Three new flags for export_all_chats.py:
- -i / --incremental: reads existing JSON, appends only new messages
(deduplicates by local_id, preserves transcription field on merge)
- --start / --end: filter messages by date range (YYYY-MM-DD or timestamp)
passes start_ts/end_ts directly to mcp_server._query_messages
- --dry-run: preview mode (shows counts without writing files)
Voice transcription in incremental mode only processes newly appended
voice messages — existing transcribed entries are untouched.
* Add export_all_chats.py: batch export all WeChat chats to JSON
Mirrors export_chat.py functionality to export every chat in the
decrypted WeChat database at once. Output format is byte-for-byte
identical to export_chat.py.
Usage: python3 export_all_chats.py [output_dir]
- Reads all sessions from decrypted/session/session.db
- Uses the same content extraction pipeline as export_chat.py:
_resolve_sender, _extract_content, _msg_type_str, sticker/video/system
- Outputs group_<display>.json or single_<display>.json with same schema
- Progress reporting every 100 exports with summary at end
* refactor: 将 7 个重复 helper 函数提取到 chat_export_helpers.py
根据 PR #77 review 反馈，将 export_chat.py 和 export_all_chats.py 中
逐字复制的消息格式化函数提取到共享模块 chat_export_helpers.py。
提取的函数:
MSG_TYPE_MAP, _msg_type_str, _resolve_sender,
_decode_sticker_desc, _format_sticker_message,
_format_system_message, _format_video_message, _extract_content
两个导出脚本现在从 chat_export_helpers import 所需函数，
消除代码漂移风险。
Davy