web_hgcg_server/web/skd/job/seopush/seopush.go

// Package seopush 为搜索引擎推送做兜底：补推已上架短剧，并重试失败的任务。
package seopush

import (
	"context"
	"strings"
	"sync"
	"time"
	"videoweb/pub/common/log"
	"videoweb/pub/models/playletmod"
	"videoweb/pub/models/seopushmod"
	"videoweb/pub/pubdef"
	"videoweb/web/service/srv_seopush"

	"go.mongodb.org/mongo-driver/bson"
	"go.mongodb.org/mongo-driver/mongo/options"
)

const (
	backfillLimit = 100
	retryLimit    = 50
	taskKeepDays  = 30
)

var jobMu sync.Mutex

// Run 定时兜底任务
func Run() {
	if !jobMu.TryLock() {
		log.Warn("seopush: skip job, previous run still in progress")
		return
	}
	defer jobMu.Unlock()

	ctx := context.Background()
	if !pubdef.Enabled() {
		discardPendingTasks(ctx)
		return
	}
	webAddr := srv_seopush.SEOBaseURL(ctx)
	if webAddr == "" {
		return
	}
	backfillVisiblePlaylets(ctx)
	retryPendingTasks(ctx, webAddr)
	cleanupExpiredTasks(ctx)
}

func discardPendingTasks(ctx context.Context) {
	filter := bson.M{"$or": []bson.M{
		{seopushmod.FN.SeoPushTask.GoogleStatus: seopushmod.PushStatusPending},
		{seopushmod.FN.SeoPushTask.IndexNowStatus: seopushmod.PushStatusPending},
	}}
	list, serr := seopushmod.FindManySeoPushTasks(ctx, filter)
	if serr != nil {
		log.Warn("seopush: find pending tasks to discard failed", log.E(serr))
		return
	}
	if len(list) == 0 {
		return
	}
	ids := make([]int64, 0, len(list))
	seen := make(map[int64]struct{}, len(list))
	for _, task := range list {
		if task == nil || task.VideoId <= 0 {
			continue
		}
		if _, ok := seen[task.VideoId]; ok {
			continue
		}
		seen[task.VideoId] = struct{}{}
		ids = append(ids, task.VideoId)
	}
	count, serr := seopushmod.DelManySeoPushTask(ctx, filter)
	if serr != nil {
		log.Warn("seopush: discard pending tasks failed", log.E(serr))
		return
	}
	if len(ids) > 0 {
		if serr := playletmod.UpdateManyPlaylet(ctx,
			bson.M{playletmod.FN.Playlet.Id: bson.M{"$in": ids}},
			bson.M{"$set": bson.M{playletmod.FN.Playlet.IndexNow: int64(0)}}); serr != nil {
			log.Warn("seopush: reset playlet indexNow failed", log.E(serr))
		}
	}
	log.Info("seopush=skip=disabled", log.Any("discarded", count))
}

func backfillVisiblePlaylets(ctx context.Context) {
	filter := bson.M{
		playletmod.FN.Playlet.Status: playletmod.StatusUp,
		"$or": []bson.M{
			{playletmod.FN.Playlet.IndexNow: 0},
			{playletmod.FN.Playlet.IndexNow: bson.M{"$exists": false}},
		},
	}
	opt := options.Find().
		SetSort(bson.M{playletmod.FN.Playlet.OnlineAt: -1}).
		SetLimit(backfillLimit)

	list, serr := playletmod.FindManyPlaylet(ctx, filter, opt)
	if serr != nil {
		log.Warn("seopush: backfill find playlets failed", log.E(serr))
		return
	}
	if len(list) == 0 {
		return
	}
	for _, item := range list {
		if item == nil {
			continue
		}
		srv_seopush.RegisterPlayletUpdated(ctx, item.Id, item.Status, item.OnlineAt)
	}
	log.Info("seopush: backfill registered", log.Any("count", len(list)))
}

func retryPendingTasks(ctx context.Context, webAddr string) {
	conds := []bson.M{
		{
			seopushmod.FN.SeoPushTask.GoogleStatus: seopushmod.PushStatusPending,
			seopushmod.FN.SeoPushTask.GoogleRetry:  bson.M{"$lt": seopushmod.MaxPushRetry},
		},
		{
			seopushmod.FN.SeoPushTask.IndexNowStatus: seopushmod.PushStatusPending,
			seopushmod.FN.SeoPushTask.IndexNowRetry:  bson.M{"$lt": seopushmod.MaxPushRetry},
		},
	}
	opt := options.Find().
		SetSort(bson.M{seopushmod.FN.SeoPushTask.CreatedAt: 1}).
		SetLimit(retryLimit)
	taskList, serr := seopushmod.FindManySeoPushTasks(ctx, bson.M{"$or": conds}, opt)
	if serr != nil {
		log.Warn("seopush: find pending tasks failed", log.E(serr))
		return
	}
	if len(taskList) == 0 {
		return
	}
	for _, task := range taskList {
		if task == nil || strings.Contains(task.Url, "?") {
			continue
		}
		srv_seopush.PushTask(ctx, webAddr, task)
	}
	log.Info("seopush: pending tasks processed", log.Any("count", len(taskList)))
}

func cleanupExpiredTasks(ctx context.Context) {
	deadline := time.Now().AddDate(0, 0, -taskKeepDays)
	count, serr := seopushmod.DelManySeoPushTask(ctx, bson.M{
		seopushmod.FN.SeoPushTask.UpdatedAt: bson.M{"$lt": deadline},
	})
	if serr != nil {
		log.Warn("seopush: cleanup expired tasks failed", log.E(serr))
		return
	}
	if count > 0 {
		log.Info("seopush: expired tasks cleaned", log.Any("count", count))
	}
}