优化

internal/handler/routes_admin.go

@@ -20,6 +20,24 @@ func registerAdminRoutes(api *gin.RouterGroup, cfg *config.Config, svc *service.
 	registerAdminAPIConfigRoutes(admin, svc)
 	registerAdminRecognitionWordRoutes(admin, svc)
 	registerAdminStrmRoutes(admin, svc)
+	registerAdminScraperRoutes(admin, svc)
+}
+
+func registerAdminScraperRoutes(admin *gin.RouterGroup, svc *service.Container) {
+	admin.GET("/scraper/queue", listScrapeQueueHandler(svc))
+	admin.POST("/scraper/queue/:id/cancel", cancelScrapeTaskHandler(svc))
+	admin.POST("/scraper/queue/:id/retry", retryScrapeTaskHandler(svc))
+	admin.DELETE("/scraper/queue/:id", deleteScrapeTaskHandler(svc))
+	admin.POST("/scraper/queue/batch", batchActionScrapeTasksHandler(svc))
+	admin.POST("/scraper/queue/clear-done", clearDoneScrapeTasksHandler(svc))
+	admin.POST("/scraper/queue/clear-finished", clearFinishedScrapeTasksHandler(svc))
+	admin.POST("/scraper/queue/clear-canceled", clearCanceledScrapeTasksHandler(svc))
+	admin.POST("/scraper/queue/retry-failed", retryAllFailedScrapeTasksHandler(svc))
+	admin.POST("/scraper/queue/cancel-pending", cancelPendingScrapeTasksHandler(svc))
+	admin.POST("/scraper/queue/enqueue-library/:id", enqueueLibraryScrapeHandler(svc))
+	admin.POST("/scraper/queue/enqueue-all", enqueueAllScrapeHandler(svc))
+	admin.POST("/media/repair-rescrape", enqueueAllScrapeHandler(svc))
+	admin.POST("/libraries/:id/repair-rescrape", enqueueLibraryScrapeHandler(svc))
 }
 
 func registerAdminStrmRoutes(admin *gin.RouterGroup, svc *service.Container) {

internal/handler/scraper_queue.go

@@ -0,0 +1,167 @@
+package handler
+
+import (
+	"net/http"
+	"strconv"
+
+	"github.com/gin-gonic/gin"
+
+	"github.com/ShukeBta/MMTL/internal/service"
+)
+
+func listScrapeQueueHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		page, _ := strconv.Atoi(c.DefaultQuery("page", "1"))
+		pageSize, _ := strconv.Atoi(c.DefaultQuery("page_size", "50"))
+		snap, err := svc.Scraper.ScrapeQueueSnapshot(c.Request.Context(), c.Query("status"), page, pageSize)
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, snap)
+	}
+}
+
+func cancelScrapeTaskHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		if err := svc.Scraper.CancelScrapeTask(c.Request.Context(), c.Param("id")); err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"ok": true})
+	}
+}
+
+func retryScrapeTaskHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		if err := svc.Scraper.RetryScrapeTask(c.Request.Context(), c.Param("id")); err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"ok": true})
+	}
+}
+
+func deleteScrapeTaskHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		if err := svc.Scraper.DeleteScrapeTask(c.Request.Context(), c.Param("id")); err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"ok": true})
+	}
+}
+
+func batchActionScrapeTasksHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		var req queueBatchReq
+		if err := c.ShouldBindJSON(&req); err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+			return
+		}
+		n, err := svc.Scraper.BatchActionScrapeTasks(c.Request.Context(), req.Action, req.IDs)
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"affected": n, "action": req.Action})
+	}
+}
+
+func clearDoneScrapeTasksHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		n, err := svc.Scraper.ClearDoneScrapeTasks(c.Request.Context())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"deleted": n})
+	}
+}
+
+func clearFinishedScrapeTasksHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		n, err := svc.Scraper.ClearFinishedScrapeTasks(c.Request.Context())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"deleted": n})
+	}
+}
+
+func clearCanceledScrapeTasksHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		n, err := svc.Scraper.ClearCanceledScrapeTasks(c.Request.Context())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"deleted": n})
+	}
+}
+
+func retryAllFailedScrapeTasksHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		n, err := svc.Scraper.RetryAllFailedScrapeTasks(c.Request.Context())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"retried": n})
+	}
+}
+
+func cancelPendingScrapeTasksHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		n, err := svc.Scraper.CancelPendingScrapeTasks(c.Request.Context())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"canceled": n})
+	}
+}
+
+type enqueueScrapeReq struct {
+	EpisodeImages  bool `json:"episode_images"`
+	EpisodeArtwork bool `json:"episode_artwork"`
+	RefreshMatched bool `json:"refresh_matched"`
+	IncludeMatched bool `json:"include_matched"`
+}
+
+func (r enqueueScrapeReq) toOptions() service.ScrapeOptions {
+	epArtwork := r.EpisodeImages || r.EpisodeArtwork
+	return service.ScrapeOptions{
+		EpisodeArtwork: &epArtwork,
+		IncludeMatched: r.IncludeMatched || r.RefreshMatched,
+		RetryNoMatch:   true,
+	}
+}
+
+func enqueueLibraryScrapeHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		var req enqueueScrapeReq
+		_ = c.ShouldBindJSON(&req)
+		libID := c.Param("id")
+		n, err := svc.Scraper.EnqueueLibrary(c.Request.Context(), libID, req.toOptions())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"enqueued": n})
+	}
+}
+
+func enqueueAllScrapeHandler(svc *service.Container) gin.HandlerFunc {
+	return func(c *gin.Context) {
+		var req enqueueScrapeReq
+		_ = c.ShouldBindJSON(&req)
+		n, err := svc.Scraper.EnqueueAll(c.Request.Context(), req.toOptions())
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"enqueued": n})
+	}
+}

internal/handler/streaming.go

@@ -2,7 +2,6 @@
 package handler
 
 import (
-	"context"
 	"errors"
 	"io"
 	"net/http"
@@ -135,28 +134,12 @@ func scrapeOneHandler(svc *service.Container) gin.HandlerFunc {
 			return
 		}
 		options.IncludeMatched = true
-		m, err := svc.Repo.Media.FindByID(c.Request.Context(), c.Param("id"))
-		if err != nil || m == nil {
-			c.JSON(http.StatusNotFound, gin.H{"error": "not found"})
-			return
-		}
-		task := startScrapeHTTPTask(svc, "手动刮削媒体", m.Title, m.Path)
-		if err := svc.Scraper.EnrichOneWithOptions(c.Request.Context(), m, options); err != nil {
-			finishHTTPTask(task, err, "scrape", "手动刮削媒体失败", nil, nil)
+		task, err := svc.Scraper.EnqueueMedia(c.Request.Context(), c.Param("id"), options)
+		if err != nil {
 			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
 			return
 		}
-		reclassified := reclassifyMediaAfterScrape(c.Request.Context(), svc, m.ID)
-		refreshed, _ := svc.Repo.Media.FindByID(c.Request.Context(), m.ID)
-		metrics := map[string]int64{"processed": 1}
-		if refreshed != nil && refreshed.ScrapeStatus == "matched" {
-			metrics["matched"] = 1
-		}
-		if reclassified > 0 {
-			metrics["reclassified"] = int64(reclassified)
-		}
-		finishHTTPTask(task, nil, "completed", "手动刮削媒体结束", metrics, nil)
-		c.JSON(http.StatusOK, refreshed)
+		c.JSON(http.StatusOK, task)
 	}
 }
 
@@ -170,40 +153,12 @@ func scrapeLibraryHandler(svc *service.Container) gin.HandlerFunc {
 			return
 		}
 		options.IncludeMatched = true
-		var task *service.TaskHandle
-		if lib, err := svc.Repo.Library.FindByID(c.Request.Context(), libID); err == nil && lib != nil {
-			task = startScrapeHTTPTask(svc, "手动刮削媒体库", lib.Name, lib.Path)
-		} else {
-			task = startScrapeHTTPTask(svc, "手动刮削媒体库", libID, "")
+		n, err := svc.Scraper.EnqueueLibrary(c.Request.Context(), libID, options)
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
 		}
-		// Run in the background so HTTP returns instantly; the WS hub
-		// pushes per-item progress on the "scrape" topic.
-		go func(libID string, task *service.TaskHandle, options service.ScrapeOptions) {
-			result, err := svc.Scraper.EnrichLibraryDetailedWithOptions(context.Background(), libID, options)
-			reclassified := 0
-			if result.Processed > 0 {
-				reclassified = reclassifyLibraryAfterScrape(context.Background(), svc, libID)
-			}
-			metrics := map[string]int64{
-				"matched":    int64(result.Matched),
-				"processed":  int64(result.Processed),
-				"candidates": int64(result.Candidates),
-			}
-			if reclassified > 0 {
-				metrics["reclassified"] = int64(reclassified)
-			}
-			if result.Failed > 0 {
-				metrics["errors"] = int64(result.Failed)
-			}
-			stage := "completed"
-			message := "手动刮削媒体库结束"
-			if err != nil {
-				stage = "scrape"
-				message = "手动刮削媒体库失败"
-			}
-			finishHTTPTask(task, err, stage, message, metrics, nil)
-		}(libID, task, options)
-		c.JSON(http.StatusAccepted, gin.H{"status": "scraping"})
+		c.JSON(http.StatusOK, gin.H{"status": "queued", "enqueued": n})
 	}
 }