feat(video-pipeline): 用 ffprobe 实际测量音视频时长并统一字段名为 script

将项目中的 `narration` 字段统一重命名为 `script`，并新增 `getAudioDurationSec` 函数通过 `ffprobe` 实际测量音频和视频文件的时长，替代 Manifest 中的估计值，提高时间线组装的准确性。同时优化字幕逻辑，仅在有 TTS 音频时调整视频速度。

.claude/skills/video-from-script/scripts/capcut_assemble.js

@@ -123,6 +123,19 @@ async function batchUploadToOSS(inputDir, files) {
   return urls
 }
 
+function getAudioDurationSec(filePath) {
+  return new Promise((resolve) => {
+    execFile('ffprobe', [
+      '-v', 'quiet', '-show_entries', 'format=duration',
+      '-of', 'csv=p=0', filePath
+    ], (err, stdout) => {
+      if (err) { resolve(null); return }
+      const dur = parseFloat(stdout.trim())
+      resolve(dur > 0 ? dur : null)
+    })
+  })
+}
+
 // ============================================================================
 // 主流程
 // ============================================================================
@@ -131,7 +144,7 @@ function buildTimeline(items, defaultDurationUs) {
   // 音频为主轴，视频调速适配（≤2x 加速，>2x 截断）
   let offset = 0
   return items.map(item => {
-    const audioDur = (item.audioDuration != null) ? item.audioDuration * US : (item.duration != null) ? item.duration * US : 0
+    const audioDur = (item.audioDuration != null) ? item.audioDuration * US : 0
     const videoDur = (item.videoDuration != null) ? item.videoDuration * US : 0
     // 无 TTS：用视频时长或固定时长
     if (audioDur <= 0) {
@@ -193,10 +206,37 @@ async function assemble(args) {
 
   if (items.length === 0) throw new Error('没有可用的素材文件')
 
-  // 统一时间线：由 duration 驱动（TTS 音频时长）或 fallback 到固定时长
+  // 用 ffprobe 测量实际音频/视频时长，替代 manifest 中的估计值
+  let audioMeasured = 0, videoMeasured = 0
+  for (const item of items) {
+    // 测量 TTS 音频实际时长
+    if (item.audio && !item.audio.startsWith('http')) {
+      const audioPath = path.isAbsolute(item.audio)
+        ? item.audio
+        : path.resolve(inputDir, item.audio)
+      if (fs.existsSync(audioPath)) {
+        const actualDur = await getAudioDurationSec(audioPath)
+        if (actualDur != null) { item.audioDuration = actualDur; audioMeasured++ }
+      }
+    }
+    // 测量视频文件实际时长（生成器返回的是硬编码常量，不准确）
+    if (item.video) {
+      const videoPath = path.isAbsolute(item.video)
+        ? item.video
+        : path.resolve(inputDir, item.video)
+      if (fs.existsSync(videoPath)) {
+        const actualDur = await getAudioDurationSec(videoPath) // ffprobe 对音视频通用
+        if (actualDur != null) { item.videoDuration = actualDur; videoMeasured++ }
+      }
+    }
+  }
+  if (audioMeasured > 0 || videoMeasured > 0) {
+    console.log(`  实际时长测量: 音频 ${audioMeasured} 个, 视频 ${videoMeasured} 个`)
+  }
+
   const timeline = buildTimeline(items, defaultDurationUs)
   const totalDurationUs = timeline.length > 0 ? timeline[timeline.length - 1].end : 0
-  const hasTTS = items.some(item => item.audio && (item.audioDuration != null || item.duration != null))
+  const hasTTS = items.some(item => item.audio && item.audioDuration != null)
 
   console.log(`\nCapCut 成片组装`)
   console.log(`  模式: ${mode}  画幅: ${format} (${width}x${height})`)
@@ -280,7 +320,7 @@ async function assemble(args) {
           if (manifestFile) {
             try {
               const m = JSON.parse(fs.readFileSync(manifestFile, 'utf-8'))
-              const mi = m.items.find(i => i.id === item.id || i.narration === (item.narration || item.text) || i.text === (item.narration || item.text))
+              const mi = m.items.find(i => i.id === item.id || i.script === item.script || i.text === item.text)
               if (mi) { mi.videoUrl = url; fs.writeFileSync(manifestFile, JSON.stringify(m, null, 2)) }
             } catch (_) {}
           }
@@ -316,7 +356,7 @@ async function assemble(args) {
 
   // -- 添加字幕 --
   step++; console.log(`[${step}/${totalSteps}] 添加字幕...`)
-  if (subtitles === 'true' && items.some(i => i.narration || i.text)) {
+  if (subtitles === 'true' && items.some(i => i.script || i.text)) {
     await addSubtitles(draftUrl, items, timeline, subtitleStyle)
   } else {
     console.log('   跳过')
@@ -669,7 +709,7 @@ async function addSubtitles(draftUrl, items, timeline, style = {}) {
 
   for (let i = 0; i < items.length; i++) {
     const item = items[i]
-    const text = item.narration || item.text || item.caption || ''
+    const text = item.script || item.text || item.caption || ''
     if (!text) continue
 
     const tl = timeline[i]