WhatsApp-Chatexport einlesen und Videos chronologisch ordnen

Erkennt iPhone- und Android-Exporte auf Deutsch und Englisch, ordnet Videos nach Zeitstempel statt nach Chatreihenfolge, meldet fehlende Dateien, markiert doppelt gepostete Videos und filtert auf Montag bis Freitag der Vorwoche. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-29 09:25:22 +00:00
1 parent f7d249f28b
commit 899b212cba
@@ -0,0 +1,209 @@
+// Liest WhatsApp-Chatexporte (iPhone und Android, deutsch und englisch) und
+// ordnet die Videos des Exports chronologisch. Läuft im Browser und in Node.
+
+// Unsichtbare Richtungszeichen, die WhatsApp vor Zeilen und Namen setzt.
+const INVISIBLE = /[\u200e\u200f\u202a-\u202e\u2066-\u2069]/g;
+const ODD_SPACES = /[\u00a0\u202f]/g;
+
+// iPhone: "[24.09.26, 16:36:19] Name: Text"
+const IOS_HEADER =
+  /^\[(\d{1,2})([./])(\d{1,2})\2(\d{2,4}),? (\d{1,2}):(\d{2})(?::(\d{2}))?(?: ?([AP]M))?\] (.*)$/i;
+// Android: "24.09.26, 16:36 - Name: Text"
+const ANDROID_HEADER =
+  /^(\d{1,2})([./])(\d{1,2})\2(\d{2,4}),? (\d{1,2}):(\d{2})(?::(\d{2}))?(?: ?([AP]M))? - (.*)$/i;
+
+// "<Anhang: 00001338-VIDEO-2026-09-24-16-36-19.mp4>" bzw. "<attached: …>"
+const TAGGED_ATTACHMENT = /<[^<>:]{1,40}:\s*([^<>]+?\.[A-Za-z0-9]{2,5})>/g;
+// "VID-20260924-WA0012.mp4 (Datei angehängt)" bzw. "(file attached)"
+const ANDROID_ATTACHMENT = /([^\s<>]+\.[A-Za-z0-9]{2,5}) \((?:Datei angehängt|file attached)\)/g;
+
+const VIDEO_EXTENSION = /\.(mp4|mov|m4v|3gp|avi|mkv|webm)$/i;
+
+const pad = (value) => String(value).padStart(2, '0');
+
+function cleanLine(line) {
+  return line.replace(INVISIBLE, '').replace(ODD_SPACES, ' ');
+}
+
+export function isVideoFile(name) {
+  return VIDEO_EXTENSION.test(name);
+}
+
+export function baseName(path) {
+  return path.split(/[\\/]/).pop();
+}
+
+// Bei Datumsangaben mit Schrägstrich ist die Reihenfolge nicht eindeutig
+// (US: M/D, sonst D/M). Ein Tag über 12 irgendwo im Chat entscheidet.
+function detectDayFirst(lines) {
+  for (const line of lines) {
+    const match = line.match(IOS_HEADER) || line.match(ANDROID_HEADER);
+    if (match && match[2] === '/' && Number(match[1]) > 12) return true;
+  }
+  return false;
+}
+
+function toTimestamp(match, dayFirst) {
+  const [, first, separator, second, rawYear, rawHour, minute, second_ = '0', meridiem] = match;
+  let day = Number(first);
+  let month = Number(second);
+  if (separator === '/' && !dayFirst) [day, month] = [month, day];
+  let year = Number(rawYear);
+  if (year < 100) year += 2000;
+  let hour = Number(rawHour);
+  if (meridiem) {
+    const pm = meridiem.toUpperCase() === 'PM';
+    if (hour === 12) hour = pm ? 12 : 0;
+    else if (pm) hour += 12;
+  }
+  if (month < 1 || month > 12 || day < 1 || day > 31 || hour > 23) return null;
+  return `${year}-${pad(month)}-${pad(day)}T${pad(hour)}:${minute}:${pad(second_)}`;
+}
+
+function splitSender(rest) {
+  const index = rest.indexOf(': ');
+  if (index > 0) return { sender: rest.slice(0, index).trim(), text: rest.slice(index + 2) };
+  if (rest.endsWith(':')) return { sender: rest.slice(0, -1).trim(), text: '' };
+  // Systemmeldung ohne Absender, z. B. "Max hat die Gruppe verlassen"
+  return { sender: null, text: rest };
+}
+
+export function findAttachments(text) {
+  const names = [];
+  for (const pattern of [TAGGED_ATTACHMENT, ANDROID_ATTACHMENT]) {
+    for (const match of text.matchAll(pattern)) names.push(match[1].trim());
+  }
+  return names;
+}
+
+function captionOf(text) {
+  return text.replace(TAGGED_ATTACHMENT, '').replace(ANDROID_ATTACHMENT, '').trim();
+}
+
+export function parseChat(rawText) {
+  const lines = rawText.replace(/\r\n?/g, '\n').split('\n').map(cleanLine);
+  const dayFirst = detectDayFirst(lines);
+  const messages = [];
+
+  for (const line of lines) {
+    const match = line.match(IOS_HEADER) || line.match(ANDROID_HEADER);
+    const timestamp = match && toTimestamp(match, dayFirst);
+    if (timestamp) {
+      messages.push({ index: messages.length, timestamp, ...splitSender(match[9]) });
+    } else if (messages.length > 0) {
+      messages[messages.length - 1].text += `\n${line}`;
+    }
+  }
+
+  return messages.map((message) => ({
+    ...message,
+    text: message.text.trim(),
+    attachments: findAttachments(message.text),
+  }));
+}
+
+// Fallback für Videos ohne Chatzeile: der Dateiname trägt den Zeitpunkt.
+export function timestampFromFileName(name) {
+  const ios = name.match(/(\d{4})-(\d{2})-(\d{2})-(\d{2})-(\d{2})-(\d{2})\.\w+$/);
+  if (ios) return `${ios[1]}-${ios[2]}-${ios[3]}T${ios[4]}:${ios[5]}:${ios[6]}`;
+  const android = name.match(/-(\d{4})(\d{2})(\d{2})-WA\d+/);
+  if (android) return `${android[1]}-${android[2]}-${android[3]}T00:00:00`;
+  return null;
+}
+
+function compareClips(a, b) {
+  return (
+    a.timestamp.localeCompare(b.timestamp) ||
+    a.order - b.order ||
+    a.file.localeCompare(b.file)
+  );
+}
+
+// Doppelt gepostete Videos sind bytegleich; die Dateigröße genügt als Merkmal.
+function markDuplicates(clips) {
+  const firstBySize = new Map();
+  for (const clip of clips) {
+    if (!clip.size) continue;
+    const first = firstBySize.get(clip.size);
+    if (first) clip.duplicateOf = first.file;
+    else firstBySize.set(clip.size, clip);
+  }
+}
+
+// files: [{ name, size }] aus dem Export (Pfade erlaubt).
+// Liefert die Videos chronologisch, dazu im Chat genannte, aber fehlende Dateien.
+export function collectClips(messages, files) {
+  const videos = new Map();
+  for (const file of files) {
+    if (isVideoFile(file.name)) videos.set(baseName(file.name), file);
+  }
+
+  const clips = [];
+  const missing = [];
+  const used = new Set();
+
+  for (const message of messages) {
+    for (const attachment of message.attachments) {
+      if (!isVideoFile(attachment)) continue;
+      const name = baseName(attachment);
+      const file = videos.get(name);
+      if (!file) {
+        missing.push({ file: name, timestamp: message.timestamp, sender: message.sender });
+        continue;
+      }
+      if (used.has(name)) continue;
+      used.add(name);
+      clips.push({
+        file: name,
+        path: file.name,
+        size: file.size ?? 0,
+        timestamp: message.timestamp,
+        sender: message.sender,
+        caption: captionOf(message.text),
+        order: message.index,
+      });
+    }
+  }
+
+  for (const [name, file] of videos) {
+    if (used.has(name)) continue;
+    const timestamp = timestampFromFileName(name);
+    if (!timestamp) continue;
+    clips.push({
+      file: name,
+      path: file.name,
+      size: file.size ?? 0,
+      timestamp,
+      sender: null,
+      caption: '',
+      order: Number.MAX_SAFE_INTEGER,
+    });
+  }
+
+  clips.sort(compareClips);
+  markDuplicates(clips);
+  return { clips, missing };
+}
+
+// from/to als "JJJJ-MM-TT", beide Tage eingeschlossen.
+export function clipsInRange(clips, from, to) {
+  return clips.filter((clip) => {
+    const day = clip.timestamp.slice(0, 10);
+    return (!from || day >= from) && (!to || day <= to);
+  });
+}
+
+export function toDateString(date) {
+  return `${date.getFullYear()}-${pad(date.getMonth() + 1)}-${pad(date.getDate())}`;
+}
+
+// Montag bis Freitag der Kalenderwoche vor der von `today`.
+export function previousWorkWeek(today = new Date()) {
+  const date = new Date(today.getFullYear(), today.getMonth(), today.getDate());
+  const daysSinceMonday = (date.getDay() + 6) % 7;
+  const monday = new Date(date);
+  monday.setDate(date.getDate() - daysSinceMonday - 7);
+  const friday = new Date(monday);
+  friday.setDate(monday.getDate() + 4);
+  return { from: toDateString(monday), to: toDateString(friday) };
+}