web_hgdjc_server/web/service/srv_seopush/push.go

// Package srv_seopush 负责短剧新上架/下架时向搜索引擎推送页面地址。
// 业务回调里即时推送一次，失败的任务由 web/skd/job/seopush 兜底重试。
package srv_seopush

import (
	"context"
	"errors"
	"strings"
	"time"
	"videoweb/pub/common/godef"
	"videoweb/pub/common/log"
	"videoweb/pub/models/playletmod"
	"videoweb/pub/models/seopushmod"
	"videoweb/pub/models/videoinformationmod"
	"videoweb/pub/pubdef"
	"videoweb/pub/seo/googleindex"
	"videoweb/pub/seo/h5path"
	seoindexnow "videoweb/pub/seo/indexnow"
	"videoweb/web/service/srv_cfg"
	"videoweb/web/webg"

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

// channelFields 把两个推送通道的字段名抽象出来，共用同一套状态更新逻辑
type channelFields struct {
	name   string
	status string
	retry  string
	errMsg string
	at     string
}

var googleChannel = channelFields{
	name:   "google",
	status: seopushmod.FN.SeoPushTask.GoogleStatus,
	retry:  seopushmod.FN.SeoPushTask.GoogleRetry,
	errMsg: seopushmod.FN.SeoPushTask.GoogleErr,
	at:     seopushmod.FN.SeoPushTask.GoogleAt,
}

var indexNowChannel = channelFields{
	name:   "indexnow",
	status: seopushmod.FN.SeoPushTask.IndexNowStatus,
	retry:  seopushmod.FN.SeoPushTask.IndexNowRetry,
	errMsg: seopushmod.FN.SeoPushTask.IndexNowErr,
	at:     seopushmod.FN.SeoPushTask.IndexNowAt,
}

// PlayletDetailURL 拼装短剧详情页完整地址，与 sitemap 一致：/detail/{slug}/，无 slug 回退 ID。
func PlayletDetailURL(webAddr, slug string, playletId int64) string {
	if h5path.Segment(slug, playletId) == "" {
		return ""
	}
	return h5path.SitemapAbs(webAddr, h5path.Entity("/detail", slug, playletId))
}

// VideoPlayURL 拼装视频播放页完整地址，路径与 H5 / sitemap 一致：/series-play/{slug}。
func VideoPlayURL(webAddr, slug string, videoId int64) string {
	path := h5path.VideoPlay(slug, videoId)
	if path == "" {
		return ""
	}
	return h5path.Abs(webAddr, path)
}

// PageVisible 判断详情页当前是否对外可访问。
func PageVisible(status playletmod.Status, onlineAt time.Time) bool {
	if status != playletmod.StatusUp {
		return false
	}
	if onlineAt.IsZero() {
		return true
	}
	return !onlineAt.After(time.Now())
}

// SEOBaseURL 取站点对外地址
func SEOBaseURL(ctx context.Context) string {
	sysCfg, serr := srv_cfg.GetSystemConfig(ctx)
	if serr != nil || sysCfg == nil {
		log.Warn("seopush: get system config failed", log.E(serr))
		return ""
	}
	base := sysCfg.SEOBaseURL()
	if base == "" {
		log.Warn("seopush: SEO base URL is empty, skip push")
	}
	return base
}

func canPushExternally() bool {
	return pubdef.Enabled() && webg.Conf.RunMod == "prod"
}

// EnqueuePlayletUpdated 短剧新上架后登记并立即推送，仅在页面已可访问时生效
func EnqueuePlayletUpdated(ctx context.Context, playletId int64,
	status playletmod.Status, onlineAt time.Time) {
	if !pubdef.Enabled() || !PageVisible(status, onlineAt) {
		return
	}
	enqueue(ctx, playletId, seopushmod.ActionUpdated, true)
}

// RegisterPlayletUpdated 只登记不立即推送，供批量补推使用
func RegisterPlayletUpdated(ctx context.Context, playletId int64,
	status playletmod.Status, onlineAt time.Time) {
	if !pubdef.Enabled() || !PageVisible(status, onlineAt) {
		return
	}
	enqueue(ctx, playletId, seopushmod.ActionUpdated, false)
}

// EnqueuePlayletDeleted 短剧下架/删除后登记并立即推送
func EnqueuePlayletDeleted(ctx context.Context, playletId int64) {
	if !pubdef.Enabled() {
		return
	}
	enqueue(ctx, playletId, seopushmod.ActionDeleted, true)
}

// VideoPageVisible 判断视频播放页当前是否对外可访问。
func VideoPageVisible(status videoinformationmod.InformationSatus) bool {
	return status == videoinformationmod.InformationSatusVerifyOK
}

// EnqueueVideoUpdated 视频上架后登记并立即推送。
func EnqueueVideoUpdated(ctx context.Context, videoId int64, status videoinformationmod.InformationSatus) {
	if !pubdef.Enabled() || !VideoPageVisible(status) {
		return
	}
	enqueueVideo(ctx, videoId, seopushmod.ActionUpdated, true)
}

// EnqueueVideoDeleted 视频下架/删除后登记并立即推送。
func EnqueueVideoDeleted(ctx context.Context, videoId int64) {
	if !pubdef.Enabled() {
		return
	}
	enqueueVideo(ctx, videoId, seopushmod.ActionDeleted, true)
}

func enqueue(ctx context.Context, playletId int64, action string, pushNow bool) {
	webAddr := SEOBaseURL(ctx)
	if webAddr == "" {
		return
	}
	pageURL := PlayletDetailURL(webAddr, loadPlayletSlug(ctx, playletId), playletId)
	if pageURL == "" {
		return
	}
	now := time.Now()
	task := &seopushmod.SeoPushTask{
		Url:       pageURL,
		VideoId:   playletId,
		Action:    action,
		CreatedAt: now,
		UpdatedAt: now,
	}

	if action == seopushmod.ActionDeleted {
		_, _ = seopushmod.DelManySeoPushTask(ctx, bson.M{
			seopushmod.FN.SeoPushTask.Url:    pageURL,
			seopushmod.FN.SeoPushTask.Action: seopushmod.ActionUpdated,
		})
	}
	if serr := seopushmod.UpsertPendingSeoPushTask(ctx, task); serr != nil {
		log.Warn("seopush: enqueue task failed", log.Any("url", pageURL), log.E(serr))
		return
	}
	if action == seopushmod.ActionUpdated {
		markPlayletQueued(ctx, playletId)
	}
	if !pushNow {
		return
	}

	godef.Go(func() {
		PushTask(context.Background(), webAddr, task)
	})
}

func enqueueVideo(ctx context.Context, videoId int64, action string, pushNow bool) {
	webAddr := SEOBaseURL(ctx)
	if webAddr == "" {
		return
	}
	pageURL := VideoPlayURL(webAddr, loadVideoSlug(ctx, videoId), videoId)
	if pageURL == "" {
		return
	}
	now := time.Now()
	task := &seopushmod.SeoPushTask{
		Url:       pageURL,
		VideoId:   videoId,
		Action:    action,
		CreatedAt: now,
		UpdatedAt: now,
	}
	if action == seopushmod.ActionDeleted {
		_, _ = seopushmod.DelManySeoPushTask(ctx, bson.M{
			seopushmod.FN.SeoPushTask.Url:    pageURL,
			seopushmod.FN.SeoPushTask.Action: seopushmod.ActionUpdated,
		})
	}
	if serr := seopushmod.UpsertPendingSeoPushTask(ctx, task); serr != nil {
		log.Warn("seopush: enqueue video task failed", log.Any("url", pageURL), log.E(serr))
		return
	}
	if action == seopushmod.ActionUpdated {
		markVideoQueued(ctx, videoId)
	}
	if !pushNow {
		return
	}
	godef.Go(func() {
		PushTask(context.Background(), webAddr, task)
	})
}

func loadPlayletSlug(ctx context.Context, playletId int64) string {
	if playletId <= 0 {
		return ""
	}
	item, err := playletmod.FindOnePlaylet(ctx, bson.M{playletmod.FN.Playlet.Id: playletId},
		options.FindOne().SetProjection(bson.D{{Key: playletmod.FN.Playlet.Slug, Value: 1}}))
	if err != nil || item == nil {
		return ""
	}
	return strings.TrimSpace(item.Slug)
}

func loadVideoSlug(ctx context.Context, videoId int64) string {
	if videoId <= 0 {
		return ""
	}
	item, err := videoinformationmod.FindOneVideoInformationWithOptions(ctx,
		bson.M{videoinformationmod.FN.VideoInformation.Id: videoId},
		options.FindOne().SetProjection(bson.D{{Key: videoinformationmod.FN.VideoInformation.Slug, Value: 1}}))
	if err != nil || item == nil {
		return ""
	}
	return strings.TrimSpace(item.Slug)
}

func isObsoletePushURL(raw string) bool {
	raw = strings.TrimSpace(raw)
	if raw == "" {
		return true
	}
	if strings.Contains(raw, "?") {
		return true
	}
	if strings.Contains(raw, "/lieyan/av/detail") {
		return true
	}
	// 旧推送路径 /play/{slug}。/series-play 是视频页，/detail/{slug}/ 是当前详情页。
	path := requestPath(raw)
	return path == "/play" || strings.HasPrefix(path, "/play/")
}

func requestPath(raw string) string {
	raw = strings.TrimSpace(raw)
	if i := strings.IndexAny(raw, "?#"); i >= 0 {
		raw = raw[:i]
	}
	if i := strings.Index(raw, "://"); i >= 0 {
		raw = raw[i+3:]
		if j := strings.Index(raw, "/"); j >= 0 {
			return raw[j:]
		}
		return "/"
	}
	if raw == "" {
		return "/"
	}
	if !strings.HasPrefix(raw, "/") {
		return "/" + raw
	}
	return raw
}

func markPlayletQueued(ctx context.Context, playletId int64) {
	if serr := playletmod.UpdateOnePlaylet(ctx,
		bson.M{playletmod.FN.Playlet.Id: playletId},
		bson.M{"$set": bson.M{playletmod.FN.Playlet.IndexNow: int64(1)}}); serr != nil {
		log.Warn("seopush: mark playlet queued failed", log.Any("playletId", playletId), log.E(serr))
	}
}

func markVideoQueued(ctx context.Context, videoId int64) {
	if serr := videoinformationmod.UpdateOneVideoInformation(ctx,
		bson.M{videoinformationmod.FN.VideoInformation.Id: videoId},
		bson.M{"$set": bson.M{videoinformationmod.FN.VideoInformation.IndexNow: int64(1)}}); serr != nil {
		log.Warn("seopush: mark video queued failed", log.Any("videoId", videoId), log.E(serr))
	}
}

// PushTask 推送任务中尚未成功的通道
func PushTask(ctx context.Context, webAddr string, task *seopushmod.SeoPushTask) {
	if task == nil || !pubdef.Enabled() {
		return
	}
	// 旧路径不再推送：查询参数、/play/{slug}、/lieyan/av/detail。
	if isObsoletePushURL(task.Url) {
		log.Warn("seopush: drop obsolete-url task", log.Any("url", task.Url), log.Any("action", task.Action))
		_, _ = seopushmod.DelManySeoPushTask(ctx, taskFilter(task))
		return
	}
	if !canPushExternally() {
		dryRun(ctx, task)
		return
	}
	if task.GoogleStatus != seopushmod.PushStatusSuccess {
		pushGoogle(ctx, task)
	}
	if task.IndexNowStatus != seopushmod.PushStatusSuccess {
		pushIndexNow(ctx, webAddr, task)
	}
}

func dryRun(ctx context.Context, task *seopushmod.SeoPushTask) {
	log.Info("seopush: dry-run, external push skipped in non-prod",
		log.Any("url", task.Url), log.Any("action", task.Action))
	if task.GoogleStatus != seopushmod.PushStatusSuccess {
		markSuccess(ctx, task, googleChannel)
	}
	if task.IndexNowStatus != seopushmod.PushStatusSuccess {
		markSuccess(ctx, task, indexNowChannel)
	}
}

func pushGoogle(ctx context.Context, task *seopushmod.SeoPushTask) {
	notifyType := googleindex.TypeUpdated
	if task.Action == seopushmod.ActionDeleted {
		notifyType = googleindex.TypeDeleted
	}

	resp, err := googleindex.Publish(ctx, task.Url, notifyType)
	if err == nil {
		markSuccess(ctx, task, googleChannel)
		log.Info("seopush: google pushed",
			log.Any("url", task.Url),
			log.Any("type", notifyType),
			log.Any("resp", resp))
		return
	}
	if errors.Is(err, googleindex.ErrQuotaExceeded) {
		log.Warn("seopush: google daily quota exceeded, keep pending",
			log.Any("url", task.Url), log.Any("resp", err.Error()))
		return
	}
	if errors.Is(err, googleindex.ErrNotEnabled) {
		log.Warn("seopush: google client not ready, keep pending", log.Any("url", task.Url))
		return
	}
	markFailed(ctx, task, googleChannel, err)
}

func pushIndexNow(ctx context.Context, webAddr string, task *seopushmod.SeoPushTask) {
	if strings.TrimSpace(webAddr) == "" {
		return
	}
	config, err := seoindexnow.BuildConfig(webAddr)
	if err != nil {
		log.Warn("seopush: build indexnow config failed", log.Any("webAddr", webAddr), log.E(err))
		return
	}

	result := seoindexnow.NewBatchIndexNowSubmitter(config).SubmitBatch([]string{task.Url})
	if result.SuccessCount > 0 {
		markSuccess(ctx, task, indexNowChannel)
		log.Info("seopush: indexnow pushed",
			log.Any("url", task.Url),
			log.Any("statusCode", result.StatusCodes))
		return
	}
	markFailed(ctx, task, indexNowChannel, errors.New(strings.Join(result.Errors, "; ")))
}

func markSuccess(ctx context.Context, task *seopushmod.SeoPushTask, ch channelFields) {
	now := time.Now()
	if serr := seopushmod.UpdateOneSeoPushTask(ctx, taskFilter(task), bson.M{
		"$set": bson.M{
			ch.status:                           seopushmod.PushStatusSuccess,
			ch.errMsg:                           "",
			ch.at:                               now,
			seopushmod.FN.SeoPushTask.UpdatedAt: now,
		},
	}); serr != nil {
		log.Warn("seopush: mark success failed",
			log.Any("channel", ch.name), log.Any("url", task.Url), log.E(serr))
	}
}

func markFailed(ctx context.Context, task *seopushmod.SeoPushTask, ch channelFields, pushErr error) {
	now := time.Now()
	errText := ""
	if pushErr != nil {
		errText = pushErr.Error()
	}
	log.Warn("seopush: push failed",
		log.Any("channel", ch.name), log.Any("url", task.Url), log.Any("err", errText))

	if serr := seopushmod.UpdateOneSeoPushTask(ctx, taskFilter(task), bson.M{
		"$set": bson.M{
			ch.errMsg:                           errText,
			ch.at:                               now,
			seopushmod.FN.SeoPushTask.UpdatedAt: now,
		},
		"$inc": bson.M{ch.retry: 1},
	}); serr != nil {
		log.Warn("seopush: mark failed failed",
			log.Any("channel", ch.name), log.Any("url", task.Url), log.E(serr))
	}
}

func taskFilter(task *seopushmod.SeoPushTask) bson.M {
	return bson.M{
		seopushmod.FN.SeoPushTask.Url:    task.Url,
		seopushmod.FN.SeoPushTask.Action: task.Action,
	}
}