStatusJSON(http.StatusForbidden, gin.H{"code": 40304, "message": "token scope denied"})
 			return
 		}
+		syncAccessTokenCookie(c, raw, claims)
 		c.Set(CtxUserID, claims.UserID)
 		c.Set(CtxUserRole, claims.Role)
 		c.Set(CtxUserTier, claims.Tier)
@@ -208,6 +214,48 @@ func AuthRequired(secret string) gin.HandlerFunc {
 	}
 }

+func syncAccessTokenCookie(c *gin.Context, raw string, claims *Claims) {
+	if c == nil || claims == nil || strings.TrimSpace(raw) == "" || strings.TrimSpace(claims.Purpose) != "" {
+		return
+	}
+	if existing, err := c.Cookie(AccessTokenCookieName); err == nil && existing == raw {
+		return
+	}
+	maxAge := int(time.Hour.Seconds())
+	expires := time.Now().Add(time.Hour)
+	if claims.ExpiresAt != nil {
+		expires = claims.ExpiresAt.Time
+		ttl := time.Until(expires)
+		if ttl <= 0 {
+			return
+		}
+		maxAge = int(ttl.Seconds())
+		if maxAge < 1 {
+			maxAge = 1
+		}
+	}
+	http.SetCookie(c.Writer, &http.Cookie{
+		Name:     AccessTokenCookieName,
+		Value:    raw,
+		Path:     AccessTokenCookiePath,
+		MaxAge:   maxAge,
+		Expires:  expires,
+		HttpOnly: true,
+		SameSite: http.SameSiteLaxMode,
+		Secure:   requestIsHTTPS(c),
+	})
+}
+
+func requestIsHTTPS(c *gin.Context) bool {
+	if c == nil || c.Request == nil {
+		return false
+	}
+	if c.Request.TLS != nil {
+		return true
+	}
+	return strings.EqualFold(c.GetHeader("X-Forwarded-Proto"), "https")
+}
+
 func scopedTokenAllowedForRequest(c *gin.Context, claims *Claims) bool {
 	if claims == nil || strings.TrimSpace(claims.Purpose) == "" {
 		return true
@@ -320,5 +368,8 @@ func extractToken(c *gin.Context) string {
 			return value
 		}
 	}
+	if cookie, err := c.Cookie(AccessTokenCookieName); err == nil {
+		return strings.TrimSpace(cookie)
+	}
 	return ""
 }
@@ -4,8 +4,10 @@ import (
 	"net/http"
 	"net/http/httptest"
 	"testing"
+	"time"

 	"github.com/gin-gonic/gin"
+	"github.com/golang-jwt/jwt/v5"
 )

 func TestCORSWildcardOriginAllowsProductionPreflight(t *testing.T) {
@@ -29,3 +31,161 @@ func TestCORSWildcardOriginAllowsProductionPreflight(t *testing.T) {
 		t.Fatalf("Access-Control-Allow-Origin = %q, want *", got)
 	}
 }
+
+func TestAuthRequiredAcceptsAccessTokenCookie(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	secret := "cookie-secret"
+	token := signedMiddlewareTestToken(t, secret, Claims{
+		UserID: "user-1",
+		Role:   "admin",
+		RegisteredClaims: jwt.RegisteredClaims{
+			ExpiresAt: jwt.NewNumericDate(time.Now().Add(time.Hour)),
+		},
+	})
+
+	router := gin.New()
+	router.Use(AuthRequired(secret))
+	router.GET("/api/img", func(c *gin.Context) {
+		c.JSON(http.StatusOK, gin.H{"user_id": c.GetString(CtxUserID)})
+	})
+
+	req := httptest.NewRequest(http.MethodGet, "/api/img?url=https%3A%2F%2Fexample.test%2Fposter.jpg", nil)
+	req.AddCookie(&http.Cookie{Name: AccessTokenCookieName, Value: token})
+	w := httptest.NewRecorder()
+	router.ServeHTTP(w, req)
+
+	if w.Code != http.StatusOK {
+		t.Fatalf("status = %d body=%s", w.Code, w.Body.String())
+	}
+}
+
+func TestAuthRequiredKeepsExplicitQueryTokenPriority(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	secret := "cookie-secret"
+	accountToken := signedMiddlewareTestToken(t, secret, Claims{
+		UserID: "user-1",
+		Role:   "admin",
+		RegisteredClaims: jwt.RegisteredClaims{
+			ExpiresAt: jwt.NewNumericDate(time.Now().Add(time.Hour)),
+		},
+	})
+	scopedToken := signedMiddlewareTestToken(t, secret, Claims{
+		UserID:  "user-1",
+		Role:    "admin",
+		Purpose: "external_play",
+		MediaID: "media-1",
+		RegisteredClaims: jwt.RegisteredClaims{
+			ExpiresAt: jwt.NewNumericDate(time.Now().Add(time.Hour)),
+		},
+	})
+
+	router := gin.New()
+	router.Use(AuthRequired(secret))
+	router.GET("/api/me", func(c *gin.Context) {
+		c.JSON(http.StatusOK, gin.H{"ok": true})
+	})
+
+	req := httptest.NewRequest(http.MethodGet, "/api/me?token="+scopedToken, nil)
+	req.AddCookie(&http.Cookie{Name: AccessTokenCookieName, Value: accountToken})
+	w := httptest.NewRecorder()
+	router.ServeHTTP(w, req)
+
+	if w.Code != http.StatusForbidden {
+		t.Fatalf("status = %d body=%s", w.Code, w.Body.String())
+	}
+}
+
+func TestAuthRequiredSyncsAccessTokenCookieFromBearer(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	secret := "cookie-secret"
+	token := signedMiddlewareTestToken(t, secret, Claims{
+		UserID: "user-1",
+		Role:   "admin",
+		RegisteredClaims: jwt.RegisteredClaims{
+			ExpiresAt: jwt.NewNumericDate(time.Now().Add(time.Hour)),
+		},
+	})
+
+	router := gin.New()
+	router.Use(AuthRequired(secret))
+	router.GET("/api/discover/feed", func(c *gin.Context) {
+		c.JSON(http.StatusOK, gin.H{"ok": true})
+	})
+
+	req := httptest.NewRequest(http.MethodGet, "/api/discover/feed", nil)
+	req.Header.Set("Authorization", "Bearer "+token)
+	w := httptest.NewRecorder()
+	router.ServeHTTP(w, req)
+
+	if w.Code != http.StatusOK {
+		t.Fatalf("status = %d body=%s", w.Code, w.Body.String())
+	}
+	cookie := middlewareTestResponseCookie(t, w, AccessTokenCookieName)
+	if cookie.Value != token {
+		t.Fatal("synced cookie should contain the bearer token")
+	}
+	if cookie.Path != AccessTokenCookiePath {
+		t.Fatalf("cookie path = %q, want %q", cookie.Path, AccessTokenCookiePath)
+	}
+	if !cookie.HttpOnly || cookie.SameSite != http.SameSiteLaxMode {
+		t.Fatalf("cookie flags not suitable: httpOnly=%v sameSite=%v", cookie.HttpOnly, cookie.SameSite)
+	}
+}
+
+func TestAuthRequiredDoesNotSyncScopedPlaybackTokenCookie(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	secret := "cookie-secret"
+	token := signedMiddlewareTestToken(t, secret, Claims{
+		UserID:  "user-1",
+		Role:    "admin",
+		Purpose: "external_play",
+		MediaID: "media-1",
+		RegisteredClaims: jwt.RegisteredClaims{
+			ExpiresAt: jwt.NewNumericDate(time.Now().Add(time.Hour)),
+		},
+	})
+
+	router := gin.New()
+	router.Use(AuthRequired(secret))
+	router.GET("/api/stream/media-1", func(c *gin.Context) {
+		c.Status(http.StatusNoContent)
+	})
+
+	req := httptest.NewRequest(http.MethodGet, "/api/stream/media-1?token="+token, nil)
+	w := httptest.NewRecorder()
+	router.ServeHTTP(w, req)
+
+	if w.Code != http.StatusNoContent {
+		t.Fatalf("status = %d body=%s", w.Code, w.Body.String())
+	}
+	if cookie := optionalMiddlewareTestResponseCookie(w, AccessTokenCookieName); cookie != nil {
+		t.Fatalf("scoped playback token should not be synced as web cookie: %#v", cookie)
+	}
+}
+
+func signedMiddlewareTestToken(t *testing.T, secret string, claims Claims) string {
+	t.Helper()
+	token, err := jwt.NewWithClaims(jwt.SigningMethodHS256, claims).SignedString([]byte(secret))
+	if err != nil {
+		t.Fatalf("sign token: %v", err)
+	}
+	return token
+}
+
+func middlewareTestResponseCookie(t *testing.T, w *httptest.ResponseRecorder, name string) *http.Cookie {
+	t.Helper()
+	cookie := optionalMiddlewareTestResponseCookie(w, name)
+	if cookie == nil {
+		t.Fatalf("missing response cookie %q", name)
+	}
+	return cookie
+}
+
+func optionalMiddlewareTestResponseCookie(w *httptest.ResponseRecorder, name string) *http.Cookie {
+	for _, cookie := range w.Result().Cookies() {
+		if cookie.Name == name {
+			return cookie
+		}
+	}
+	return nil
+}