improvement(sns): 精简朋友圈媒体链路并移除手动挑图入口

- 移除 Server-Timing/X-SNS 阶段透出与前端媒体来源标记 - 移除本地候选图与手动匹配接口及预览交互 - /api/sns/media 默认仅保留远程下载解密路径，未命中直接返回 404 - 将导出所需的本地缓存匹配逻辑下沉到 sns_export_service
2026-06-18 15:54:08 +08:00 · 2026-04-16 01:19:09 +08:00
parent 94a8b124c3
commit 7870cd54fc
8 changed files with 343 additions and 1267 deletions
@@ -357,30 +357,6 @@ export const useApi = () => {
    return await request(url)
  }

-  // 朋友圈图片本地缓存候选（用于错图时手动选择）
-  const listSnsMediaCandidates = async (params = {}) => {
-    const query = new URLSearchParams()
-    if (params && params.account) query.set('account', params.account)
-    if (params && params.create_time != null) query.set('create_time', String(params.create_time))
-    if (params && params.width != null) query.set('width', String(params.width))
-    if (params && params.height != null) query.set('height', String(params.height))
-    if (params && params.limit != null) query.set('limit', String(params.limit))
-    if (params && params.offset != null) query.set('offset', String(params.offset))
-    const url = '/sns/media_candidates' + (query.toString() ? `?${query.toString()}` : '')
-    return await request(url)
-  }
-
-  // 保存朋友圈图片手动匹配结果（本机）
-  const saveSnsMediaPicks = async (data = {}) => {
-    return await request('/sns/media_picks', {
-      method: 'POST',
-      body: {
-        account: data.account || null,
-        picks: (data && data.picks && typeof data.picks === 'object' && !Array.isArray(data.picks)) ? data.picks : {}
-      }
-    })
-  }
-
  const openChatMediaFolder = async (params = {}) => {
    const query = new URLSearchParams()
    if (params && params.account) query.set('account', params.account)
@@ -667,8 +643,6 @@ export const useApi = () => {
    resolveAppMsg,
    listSnsTimeline,
    listSnsUsers,
-    listSnsMediaCandidates,
-    saveSnsMediaPicks,
    openChatMediaFolder,
    downloadChatEmoji,
    saveMediaKeys,
@@ -114,20 +114,7 @@
                    :src="getSnsMediaUrl(activeCover, activeCover.media[0], 0, activeCover.media[0].url)"
                    class="w-full h-full object-cover"
                    alt="朋友圈封面"
-                    @load="onCoverMediaLoaded(activeCover, $event)"
                />
-                <div
-                    v-if="snsMediaStageLabel(snsCoverStageKey(activeCover)) || snsMediaStageLoading[snsCoverStageKey(activeCover)]"
-                    class="absolute top-3 left-3 z-20 opacity-0 group-hover:opacity-100 transition-opacity pointer-events-none"
-                >
-                  <div
-                      class="text-[10px] px-2 py-0.5 rounded backdrop-blur-sm shadow-sm"
-                      :class="snsMediaStageBadgeColorClass(snsCoverStageKey(activeCover))"
-                      :title="snsMediaStageBadgeTitle(snsCoverStageKey(activeCover))"
-                  >
-                    {{ snsMediaStageLabel(snsCoverStageKey(activeCover)) || '识别中' }}
-                  </div>
-                </div>

                <div
                    v-if="(activeCover && Number(activeCover.createTime || 0)) || (covers && covers.length > 1)"
@@ -347,7 +334,7 @@
                          loop
                          muted
                          playsinline
-                          @loadeddata="onLocalVideoLoaded(post.id, post.media[0].id); onSnsMediaLoaded(post, post.media[0], 0)"
+                          @loadeddata="onLocalVideoLoaded(post.id, post.media[0].id)"
                          @error="onLocalVideoError(post.id, post.media[0].id)"
                      ></video>

@@ -361,7 +348,6 @@
                          loop
                          :muted="livePhotoHoverMuted"
                          playsinline
-                          @loadeddata="onSnsMediaLoaded(post, post.media[0], 0)"
                          @error="onLivePhotoVideoError(post.id, 0)"
                      ></video>

@@ -372,22 +358,8 @@
                          alt=""
                          loading="lazy"
                          referrerpolicy="no-referrer"
-                          @load="onSnsMediaLoaded(post, post.media[0], 0, $event)"
                          @error="onMediaError(post.id, 0)"
                      />
-
-                      <div
-                          v-if="snsMediaStageLabel(snsMediaStageKey(post.id, 0, 'thumb')) || snsMediaStageLoading[snsMediaStageKey(post.id, 0, 'thumb')]"
-                          class="absolute top-2 left-2 z-20 opacity-0 group-hover:opacity-100 transition-opacity pointer-events-none"
-                      >
-                        <div
-                            class="text-[10px] px-2 py-0.5 rounded backdrop-blur-sm shadow-sm"
-                            :class="snsMediaStageBadgeColorClass(snsMediaStageKey(post.id, 0, 'thumb'))"
-                            :title="snsMediaStageBadgeTitle(snsMediaStageKey(post.id, 0, 'thumb'))"
-                        >
-                          {{ snsMediaStageLabel(snsMediaStageKey(post.id, 0, 'thumb')) || '识别中' }}
-                        </div>
-                      </div>
                      <div
                          v-if="Number(post.media[0]?.type || 0) === 6 && !isLocalVideoLoaded(post.id, post.media[0].id)"
                          class="absolute inset-0 flex items-center justify-center pointer-events-none"
@@ -451,7 +423,7 @@
                          loop
                          muted
                          playsinline
-                          @loadeddata="onLocalVideoLoaded(post.id, m.id); onSnsMediaLoaded(post, m, idx)"
+                          @loadeddata="onLocalVideoLoaded(post.id, m.id)"
                          @error="onLocalVideoError(post.id, m.id)"
                      ></video>
                      <video
@@ -464,7 +436,6 @@
                          loop
                          :muted="livePhotoHoverMuted"
                          playsinline
-                          @loadeddata="onSnsMediaLoaded(post, m, idx)"
                          @error="onLivePhotoVideoError(post.id, idx)"
                      ></video>
                      <img
@@ -474,22 +445,8 @@
                          alt=""
                          loading="lazy"
                          referrerpolicy="no-referrer"
-                          @load="onSnsMediaLoaded(post, m, idx, $event)"
                          @error="onMediaError(post.id, idx)"
                      />
-
-                      <div
-                          v-if="snsMediaStageLabel(snsMediaStageKey(post.id, idx, 'thumb')) || snsMediaStageLoading[snsMediaStageKey(post.id, idx, 'thumb')]"
-                          class="absolute top-1 left-1 z-20 opacity-0 group-hover:opacity-100 transition-opacity pointer-events-none"
-                      >
-                        <div
-                            class="text-[10px] px-1.5 py-0.5 rounded backdrop-blur-sm shadow-sm"
-                            :class="snsMediaStageBadgeColorClass(snsMediaStageKey(post.id, idx, 'thumb'))"
-                            :title="snsMediaStageBadgeTitle(snsMediaStageKey(post.id, idx, 'thumb'))"
-                        >
-                          {{ snsMediaStageLabel(snsMediaStageKey(post.id, idx, 'thumb')) || '识别中' }}
-                        </div>
-                      </div>
                      <!-- 不知道微信朋友圈可不可以发多视频，先这样写吧-->
                      <span v-else class="text-[10px] text-gray-400">图片失败</span>

@@ -630,7 +587,7 @@
      </button>
    </div>

-	    <!-- 图片预览弹窗 + 候选匹配选择 -->
+	    <!-- 图片预览弹窗 -->
 	    <div
 	      v-if="previewCtx"
 	      class="fixed inset-0 z-[60] bg-black/90 flex items-center justify-center"
@@ -711,17 +668,6 @@ import { reportServerErrorFromError } from '~/lib/server-error-logging'

 useHead({ title: '朋友圈 - 微信数据分析助手' })

-// Nuxt dev mode can load hundreds of module resources, quickly filling the default
-// ResourceTiming buffer (150). If it overflows, `<img>` requests may not produce
-// entries, making Server-Timing based stage detection always fall back to "unknown".
-if (process.client) {
-  try {
-    if (typeof performance !== 'undefined' && performance?.setResourceTimingBufferSize) {
-      performance.setResourceTimingBufferSize(5000)
-    }
-  } catch {}
-}
-
 const api = useApi()

 const chatAccounts = useChatAccountsStore()
@@ -937,186 +883,6 @@ const onMediaError = (postId, idx) => {
  mediaErrors.value[mediaErrorKey(postId, idx)] = true
 }

-// Hover badge: show which SNS media pipeline stage produced the image.
-// Backend provides `X-SNS-Source` (and optional `X-SNS-Hit-Type`, `X-SNS-X-Enc`) on `/api/sns/media` responses.
-const snsMediaStage = ref({}) // stageKey -> { source, hitType, xEnc }
-const snsMediaStageLoading = ref({}) // stageKey -> boolean
-const snsMediaStageInFlight = new Set()
-
-const isSnsMediaApiUrl = (url) => {
-  const u = String(url || '').trim()
-  return !!u && u.includes('/api/sns/media')
-}
-
-const snsMediaStageKey = (postId, idx, kind = 'thumb') => {
-  const acc = String(selectedAccount.value || '').trim()
-  const pid = String(postId || '').trim()
-  return `sns:${acc}:${pid}:${String(Number(idx) || 0)}:${String(kind || 'thumb')}`
-}
-
-const snsCoverStageKey = (cover) => {
-  const acc = String(selectedAccount.value || '').trim()
-  const cid = String(cover?.id || cover?.tid || cover?.createTime || '').trim()
-  return `sns:${acc}:cover:${cid || '0'}`
-}
-
-const snsMediaStageLabel = (key) => {
-  const k = String(key || '').trim()
-  if (!k) return ''
-  const info = snsMediaStage.value[k]
-  if (!info || typeof info !== 'object') return ''
-
-  const source = String(info?.source || '').trim()
-  const hitType = String(info?.hitType || '').trim()
-
-  if (source === 'remote-cache') return '远程缓存'
-  if (source === 'remote-decrypt') return '远程解密'
-  if (source === 'remote') return '远程直出'
-  if (source === 'deterministic-hash') return hitType ? `本地命中(${hitType})` : '本地命中'
-  if (source === 'manual-pick') return '手动匹配'
-  if (source === 'local-heuristic') return '本地兜底'
-  if (source === 'local-heuristic-next') return '本地兜底(跳过)'
-  if (source === 'browser-cache') return '浏览器缓存'
-  if (source === 'bkg-cover') return '封面缓存'
-  if (source === 'proxy') return '远程代理'
-  if (source === 'unknown') return '未知'
-  if (source === 'error') return '获取失败'
-  return source || '未知'
-}
-
-const snsMediaStageBadgeColorClass = (key) => {
-  const k = String(key || '').trim()
-  const source = String(snsMediaStage.value?.[k]?.source || '').trim()
-
-  if (source.startsWith('remote')) return 'bg-emerald-600/85 text-white'
-  if (source === 'deterministic-hash') return 'bg-sky-600/85 text-white'
-  if (source.startsWith('local')) return 'bg-blue-600/85 text-white'
-  if (source === 'manual-pick') return 'bg-amber-600/90 text-white'
-  if (source === 'browser-cache') return 'bg-slate-600/85 text-white'
-  if (source === 'proxy') return 'bg-fuchsia-600/85 text-white'
-  if (source === 'bkg-cover') return 'bg-indigo-600/85 text-white'
-  if (source === 'error') return 'bg-red-600/85 text-white'
-  return 'bg-black/50 text-white'
-}
-
-const snsMediaStageBadgeTitle = (key) => {
-  const k = String(key || '').trim()
-  const info = snsMediaStage.value?.[k]
-  if (!info || typeof info !== 'object') return ''
-  const source = String(info?.source || '').trim()
-  const hitType = String(info?.hitType || '').trim()
-  const xEnc = String(info?.xEnc || '').trim()
-
-  const parts = []
-  if (source) parts.push(`source=${source}`)
-  if (hitType) parts.push(`hit=${hitType}`)
-  if (xEnc) parts.push(`x-enc=${xEnc}`)
-  return parts.join(' · ')
-}
-
-const readSnsStageFromResourceTiming = (url) => {
-  try {
-    if (!process.client) return null
-    if (typeof performance === 'undefined' || !performance?.getEntriesByName) return null
-    const u = String(url || '').trim()
-    if (!u) return null
-    const entries = performance.getEntriesByName(u) || []
-    const latest = [...entries].reverse().find((e) => String(e?.entryType || '') === 'resource')
-    if (!latest) return null
-
-    // Prefer backend-injected stage info from `Server-Timing`.
-    const st = latest?.serverTiming
-    if (Array.isArray(st) && st.length > 0) {
-      let source = ''
-      let hitType = ''
-      let xEnc = ''
-      for (const item of st) {
-        const name = String(item?.name || '').trim()
-        const desc = String(item?.description || '').trim()
-        if (name === 'sns_source' && desc) source = desc
-        else if (name.startsWith('sns_source_')) source = name.slice('sns_source_'.length) || desc
-        else if (name === 'sns_hit' && desc) hitType = desc
-        else if (name.startsWith('sns_hit_')) hitType = name.slice('sns_hit_'.length) || desc
-        else if (name === 'sns_xenc' && desc) xEnc = desc
-        else if (name.startsWith('sns_xenc_')) xEnc = name.slice('sns_xenc_'.length) || desc
-      }
-      if (source) return { source, hitType, xEnc }
-    }
-
-    // When DevTools shows "(from disk cache)", browsers may not expose `serverTiming` at all.
-    // Best-effort: infer a browser cache hit from ResourceTiming sizes.
-    const transferSize = Number(latest?.transferSize)
-    if (Number.isFinite(transferSize) && transferSize === 0) {
-      return { source: 'browser-cache', hitType: 'transfer=0', xEnc: '' }
-    }
-
-    return null
-  } catch {
-    return null
-  }
-}
-
-const ensureSnsMediaStage = async (key, url) => {
-  if (!process.client) return
-  const k = String(key || '').trim()
-  const u = String(url || '').trim()
-  if (!k || !u) return
-  if (!isSnsMediaApiUrl(u)) return
-
-  const existingSource = String(snsMediaStage.value?.[k]?.source || '').trim()
-  if (existingSource && existingSource !== 'unknown') return
-  if (snsMediaStageLoading.value[k]) return
-  if (snsMediaStageInFlight.has(k)) return
-
-  snsMediaStageInFlight.add(k)
-  snsMediaStageLoading.value[k] = true
-
-  try {
-    // Prefer stage info from the *same* request that loaded the <img>/<video> element
-    // (via Server-Timing + Timing-Allow-Origin), to avoid a non-idempotent extra fetch.
-    let info = null
-    for (const delayMs of [0, 0, 16, 50, 120, 250, 500]) {
-      if (delayMs) await new Promise((resolve) => setTimeout(resolve, delayMs))
-      info = readSnsStageFromResourceTiming(u)
-      if (info) break
-    }
-    snsMediaStage.value[k] = info || { source: 'unknown', hitType: '', xEnc: '' }
-  } finally {
-    snsMediaStageLoading.value[k] = false
-    snsMediaStageInFlight.delete(k)
-  }
-}
-
-const eventCurrentSrc = (ev) => {
-  try {
-    const el = ev?.target || ev?.currentTarget
-    return String(el?.currentSrc || el?.src || '').trim()
-  } catch {
-    return ''
-  }
-}
-
-const onSnsMediaLoaded = (post, m, idx = 0, ev) => {
-  const pid = String(post?.id || '').trim()
-  if (!pid) return
-  const key = snsMediaStageKey(pid, idx, 'thumb')
-  const u = eventCurrentSrc(ev) || getMediaThumbSrc(post, m, idx)
-  ensureSnsMediaStage(key, u)
-}
-
-const onCoverMediaLoaded = (cover, ev) => {
-  const c = cover || activeCover.value
-  if (!c || !Array.isArray(c.media) || c.media.length <= 0) return
-  const u = eventCurrentSrc(ev) || getSnsMediaUrl(c, c.media[0], 0, c.media[0].url)
-  ensureSnsMediaStage(snsCoverStageKey(c), u)
-}
-
-watch([selectedAccount, snsUseCache], () => {
-  snsMediaStage.value = {}
-  snsMediaStageLoading.value = {}
-  snsMediaStageInFlight.clear()
-})
-
 // Article card thumbnail is best-effort: try SNS media thumb first, then fall back to
 // extracting the cover from mp.weixin.qq.com HTML. Track per-post stage so we don't
 // keep showing a broken <img>.
@@ -1504,36 +1270,6 @@ const upgradeTencentHttps = (u) => {
  return raw
 }

-const normalizeHex32 = (value) => {
-  const raw = String(value ?? '').trim()
-  if (!raw) return ''
-  const hex = raw.replace(/[^0-9a-fA-F]/g, '').toLowerCase()
-  return hex.length >= 32 ? hex.slice(0, 32) : ''
-}
-
-const mediaSizeKey = (m) => {
-  const t = String(m?.type ?? '')
-  const w = String(m?.size?.width || m?.size?.w || '').trim()
-  const h = String(m?.size?.height || m?.size?.h || '').trim()
-  if (!w || !h) return ''
-  return `${t}:${w}x${h}`
-}
-
-// Our backend matches SNS cache images by width/height and then uses `idx` to
-// pick the N-th match. `idx` must be the index within the same size-group,
-// not the global media index in the post, otherwise images can shift.
-const mediaSizeGroupIndex = (post, m, idx) => {
-  const list = Array.isArray(post?.media) ? post.media : []
-  const key = mediaSizeKey(m)
-  const i0 = Number(idx) || 0
-  if (!key || i0 <= 0) return i0
-  let count = 0
-  for (let i = 0; i < i0; i++) {
-    if (mediaSizeKey(list[i]) === key) count++
-  }
-  return count
-}
-
 const getSnsMediaUrl = (post, m, idx, rawUrl) => {
  const raw = upgradeTencentHttps(String(rawUrl || '').trim())
  if (!raw) return ''
@@ -1550,36 +1286,12 @@ const getSnsMediaUrl = (post, m, idx, rawUrl) => {
      const host = new URL(raw).hostname.toLowerCase()
      if (host.endsWith('.qpic.cn') || host.endsWith('.qlogo.cn') || host.endsWith('.tc.qq.com')) {
        const acc = String(selectedAccount.value || '').trim()
-        const ct = String(post?.createTime || '').trim()
-        const w = String(m?.size?.width || m?.size?.w || '').trim()
-        const h = String(m?.size?.height || m?.size?.h || '').trim()
-        const ts = String(m?.size?.totalSize || m?.size?.total_size || m?.size?.total || '').trim()
-        const sizeIdx = mediaSizeGroupIndex(post, m, idx)
-        // const pick = getSnsMediaOverridePick(post?.id, idx)
-        let md5 = normalizeHex32(m?.urlAttrs?.md5 || m?.thumbAttrs?.md5 || m?.urlAttrs?.MD5 || m?.thumbAttrs?.MD5)
-        if (!md5) {
-          const match = /[?&]md5=([0-9a-fA-F]{16,32})/.exec(raw)
-          if (match?.[1]) md5 = normalizeHex32(match[1])
-        }
+        // Match WeFlow's image pipeline: use a stable URL + key/token and let the
+        // backend handle cache-first remote fetch/decrypt. Avoid attaching legacy
+        // local-match metadata to the main image path so browser caching can reuse
+        // the same request URL for list + preview.
        const parts = new URLSearchParams()
        if (acc) parts.set('account', acc)
-        if (ct) parts.set('create_time', ct)
-        if (w) parts.set('width', w)
-        if (h) parts.set('height', h)
-        if (/^\d+$/.test(ts)) parts.set('total_size', ts)
-        parts.set('idx', String(Number(sizeIdx) || 0))
-        const pid = String(post?.id || '').trim()
-        if (pid) parts.set('post_id', pid)
-
-        const mid = String(m?.id || '').trim()
-        if (mid) parts.set('media_id', mid)
-
-        const postType = String(post?.type || '1').trim()
-        if (postType) parts.set('post_type', postType)
-
-        const mediaType = String(m?.type || '2').trim()
-        if (mediaType) parts.set('media_type', mediaType)
-
        const token = String(m?.token || m?.urlAttrs?.token || m?.thumbAttrs?.token || '').trim()
        if (token) parts.set('token', token)

@@ -1589,10 +1301,8 @@ const getSnsMediaUrl = (post, m, idx, rawUrl) => {
        parts.set('use_cache', snsUseCache.value ? '1' : '0')
        // When cache is disabled, bust browser caching so backend really downloads+decrypts each time.
        if (!snsUseCache.value) parts.set('_t', String(Date.now()))
-
-        if (md5) parts.set('md5', md5)
-        // Bump this when changing backend matching logic to avoid stale cached wrong images.
-        parts.set('v', '9')
+        // Bump this when changing the WeFlow-aligned image pipeline to avoid stale browser caches.
+        parts.set('v', '10')
        parts.set('url', raw)
        return `${apiBase}/sns/media?${parts.toString()}`
      }
@@ -1607,7 +1317,9 @@ const getMediaThumbSrc = (post, m, idx = 0) => {
 }

 const getMediaPreviewSrc = (post, m, idx = 0) => {
-  return getSnsMediaUrl(post, m, idx, m?.url || m?.thumb)
+  // Align with WeFlow: preview reuses the same prepared image source as the grid
+  // instead of issuing a second "original image" request on click.
+  return getMediaThumbSrc(post, m, idx)
 }


@@ -1755,26 +1467,8 @@ const getLivePhotoVideoSrc = (post, m, idx = 0) => {
  return `${apiBase}/sns/video_remote?${parts.toString()}`
 }

-// 图片预览 + 候选匹配选择
+// 图片预览
 const previewCtx = ref(null) // { post, media, idx }
-const previewCandidatesOpen = ref(false)
-const previewCandidates = reactive({
-  loading: false,
-  loadingMore: false,
-  error: '',
-  items: [],
-  count: 0,
-  hasMore: false
-})
-
-const resetPreviewCandidates = () => {
-  previewCandidates.loading = false
-  previewCandidates.loadingMore = false
-  previewCandidates.error = ''
-  previewCandidates.items = []
-  previewCandidates.count = 0
-  previewCandidates.hasMore = false
-}

 const previewSrc = computed(() => {
  const ctx = previewCtx.value
@@ -1897,60 +1591,7 @@ watch(
  }
 )

-
-const loadPreviewCandidates = async ({ reset }) => {
-  const ctx = previewCtx.value
-  if (!ctx) return
-  const acc = String(selectedAccount.value || '').trim()
-  if (!acc) return
-
-  const toInt = (v) => Number.parseInt(String(v || '').trim(), 10) || 0
-  const w = toInt(ctx.media?.size?.width || ctx.media?.size?.w)
-  const h = toInt(ctx.media?.size?.height || ctx.media?.size?.h)
-
-  // Without dimensions, local matching is too noisy; keep it empty.
-  if (w <= 0 || h <= 0) {
-    resetPreviewCandidates()
-    return
-  }
-
-  const limit = 24
-  const offset = reset ? 0 : (previewCandidates.items?.length || 0)
-
-  if (reset) {
-    resetPreviewCandidates()
-    previewCandidates.loading = true
-  } else {
-    previewCandidates.loadingMore = true
-  }
-  previewCandidates.error = ''
-
-  try {
-    const resp = await api.listSnsMediaCandidates({
-      account: acc,
-      create_time: Number(ctx.post?.createTime || 0),
-      width: w,
-      height: h,
-      limit,
-      offset
-    })
-    const items = Array.isArray(resp?.items) ? resp.items : []
-    previewCandidates.count = Number(resp?.count || 0)
-    previewCandidates.hasMore = !!resp?.hasMore
-    if (reset) {
-      previewCandidates.items = items
-    } else {
-      previewCandidates.items = [...(previewCandidates.items || []), ...items]
-    }
-  } catch (e) {
-    previewCandidates.error = e?.message || '加载候选失败'
-  } finally {
-    previewCandidates.loading = false
-    previewCandidates.loadingMore = false
-  }
-}
-
-const openImagePreview = async (post, m, idx = 0) => {
+const openImagePreview = (post, m, idx = 0) => {
  if (!process.client) return
  resetPreviewVideo()
  // Stop any background hover-playing live photo when opening the preview.
@@ -1965,11 +1606,7 @@ const openImagePreview = async (post, m, idx = 0) => {
    }
  }
  previewCtx.value = { post, media: m, idx: Number(idx) || 0 }
-  previewCandidatesOpen.value = false
-  resetPreviewCandidates()
  document.body.style.overflow = 'hidden'
-  // Load the first page so we can show the candidate count in the header.
-  await loadPreviewCandidates({ reset: true })
 }

 const openVideoPreview = (post, m, idx = 0) => {
@@ -1987,8 +1624,6 @@ const openVideoPreview = (post, m, idx = 0) => {
  else previewVideoError.value = '视频地址缺失。'

  previewCtx.value = { post, media: m, idx: Number(idx) || 0 }
-  previewCandidatesOpen.value = false
-  resetPreviewCandidates()
  document.body.style.overflow = 'hidden'
 }

@@ -2021,8 +1656,6 @@ const onPreviewVideoError = () => {
 const closeImagePreview = () => {
  if (!process.client) return
  previewCtx.value = null
-  previewCandidatesOpen.value = false
-  resetPreviewCandidates()
  resetPreviewVideo()
  document.body.style.overflow = ''
 }
@@ -2038,7 +1671,7 @@ const onMediaClick = (post, m, idx = 0) => {
  }

  // 图片：打开预览
-  void openImagePreview(post, m, idx)
+  openImagePreview(post, m, idx)
 }

 const formatRelativeTime = (tsSeconds) => {
@@ -35,7 +35,6 @@ from .routers.sns_export import router as _sns_export_router
 from .routers.wechat_detection import router as _wechat_detection_router
 from .routers.wrapped import router as _wrapped_router
 from .request_logging import log_server_errors_middleware
-from .sns_stage_timing import add_sns_stage_timing_headers
 from .wcdb_realtime import WCDB_REALTIME, shutdown as _wcdb_shutdown
 from .routers.biz import router as _biz_router
 from .routers.system import router as _system_router
@@ -56,31 +55,9 @@ app.add_middleware(
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
-    expose_headers=["X-SNS-Source", "X-SNS-Hit-Type", "X-SNS-X-Enc"],
 )


-@app.middleware("http")
-async def _add_sns_stage_timing_headers(request: Request, call_next):
-    """Expose SNS stage metadata to the frontend without extra requests.
-
-    `<img>` elements can't read response headers, but browsers can surface `Server-Timing`
-    via `performance.getEntriesByName(...).serverTiming` when `Timing-Allow-Origin` is set.
-    """
-
-    response = await call_next(request)
-    try:
-        add_sns_stage_timing_headers(
-            response.headers,
-            source=str(response.headers.get("X-SNS-Source") or ""),
-            hit_type=str(response.headers.get("X-SNS-Hit-Type") or ""),
-            x_enc=str(response.headers.get("X-SNS-X-Enc") or ""),
-        )
-    except Exception:
-        pass
-    return response
-
-
@app.middleware("http")
 async def _log_server_errors(request: Request, call_next):
    return await log_server_errors_middleware(request_logger, request, call_next)
@@ -1,4 +1,3 @@
-from bisect import bisect_left, bisect_right
 from functools import lru_cache
 from pathlib import Path
 import os
@@ -20,7 +19,6 @@ from starlette.background import BackgroundTask

 from fastapi import APIRouter, HTTPException
 from fastapi.responses import Response, FileResponse  # 返回视频文件
-from pydantic import BaseModel, Field

 from ..chat_helpers import _load_contact_rows, _pick_display_name, _resolve_account_dir
 from ..logging_config import get_logger
@@ -44,8 +42,6 @@ logger = get_logger(__name__)

 router = APIRouter(route_class=PathFixRoute)

-SNS_MEDIA_PICKS_FILE = "_sns_media_picks.json"
-
 _SNS_VIDEO_KEY_RE = re.compile(r'<enc\s+key="(\d+)"', flags=re.IGNORECASE)
 _MP_BIZ_RE = re.compile(r"__biz=([A-Za-z0-9_=+-]+)")
 _ZSTD_MAGIC = b"\x28\xb5\x2f\xfd"
@@ -860,233 +856,6 @@ def _parse_timeline_xml(xml_text: str, fallback_username: str) -> dict[str, Any]
    return out


-def _image_size_from_bytes(data: bytes, media_type: str) -> tuple[int, int]:
-    mt = str(media_type or "").lower()
-    if mt == "image/png":
-        # PNG IHDR width/height are stored at byte offsets 16..24
-        if len(data) >= 24 and data.startswith(b"\x89PNG\r\n\x1a\n"):
-            try:
-                w = int.from_bytes(data[16:20], "big")
-                h = int.from_bytes(data[20:24], "big")
-                return w, h
-            except Exception:
-                return 0, 0
-        return 0, 0
-    if mt in {"image/jpeg", "image/jpg"}:
-        # Minimal JPEG SOF parser.
-        if len(data) < 4 or (not data.startswith(b"\xFF\xD8")):
-            return 0, 0
-        i = 2
-        while i + 3 < len(data):
-            if data[i] != 0xFF:
-                i += 1
-                continue
-            # Skip padding 0xFF bytes.
-            while i < len(data) and data[i] == 0xFF:
-                i += 1
-            if i >= len(data):
-                break
-            marker = data[i]
-            i += 1
-            # Markers without a segment length.
-            if marker in (0xD8, 0xD9):
-                continue
-            if marker == 0xDA:  # Start of scan.
-                break
-            if i + 1 >= len(data):
-                break
-            seg_len = (data[i] << 8) + data[i + 1]
-            i += 2
-            if seg_len < 2:
-                break
-            # SOF markers which contain width/height.
-            if marker in {
-                0xC0,
-                0xC1,
-                0xC2,
-                0xC3,
-                0xC5,
-                0xC6,
-                0xC7,
-                0xC9,
-                0xCA,
-                0xCB,
-                0xCD,
-                0xCE,
-                0xCF,
-            }:
-                # segment: [precision(1), height(2), width(2), ...]
-                if i + 4 < len(data):
-                    try:
-                        h = (data[i + 1] << 8) + data[i + 2]
-                        w = (data[i + 3] << 8) + data[i + 4]
-                        return w, h
-                    except Exception:
-                        return 0, 0
-            i += seg_len - 2
-        return 0, 0
-    return 0, 0
-
-
-@lru_cache(maxsize=16)
-def _sns_img_time_index(wxid_dir_str: str) -> tuple[list[float], list[str]]:
-    """Build a (mtime_sorted, path_sorted) index for local Moments cache images.
-
-    WeChat stores encrypted SNS cache images under:
-      `{wxid_dir}/cache/YYYY-MM/Sns/Img/<2hex>/<30hex>`
-    """
-    wxid_dir = Path(str(wxid_dir_str or "").strip())
-    out: list[tuple[float, str]] = []
-
-    cache_root = wxid_dir / "cache"
-    try:
-        month_dirs = [p for p in cache_root.iterdir() if p.is_dir()]
-    except Exception:
-        month_dirs = []
-
-    for mdir in month_dirs:
-        img_root = mdir / "Sns" / "Img"
-        try:
-            if not (img_root.exists() and img_root.is_dir()):
-                continue
-        except Exception:
-            continue
-        # The Img dir uses a 2-level layout; keep this tight (no global rglob).
-        try:
-            for sub in img_root.iterdir():
-                if not sub.is_dir():
-                    continue
-                for f in sub.iterdir():
-                    try:
-                        if not f.is_file():
-                            continue
-                        st = f.stat()
-                        out.append((float(st.st_mtime), str(f)))
-                    except Exception:
-                        continue
-        except Exception:
-            continue
-
-    out.sort(key=lambda x: x[0])
-    mtimes = [m for m, _p in out]
-    paths = [_p for _m, _p in out]
-    return mtimes, paths
-
-
-def _normalize_hex32(value: Optional[str]) -> str:
-    """Return the first 32 hex chars from value, or '' if not present."""
-    s = str(value or "").strip().lower()
-    if not s:
-        return ""
-    # Keep only hex chars. Some attrs may contain separators or be wrapped.
-    s = re.sub(r"[^0-9a-f]", "", s)
-    if len(s) < 32:
-        return ""
-    return s[:32]
-
-
-def _sns_media_picks_path(account_dir: Path) -> Path:
-    return account_dir / SNS_MEDIA_PICKS_FILE
-
-
-def _sns_post_id_from_media_key(media_key: str) -> str:
-    # Frontend stores picks under `${postId}:${idx}`.
-    s = str(media_key or "").strip()
-    if not s:
-        return ""
-    return s.split(":", 1)[0].strip()
-
-
-@lru_cache(maxsize=32)
-def _load_sns_media_picks_cached(path_str: str, mtime: float) -> dict[str, str]:
-    p = Path(str(path_str or "").strip())
-    try:
-        raw = p.read_text(encoding="utf-8")
-    except Exception:
-        return {}
-
-    try:
-        obj = json.loads(raw)
-    except Exception:
-        return {}
-
-    picks_obj = obj.get("picks") if isinstance(obj, dict) else None
-    if not isinstance(picks_obj, dict):
-        return {}
-
-    out: dict[str, str] = {}
-    for k, v in picks_obj.items():
-        mk = str(k or "").strip()
-        if not mk:
-            continue
-        ck = _normalize_hex32(str(v or ""))
-        if not ck:
-            continue
-        out[mk] = ck
-    return out
-
-
-def _load_sns_media_picks(account_dir: Path) -> dict[str, str]:
-    p = _sns_media_picks_path(account_dir)
-    try:
-        st = p.stat()
-        mtime = float(st.st_mtime)
-    except Exception:
-        mtime = 0.0
-    return _load_sns_media_picks_cached(str(p), mtime)
-
-
-def _save_sns_media_picks(account_dir: Path, picks: dict[str, str]) -> int:
-    # Normalize + keep it stable for easier diff/debugging.
-    out: dict[str, str] = {}
-    for k, v in (picks or {}).items():
-        mk = str(k or "").strip()
-        if not mk:
-            continue
-        ck = _normalize_hex32(str(v or ""))
-        if not ck:
-            continue
-        out[mk] = ck
-
-    try:
-        payload = {"updated_at": int(time.time()), "picks": dict(sorted(out.items(), key=lambda x: x[0]))}
-        _sns_media_picks_path(account_dir).write_text(
-            json.dumps(payload, ensure_ascii=False, indent=2),
-            encoding="utf-8",
-        )
-    except Exception:
-        pass
-
-    try:
-        _load_sns_media_picks_cached.cache_clear()
-    except Exception:
-        pass
-
-    return len(out)
-
-
-@lru_cache(maxsize=16)
-def _sns_img_roots(wxid_dir_str: str) -> tuple[str, ...]:
-    """List all month cache roots that contain `Sns/Img`."""
-    wxid_dir = Path(str(wxid_dir_str or "").strip())
-    cache_root = wxid_dir / "cache"
-    try:
-        month_dirs = [p for p in cache_root.iterdir() if p.is_dir()]
-    except Exception:
-        month_dirs = []
-
-    roots: list[str] = []
-    for mdir in month_dirs:
-        img_root = mdir / "Sns" / "Img"
-        try:
-            if img_root.exists() and img_root.is_dir():
-                roots.append(str(img_root))
-        except Exception:
-            continue
-    # Keep it stable (helps debugging and caching predictability).
-    roots.sort()
-    return tuple(roots)
-
@lru_cache(maxsize=16)
 def _sns_video_roots(wxid_dir_str: str) -> tuple[str, ...]:
    """List all month cache roots that contain `Sns/Video`."""
@@ -1139,268 +908,6 @@ def _resolve_sns_cached_video_path(

    return None

-
-def _resolve_sns_cached_image_path_by_md5(
-    *,
-    wxid_dir: Path,
-    md5: str,
-    create_time: int,
-) -> Optional[str]:
-    """Try to resolve SNS cache image by md5-based cache path layout."""
-    md5_32 = _normalize_hex32(md5)
-    if not md5_32:
-        return None
-
-    sub = md5_32[:2]
-    rest = md5_32[2:]
-    roots = _sns_img_roots(str(wxid_dir))
-    if not roots:
-        return None
-
-    best: tuple[float, str] | None = None
-    for root_str in roots:
-        try:
-            p = Path(root_str) / sub / rest
-            if not (p.exists() and p.is_file()):
-                continue
-            # Prefer the cache file closest to the post create_time (if provided),
-            # otherwise pick the newest one.
-            st = p.stat()
-            if create_time > 0:
-                score = abs(float(st.st_mtime) - float(create_time))
-            else:
-                score = -float(st.st_mtime)
-            if best is None or score < best[0]:
-                best = (score, str(p))
-        except Exception:
-            continue
-    return best[1] if best else None
-
-
-def _sns_cache_key_from_path(p: Path) -> str:
-    """Return the 32-hex cache key for a SNS cache file path, or ''."""
-    try:
-        # cache/.../Sns/Img/<2hex>/<30hex>
-        key = f"{p.parent.name}{p.name}"
-    except Exception:
-        return ""
-    return _normalize_hex32(key)
-
-
-def _generate_sns_cache_key(tid: str, media_id: str, media_type: int = 2) -> str:
-    """
-    公式: md5(tid_mediaId_type)
-    Example: 14852422213384352392_14852422213963625090_2 -> 6d479249ca5a090fab5c42c79bc56b89
-    """
-    if not tid or not media_id:
-        return ""
-
-    raw_key = f"{tid}_{media_id}_{media_type}"
-
-    try:
-        return hashlib.md5(raw_key.encode("utf-8")).hexdigest()
-    except Exception:
-        return ""
-
-def _resolve_sns_cached_image_path_by_cache_key(
-    *,
-    wxid_dir: Path,
-    cache_key: str,
-    create_time: int,
-) -> Optional[str]:
-    """Resolve SNS cache image by `<2hex>/<30hex>` cache key."""
-    key32 = _normalize_hex32(cache_key)
-    if not key32:
-        return None
-
-    sub = key32[:2]
-    rest = key32[2:]
-    roots = _sns_img_roots(str(wxid_dir))
-    if not roots:
-        return None
-
-    best: tuple[float, str] | None = None
-    for root_str in roots:
-        try:
-            p = Path(root_str) / sub / rest
-            if not (p.exists() and p.is_file()):
-                continue
-            st = p.stat()
-            if create_time > 0:
-                score = abs(float(st.st_mtime) - float(create_time))
-            else:
-                score = -float(st.st_mtime)
-            if best is None or score < best[0]:
-                best = (score, str(p))
-        except Exception:
-            continue
-    return best[1] if best else None
-
-
-@lru_cache(maxsize=4096)
-def _resolve_sns_cached_image_path(
-    *,
-    account_dir_str: str,
-    create_time: int,
-    width: int,
-    height: int,
-    idx: int,
-    total_size: int = 0,
-) -> Optional[str]:
-    """Best-effort resolve a local cached SNS image for a post+media meta."""
-    total_size_i = int(total_size or 0)
-    must_match_size = width > 0 and height > 0
-    # Without size/total_size, time-only matching is too error-prone and can easily mix images.
-    if (not must_match_size) and total_size_i <= 0:
-        return None
-
-    account_dir = Path(str(account_dir_str or "").strip())
-    if not account_dir.exists():
-        return None
-
-    wxid_dir = _resolve_account_wxid_dir(account_dir)
-    if not wxid_dir:
-        return None
-
-    mtimes, paths = _sns_img_time_index(str(wxid_dir))
-    if not mtimes:
-        return None
-
-    create_time_i = int(create_time or 0)
-    if create_time_i > 0:
-        # We don't know when the image was cached (could be close to create_time, could be hours later).
-        # Use a generous window but keep it bounded for performance.
-        window = 72 * 3600  # 72h
-        lo = create_time_i - window
-        hi = create_time_i + window
-
-        l = bisect_left(mtimes, lo)
-        r = bisect_right(mtimes, hi)
-        if l >= r:
-            # Fallback: search the newest N files if time window has no hits.
-            l = max(0, len(mtimes) - 800)
-            r = len(mtimes)
-    else:
-        # Missing createTime: only probe the newest cache entries.
-        l = max(0, len(mtimes) - 800)
-        r = len(mtimes)
-
-    # Rank by time proximity to create_time (or by recency when createTime is missing).
-    candidates: list[tuple[float, str]] = []
-    for j in range(l, r):
-        try:
-            if create_time_i > 0:
-                candidates.append((abs(mtimes[j] - float(create_time_i)), paths[j]))
-            else:
-                candidates.append((-mtimes[j], paths[j]))
-        except Exception:
-            continue
-    candidates.sort(key=lambda x: x[0])
-
-    matched: list[tuple[int, float, str]] = []
-    # Limit the work per request.
-    max_probe = 2000 if (r - l) <= 2000 else 2000
-    for _diff, pstr in candidates[:max_probe]:
-        try:
-            p = Path(pstr)
-            payload, media_type = _read_and_maybe_decrypt_media(p, account_dir)
-            if not payload or not str(media_type or "").startswith("image/"):
-                continue
-            if must_match_size:
-                w0, h0 = _image_size_from_bytes(payload, str(media_type or ""))
-                if (w0, h0) != (width, height):
-                    continue
-
-            size_diff = abs(len(payload) - total_size_i) if total_size_i > 0 else 0
-            # When totalSize is available, it tends to be a stronger discriminator than mtime.
-            matched.append((int(size_diff), float(_diff), pstr))
-        except Exception:
-            continue
-
-    if not matched:
-        return None
-    if must_match_size:
-        matched.sort(key=lambda x: (x[0], x[1], x[2]))
-        # If we have totalSize, treat it as a strong discriminator and always take the best match.
-        if total_size_i > 0:
-            return matched[0][2]
-        idx0 = max(0, int(idx or 0))
-        return matched[idx0][2] if idx0 < len(matched) else None
-    # No size: only return a best-effort match when totalSize is available.
-    if total_size_i > 0:
-        matched.sort(key=lambda x: (x[0], x[1], x[2]))
-        return matched[0][2]
-    return None
-
-
-@lru_cache(maxsize=2048)
-def _list_sns_cached_image_candidate_keys(
-    *,
-    account_dir_str: str,
-    create_time: int,
-    width: int,
-    height: int,
-) -> tuple[str, ...]:
-    """List local SNS cache candidates (as 32-hex cache keys) for a media item.
-
-    The ordering matches `_resolve_sns_cached_image_path()`'s scan order, so `idx`
-    is stable within the same (account, create_time, width, height) input.
-    """
-    if create_time <= 0 or width <= 0 or height <= 0:
-        return tuple()
-
-    account_dir = Path(str(account_dir_str or "").strip())
-    if not account_dir.exists():
-        return tuple()
-
-    wxid_dir = _resolve_account_wxid_dir(account_dir)
-    if not wxid_dir:
-        return tuple()
-
-    mtimes, paths = _sns_img_time_index(str(wxid_dir))
-    if not mtimes:
-        return tuple()
-
-    window = 72 * 3600  # 72h
-    lo = create_time - window
-    hi = create_time + window
-
-    l = bisect_left(mtimes, lo)
-    r = bisect_right(mtimes, hi)
-    if l >= r:
-        l = max(0, len(mtimes) - 800)
-        r = len(mtimes)
-
-    candidates: list[tuple[float, str]] = []
-    for j in range(l, r):
-        try:
-            candidates.append((abs(mtimes[j] - float(create_time)), paths[j]))
-        except Exception:
-            continue
-    candidates.sort(key=lambda x: x[0])
-
-    max_probe = 2000 if (r - l) <= 2000 else 2000
-    out: list[str] = []
-    seen: set[str] = set()
-    for _diff, pstr in candidates[:max_probe]:
-        try:
-            p = Path(pstr)
-            payload, media_type = _read_and_maybe_decrypt_media(p, account_dir)
-            if not payload or not str(media_type or "").startswith("image/"):
-                continue
-            w0, h0 = _image_size_from_bytes(payload, str(media_type or ""))
-            if (w0, h0) != (width, height):
-                continue
-            key = _sns_cache_key_from_path(p)
-            if not key or key in seen:
-                continue
-            seen.add(key)
-            out.append(key)
-        except Exception:
-            continue
-
-    return tuple(out)
-
 def _get_sns_covers(account_dir: Path, target_wxid: str, limit: int = 20) -> list[dict[str, Any]]:
    """无论多古老，强行揪出用户的朋友圈封面历史 (type=7)。

@@ -2575,47 +2082,6 @@ def list_sns_users(
    return {"items": items, "count": len(items), "limit": lim}


-class SnsMediaPicksSaveRequest(BaseModel):
-    account: Optional[str] = Field(None, description="账号目录名（可选，默认使用第一个）")
-    picks: dict[str, str] = Field(default_factory=dict, description="手动匹配表：`${postId}:${idx}` -> 32hex cacheKey")
-
-
-@router.post("/api/sns/media_picks", summary="保存朋友圈图片手动匹配结果（本机）")
-async def save_sns_media_picks(request: SnsMediaPicksSaveRequest):
-    account_dir = _resolve_account_dir(request.account)
-    count = _save_sns_media_picks(account_dir, request.picks or {})
-    return {"status": "success", "count": int(count)}
-
-
-@router.get("/api/sns/media_candidates", summary="获取朋友圈图片本地缓存候选")
-def list_sns_media_candidates(
-    account: Optional[str] = None,
-    create_time: int = 0,
-    width: int = 0,
-    height: int = 0,
-    limit: int = 24,
-    offset: int = 0,
-):
-    if limit <= 0:
-        raise HTTPException(status_code=400, detail="Invalid limit.")
-    if limit > 200:
-        limit = 200
-    if offset < 0:
-        offset = 0
-
-    account_dir = _resolve_account_dir(account)
-    keys = _list_sns_cached_image_candidate_keys(
-        account_dir_str=str(account_dir),
-        create_time=int(create_time or 0),
-        width=int(width or 0),
-        height=int(height or 0),
-    )
-    total = len(keys)
-    end = min(total, offset + limit)
-    items = [{"idx": i, "key": keys[i]} for i in range(offset, end)]
-    return {"count": total, "items": items, "hasMore": end < total, "limit": limit, "offset": offset}
-
-
 def _is_allowed_sns_media_host(host: str) -> bool:
    return _sns_media.is_allowed_sns_media_host(host)

@@ -2902,10 +2368,7 @@ async def _try_fetch_and_decrypt_sns_remote(
    token: str,
    use_cache: bool,
 ) -> Optional[Response]:
-    """Try remote download+decrypt first (accurate when keys are present).
-
-    Returns a Response on success, or None on failure so caller can fall back to local cache matching.
-    """
+    """Try remote download+decrypt first (accurate when keys are present)."""
    res = await _sns_media.try_fetch_and_decrypt_sns_image_remote(
        account_dir=account_dir,
        url=str(url or ""),
@@ -2918,34 +2381,18 @@ async def _try_fetch_and_decrypt_sns_remote(

    resp = Response(content=res.payload, media_type=res.media_type)
    resp.headers["Cache-Control"] = "public, max-age=86400" if use_cache else "no-store"
-    resp.headers["X-SNS-Source"] = str(res.source or "remote")
-    if res.x_enc:
-        resp.headers["X-SNS-X-Enc"] = str(res.x_enc)
    return resp


@router.get("/api/sns/media", summary="获取朋友圈图片（下载解密优先）")
 async def get_sns_media(
        account: Optional[str] = None,
-        create_time: int = 0,
-        width: int = 0,
-        height: int = 0,
-        total_size: int = 0,
-        idx: int = 0,
-        avoid_picked: int = 0,
-        post_id: Optional[str] = None,
-        media_id: Optional[str] = None,
-        post_type: int = 1,
-        media_type: int = 2,
-        pick: Optional[str] = None,
-        md5: Optional[str] = None,
        token: Optional[str] = None,
        key: Optional[str] = None,
        use_cache: int = 1,
        url: Optional[str] = None,
 ):
    account_dir = _resolve_account_dir(account)
-    wxid_dir = _resolve_account_wxid_dir(account_dir)

    try:
        use_cache_flag = bool(int(use_cache or 1))
@@ -2963,179 +2410,7 @@ async def get_sns_media(
    if remote_resp is not None:
        return remote_resp

-    # Cache disabled: do not fall back to local cache heuristics.
-    if not use_cache_flag:
-        raise HTTPException(status_code=404, detail="SNS media not found (cache disabled).")
-
-    if wxid_dir and post_id and media_id:
-        if int(post_type) == 7:
-            raw_key = f"{post_id}_{media_id}_4"  # 硬编码
-
-            md5_str = hashlib.md5(raw_key.encode("utf-8")).hexdigest()
-            bkg_path = wxid_dir / "business" / "sns" / "bkg" / md5_str[:2] / md5_str
-
-            if bkg_path.exists() and bkg_path.is_file():
-                print(f"===== Hit Bkg Cover ======= {bkg_path}")
-
-                return FileResponse(bkg_path, media_type="image/jpeg",
-                                    headers={"Cache-Control": "public, max-age=31536000", "X-SNS-Source": "bkg-cover"})
-        exact_match_path = None
-        hit_type = ""
-
-        # 尝试 1: 使用 post_type 计算 MD5
-        key_post = _generate_sns_cache_key(post_id, media_id, post_type)
-        exact_match_path = _resolve_sns_cached_image_path_by_cache_key(
-            wxid_dir=wxid_dir,
-            cache_key=key_post,
-            create_time=0
-        )
-        if exact_match_path:
-            hit_type = "post_type"
-
-        # 尝试 2: 如果没找到，并且 media_type 和 post_type 不一样，再试一次
-        if not exact_match_path and post_type != media_type:
-            key_media = _generate_sns_cache_key(post_id, media_id, media_type)
-            exact_match_path = _resolve_sns_cached_image_path_by_cache_key(
-                wxid_dir=wxid_dir,
-                cache_key=key_media,
-                create_time=0
-            )
-            if exact_match_path:
-                hit_type = "media_type"
-
-        # 如果通过这两种精确定位找到了文件，直接返回
-        if exact_match_path:
-            print(f"=====exact_match_path======={exact_match_path}============= (Hit: {hit_type})")
-            try:
-                payload, mtype = _read_and_maybe_decrypt_media(Path(exact_match_path), account_dir)
-                if payload and str(mtype or "").startswith("image/"):
-                    resp = Response(content=payload, media_type=str(mtype or "image/jpeg"))
-                    resp.headers["Cache-Control"] = "public, max-age=31536000"
-                    resp.headers["X-SNS-Source"] = "deterministic-hash"
-                    # 在 Header 里塞入到底是哪个 type 命中的，方便 F12 调试
-                    resp.headers["X-SNS-Hit-Type"] = hit_type
-                    return resp
-            except Exception:
-                pass
-
-        print("no exact match path, falling back...")
-
-    # 0) User-picked cache key override (stable across candidate ordering).
-    pick_key = _normalize_hex32(pick)
-    if pick_key:
-        wxid_dir = _resolve_account_wxid_dir(account_dir)
-        if wxid_dir:
-            local = _resolve_sns_cached_image_path_by_cache_key(
-                wxid_dir=wxid_dir,
-                cache_key=pick_key,
-                create_time=int(create_time or 0),
-            )
-            if local:
-                try:
-                    payload, media_type = _read_and_maybe_decrypt_media(Path(local), account_dir)
-                    if payload and str(media_type or "").startswith("image/"):
-                        resp = Response(content=payload, media_type=str(media_type or "image/jpeg"))
-                        resp.headers["Cache-Control"] = "public, max-age=86400"
-                        resp.headers["X-SNS-Source"] = "manual-pick"
-                        return resp
-                except Exception:
-                    pass
-
-    # Optional: avoid using a cache image that was manually pinned to another post.
-    # Only applies when frontend enables this setting and the current media has no explicit `pick`.
-    try:
-        avoid_flag = bool(int(avoid_picked or 0))
-    except Exception:
-        avoid_flag = False
-    cur_post_id = str(post_id or "").strip()
-    reserved_other: set[str] = set()
-    if avoid_flag and (not pick_key) and cur_post_id:
-        picks_map = _load_sns_media_picks(account_dir)
-        for mk, ck in (picks_map or {}).items():
-            pid = _sns_post_id_from_media_key(mk)
-            if not pid or pid == cur_post_id:
-                continue
-            if ck:
-                reserved_other.add(str(ck))
-
-    # 1) Try local decrypted cache first (works for old posts where CDN URLs return placeholders).
-    local = _resolve_sns_cached_image_path(
-        account_dir_str=str(account_dir),
-        create_time=int(create_time or 0),
-        width=int(width or 0),
-        height=int(height or 0),
-        idx=max(0, int(idx or 0)),
-        total_size=int(total_size or 0),
-    )
-    if local and reserved_other:
-        try:
-            ck0 = _sns_cache_key_from_path(Path(local))
-            if ck0 and ck0 in reserved_other:
-                local = None
-        except Exception:
-            pass
-    if local:
-        try:
-            payload, media_type = _read_and_maybe_decrypt_media(Path(local), account_dir)
-            if payload and str(media_type or "").startswith("image/"):
-                resp = Response(content=payload, media_type=str(media_type or "image/jpeg"))
-                resp.headers["Cache-Control"] = "public, max-age=86400"
-                resp.headers["X-SNS-Source"] = "local-heuristic"
-                return resp
-        except Exception:
-            pass
-
-    # 1.5) If enabled, and the default match was skipped (or not found), pick the next candidate
-    # that is not reserved by a manual pick on another post.
-    if reserved_other and int(create_time or 0) > 0 and int(width or 0) > 0 and int(height or 0) > 0:
-        wxid_dir = _resolve_account_wxid_dir(account_dir)
-        if wxid_dir:
-            keys = _list_sns_cached_image_candidate_keys(
-                account_dir_str=str(account_dir),
-                create_time=int(create_time or 0),
-                width=int(width or 0),
-                height=int(height or 0),
-            )
-            base_idx = max(0, int(idx or 0))
-            for ck in keys[base_idx:]:
-                if not ck or ck in reserved_other:
-                    continue
-                local2 = _resolve_sns_cached_image_path_by_cache_key(
-                    wxid_dir=wxid_dir,
-                    cache_key=str(ck),
-                    create_time=int(create_time or 0),
-                )
-                if not local2:
-                    continue
-                try:
-                    payload, media_type = _read_and_maybe_decrypt_media(Path(local2), account_dir)
-                    if payload and str(media_type or "").startswith("image/"):
-                        resp = Response(content=payload, media_type=str(media_type or "image/jpeg"))
-                        resp.headers["Cache-Control"] = "public, max-age=86400"
-                        resp.headers["X-SNS-Source"] = "local-heuristic-next"
-                        return resp
-                except Exception:
-                    continue
-
-    # 2) Fallback to the remote URL (may still return a Tencent placeholder image).
-    u = str(url or "").strip()
-    if not u:
-        raise HTTPException(status_code=404, detail="SNS media not found.")
-
-    # Delay-import to avoid pulling requests machinery during normal timeline listing.
-    from .chat_media import proxy_image  # pylint: disable=import-outside-toplevel
-
-    try:
-        resp0 = await proxy_image(u)
-        try:
-            resp0.headers["X-SNS-Source"] = "proxy"
-        except Exception:
-            pass
-        return resp0
-    except HTTPException:
-        raise
-    except Exception as e:
-        raise HTTPException(status_code=502, detail=f"Fetch sns media failed: {e}")
+    raise HTTPException(status_code=404, detail="SNS media not found.")


@router.get("/api/sns/article_thumb", summary="提取公众号文章封面图")
@@ -3197,8 +2472,7 @@ async def get_sns_video_remote(
    if path is None:
        raise HTTPException(status_code=404, detail="SNS remote video not found.")

-    headers = {"X-SNS-Source": "remote-video-cache" if use_cache_flag else "remote-video"}
-    headers["Cache-Control"] = "public, max-age=86400" if use_cache_flag else "no-store"
+    headers = {"Cache-Control": "public, max-age=86400" if use_cache_flag else "no-store"}

    if use_cache_flag:
        return FileResponse(str(path), media_type="video/mp4", headers=headers)
@@ -3,8 +3,10 @@ from __future__ import annotations
 """SNS (Moments) HTML export service (offline ZIP)."""

 import asyncio
+from bisect import bisect_left, bisect_right
 from dataclasses import dataclass, field
 from datetime import datetime
+from functools import lru_cache
 import hashlib
 import html
 import json
@@ -33,10 +35,6 @@ from .chat_export_service import (  # pylint: disable=protected-access

 # Reuse SNS timeline/local cache helpers.
 from .routers.sns import (  # pylint: disable=protected-access
-    _generate_sns_cache_key,
-    _resolve_sns_cached_image_path,
-    _resolve_sns_cached_image_path_by_cache_key,
-    _resolve_sns_cached_image_path_by_md5,
    _resolve_sns_cached_video_path,
    list_sns_timeline,
 )
@@ -54,6 +52,7 @@ ExportStatus = Literal["queued", "running", "done", "error", "cancelled"]
 ExportScope = Literal["selected", "all"]

 _INVALID_PATH_CHARS = re.compile(r'[<>:"/\\|?*\x00-\x1f]')
+_HEX_ONLY_RE = re.compile(r"[^0-9a-fA-F]+")


 def _safe_name(s: str, max_len: int = 80) -> str:
@@ -101,6 +100,289 @@ def _mime_to_ext(mt: str) -> str:
    }.get(m, ".bin")


+def _normalize_hex32(value: Any) -> str:
+    raw = str(value or "").strip()
+    if not raw:
+        return ""
+    hex_only = _HEX_ONLY_RE.sub("", raw).lower()
+    return hex_only[:32] if len(hex_only) >= 32 else ""
+
+
+def _image_size_from_bytes(data: bytes, media_type: str) -> tuple[int, int]:
+    mt = str(media_type or "").lower()
+    if mt == "image/png":
+        if len(data) >= 24 and data.startswith(b"\x89PNG\r\n\x1a\n"):
+            try:
+                w = int.from_bytes(data[16:20], "big")
+                h = int.from_bytes(data[20:24], "big")
+                return w, h
+            except Exception:
+                return 0, 0
+        return 0, 0
+    if mt in {"image/jpeg", "image/jpg"}:
+        if len(data) < 4 or not data.startswith(b"\xFF\xD8"):
+            return 0, 0
+        i = 2
+        while i + 3 < len(data):
+            if data[i] != 0xFF:
+                i += 1
+                continue
+            while i < len(data) and data[i] == 0xFF:
+                i += 1
+            if i >= len(data):
+                break
+            marker = data[i]
+            i += 1
+            if marker in (0xD8, 0xD9):
+                continue
+            if marker == 0xDA:
+                break
+            if i + 1 >= len(data):
+                break
+            seg_len = (data[i] << 8) + data[i + 1]
+            i += 2
+            if seg_len < 2:
+                break
+            if marker in {
+                0xC0,
+                0xC1,
+                0xC2,
+                0xC3,
+                0xC5,
+                0xC6,
+                0xC7,
+                0xC9,
+                0xCA,
+                0xCB,
+                0xCD,
+                0xCE,
+                0xCF,
+            }:
+                if i + 4 < len(data):
+                    try:
+                        h = (data[i + 1] << 8) + data[i + 2]
+                        w = (data[i + 3] << 8) + data[i + 4]
+                        return w, h
+                    except Exception:
+                        return 0, 0
+            i += seg_len - 2
+        return 0, 0
+    return 0, 0
+
+
+@lru_cache(maxsize=16)
+def _sns_img_roots(wxid_dir_str: str) -> tuple[str, ...]:
+    wxid_dir = Path(str(wxid_dir_str or "").strip())
+    cache_root = wxid_dir / "cache"
+    try:
+        month_dirs = [p for p in cache_root.iterdir() if p.is_dir()]
+    except Exception:
+        month_dirs = []
+
+    roots: list[str] = []
+    for mdir in month_dirs:
+        img_root = mdir / "Sns" / "Img"
+        try:
+            if img_root.exists() and img_root.is_dir():
+                roots.append(str(img_root))
+        except Exception:
+            continue
+    roots.sort()
+    return tuple(roots)
+
+
+@lru_cache(maxsize=16)
+def _sns_img_time_index(wxid_dir_str: str) -> tuple[list[float], list[str]]:
+    wxid_dir = Path(str(wxid_dir_str or "").strip())
+    out: list[tuple[float, str]] = []
+
+    cache_root = wxid_dir / "cache"
+    try:
+        month_dirs = [p for p in cache_root.iterdir() if p.is_dir()]
+    except Exception:
+        month_dirs = []
+
+    for mdir in month_dirs:
+        img_root = mdir / "Sns" / "Img"
+        try:
+            if not (img_root.exists() and img_root.is_dir()):
+                continue
+        except Exception:
+            continue
+        try:
+            for sub in img_root.iterdir():
+                if not sub.is_dir():
+                    continue
+                for f in sub.iterdir():
+                    try:
+                        if not f.is_file():
+                            continue
+                        st = f.stat()
+                        out.append((float(st.st_mtime), str(f)))
+                    except Exception:
+                        continue
+        except Exception:
+            continue
+
+    out.sort(key=lambda x: x[0])
+    mtimes = [m for m, _p in out]
+    paths = [_p for _m, _p in out]
+    return mtimes, paths
+
+
+def _generate_sns_cache_key(tid: str, media_id: str, media_type: int = 2) -> str:
+    if not tid or not media_id:
+        return ""
+    raw_key = f"{tid}_{media_id}_{media_type}"
+    try:
+        return hashlib.md5(raw_key.encode("utf-8")).hexdigest()
+    except Exception:
+        return ""
+
+
+def _resolve_sns_cached_image_path_by_cache_key(
+    *,
+    wxid_dir: Path,
+    cache_key: str,
+    create_time: int,
+) -> Optional[str]:
+    key32 = _normalize_hex32(cache_key)
+    if not key32:
+        return None
+
+    sub = key32[:2]
+    rest = key32[2:]
+    roots = _sns_img_roots(str(wxid_dir))
+    if not roots:
+        return None
+
+    best: tuple[float, str] | None = None
+    for root_str in roots:
+        try:
+            p = Path(root_str) / sub / rest
+            if not (p.exists() and p.is_file()):
+                continue
+            st = p.stat()
+            score = abs(float(st.st_mtime) - float(create_time)) if create_time > 0 else -float(st.st_mtime)
+            if best is None or score < best[0]:
+                best = (score, str(p))
+        except Exception:
+            continue
+    return best[1] if best else None
+
+
+def _resolve_sns_cached_image_path_by_md5(
+    *,
+    wxid_dir: Path,
+    md5: str,
+    create_time: int,
+) -> Optional[str]:
+    md5_32 = _normalize_hex32(md5)
+    if not md5_32:
+        return None
+
+    sub = md5_32[:2]
+    rest = md5_32[2:]
+    roots = _sns_img_roots(str(wxid_dir))
+    if not roots:
+        return None
+
+    best: tuple[float, str] | None = None
+    for root_str in roots:
+        try:
+            p = Path(root_str) / sub / rest
+            if not (p.exists() and p.is_file()):
+                continue
+            st = p.stat()
+            score = abs(float(st.st_mtime) - float(create_time)) if create_time > 0 else -float(st.st_mtime)
+            if best is None or score < best[0]:
+                best = (score, str(p))
+        except Exception:
+            continue
+    return best[1] if best else None
+
+
+def _resolve_sns_cached_image_path(
+    *,
+    account_dir_str: str,
+    create_time: int,
+    width: int,
+    height: int,
+    idx: int,
+    total_size: int = 0,
+) -> Optional[str]:
+    total_size_i = int(total_size or 0)
+    must_match_size = width > 0 and height > 0
+    if (not must_match_size) and total_size_i <= 0:
+        return None
+
+    account_dir = Path(str(account_dir_str or "").strip())
+    if not account_dir.exists():
+        return None
+
+    wxid_dir = _resolve_account_wxid_dir(account_dir)
+    if not wxid_dir:
+        return None
+
+    mtimes, paths = _sns_img_time_index(str(wxid_dir))
+    if not mtimes:
+        return None
+
+    create_time_i = int(create_time or 0)
+    if create_time_i > 0:
+        window = 72 * 3600
+        lo = create_time_i - window
+        hi = create_time_i + window
+        l = bisect_left(mtimes, lo)
+        r = bisect_right(mtimes, hi)
+        if l >= r:
+            l = max(0, len(mtimes) - 800)
+            r = len(mtimes)
+    else:
+        l = max(0, len(mtimes) - 800)
+        r = len(mtimes)
+
+    candidates: list[tuple[float, str]] = []
+    for j in range(l, r):
+        try:
+            if create_time_i > 0:
+                candidates.append((abs(mtimes[j] - float(create_time_i)), paths[j]))
+            else:
+                candidates.append((-mtimes[j], paths[j]))
+        except Exception:
+            continue
+    candidates.sort(key=lambda x: x[0])
+
+    matched: list[tuple[int, float, str]] = []
+    for diff, pstr in candidates[:2000]:
+        try:
+            p = Path(pstr)
+            payload, media_type = _read_and_maybe_decrypt_media(p, account_dir)
+            if not payload or not str(media_type or "").startswith("image/"):
+                continue
+            if must_match_size:
+                w0, h0 = _image_size_from_bytes(payload, str(media_type or ""))
+                if (w0, h0) != (width, height):
+                    continue
+            size_diff = abs(len(payload) - total_size_i) if total_size_i > 0 else 0
+            matched.append((int(size_diff), float(diff), pstr))
+        except Exception:
+            continue
+
+    if not matched:
+        return None
+    if must_match_size:
+        matched.sort(key=lambda x: (x[0], x[1], x[2]))
+        if total_size_i > 0:
+            return matched[0][2]
+        idx0 = max(0, int(idx or 0))
+        return matched[idx0][2] if idx0 < len(matched) else None
+    if total_size_i > 0:
+        matched.sort(key=lambda x: (x[0], x[1], x[2]))
+        return matched[0][2]
+    return None
+
+
 def _format_dt(ts_seconds: Any) -> str:
    try:
        t = int(ts_seconds or 0)
@@ -1,63 +0,0 @@
-import re
-from collections.abc import MutableMapping
-
-
-def add_sns_stage_timing_headers(
-    headers: MutableMapping[str, str],
-    *,
-    source: str,
-    hit_type: str = "",
-    x_enc: str = "",
-) -> None:
-    """Inject `Server-Timing` + `Timing-Allow-Origin` for SNS media stage inspection.
-
-    The frontend can't read `<img>` response headers, but browsers expose `Server-Timing` metrics
-    via `performance.getEntriesByName(...).serverTiming` when `Timing-Allow-Origin` allows it.
-
-    This helper is intentionally side-effect free beyond mutating `headers`.
-    """
-
-    src = str(source or "").strip()
-    if not src:
-        return
-
-    ht = str(hit_type or "").strip()
-    xe = str(x_enc or "").strip()
-
-    if "Timing-Allow-Origin" not in headers:
-        headers["Timing-Allow-Origin"] = "*"
-
-    def _esc(v: str) -> str:
-        return v.replace("\\", "\\\\").replace('"', '\\"')
-
-    def _token(v: str) -> str:
-        raw = str(v or "").strip()
-        if not raw:
-            return ""
-        raw = raw.replace(" ", "_")
-        safe = re.sub(r"[^0-9A-Za-z_.-]+", "_", raw).strip("_")
-        if not safe:
-            return ""
-        return safe[:64]
-
-    parts: list[str] = []
-    src_tok = _token(src) or "unknown"
-    parts.append(f'sns_source_{src_tok};dur=0;desc="{_esc(src)}"')
-    if ht:
-        ht_tok = _token(ht)
-        if ht_tok:
-            parts.append(f'sns_hit_{ht_tok};dur=0;desc="{_esc(ht)}"')
-    if xe:
-        xe_tok = _token(xe)
-        if xe_tok:
-            parts.append(f'sns_xenc_{xe_tok};dur=0;desc="{_esc(xe)}"')
-
-    existing = str(headers.get("Server-Timing") or "").strip()
-    # Some responses may already have upstream `Server-Timing` metrics. Always append ours so
-    # the frontend can consistently read `sns_source_*` via ResourceTiming.serverTiming.
-    if existing and re.search(r"(^|,\\s*)sns_source(_|;)", existing):
-        return
-
-    combined = ", ".join(parts)
-    headers["Server-Timing"] = f"{existing}, {combined}" if existing else combined
-
@@ -0,0 +1,39 @@
+import asyncio
+import sys
+import unittest
+from pathlib import Path
+from tempfile import TemporaryDirectory
+from unittest import mock
+
+
+ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT / "src"))
+
+
+from wechat_decrypt_tool.routers import sns  # noqa: E402  pylint: disable=wrong-import-position
+
+
+class TestSnsMediaRouteWeFlowDefault(unittest.TestCase):
+    def test_route_stops_after_remote_miss_by_default(self):
+        with TemporaryDirectory() as td:
+            account_dir = Path(td) / "acc"
+            account_dir.mkdir(parents=True, exist_ok=True)
+
+            with mock.patch("wechat_decrypt_tool.routers.sns._resolve_account_dir", return_value=account_dir):
+                with mock.patch("wechat_decrypt_tool.routers.sns._try_fetch_and_decrypt_sns_remote", return_value=None):
+                    with self.assertRaises(sns.HTTPException) as ctx:
+                        asyncio.run(
+                            sns.get_sns_media(
+                                account="acc",
+                                url="https://mmsns.qpic.cn/sns/test/0",
+                                key="123",
+                                token="tkn",
+                                use_cache=1,
+                            )
+                        )
+
+        self.assertEqual(ctx.exception.status_code, 404)
+
+
+if __name__ == "__main__":
+    unittest.main()
@@ -1,40 +0,0 @@
-import sys
-import unittest
-from pathlib import Path
-
-from starlette.responses import Response
-
-
-ROOT = Path(__file__).resolve().parents[1]
-sys.path.insert(0, str(ROOT / "src"))
-
-
-from wechat_decrypt_tool.sns_stage_timing import add_sns_stage_timing_headers  # noqa: E402  pylint: disable=wrong-import-position
-
-
-class TestSnsStageServerTiming(unittest.TestCase):
-    def test_injects_server_timing_when_missing(self):
-        resp = Response(content=b"ok")
-        add_sns_stage_timing_headers(resp.headers, source="proxy")
-        st = str(resp.headers.get("Server-Timing") or "")
-        self.assertIn("sns_source_", st)
-        self.assertIn("proxy", st)
-
-    def test_appends_when_upstream_server_timing_exists(self):
-        resp = Response(content=b"ok")
-        resp.headers["Server-Timing"] = "edge;dur=1"
-        add_sns_stage_timing_headers(resp.headers, source="proxy")
-        st = str(resp.headers.get("Server-Timing") or "")
-        self.assertIn("edge;dur=1", st)
-        self.assertIn("sns_source_", st)
-
-    def test_does_not_duplicate_existing_sns_source_metric(self):
-        resp = Response(content=b"ok")
-        resp.headers["Server-Timing"] = 'sns_source_proxy;dur=0;desc="proxy"'
-        add_sns_stage_timing_headers(resp.headers, source="proxy")
-        st = str(resp.headers.get("Server-Timing") or "")
-        self.assertEqual(st.count("sns_source_"), 1)
-
-
-if __name__ == "__main__":
-    unittest.main()