fix: support biz_message_N.db filename pattern in export_messages.py

README.md

@@ -45,9 +45,37 @@ python3 export_messages.py -s "关键词"
 
 # 导出所有会话
 python3 export_messages.py --all
+
+# 按日期范围导出（支持个人聊天 + 群聊，可组合使用）
+python3 export_messages.py --all --since 2026-03-02
+python3 export_messages.py --all --since 2026-03-01 --until 2026-03-31
+
+# 只导出个人聊天和群聊（过滤公众号）
+python3 export_messages.py --all --personal
+
+# 组合：导出 3 月 2 日至今的个人聊天，仅保留有消息的会话
+python3 export_messages.py --all --personal --since 2026-03-02 -o exported_personal
+```
+
+> **注意**：微信将个人消息存储在 `message_N.db`，公众号消息存储在 `biz_message_N.db`。首次运行 `find_key_memscan.py` 时，只有已在微信中打开过的对话对应的数据库密钥会被提取到内存中。若发现导出的个人聊天记录不完整，请在微信中打开更多历史对话后重新运行密钥提取脚本，再重新解密。
+
+### 5. 合并并清理聊天记录
+
+将导出目录中的所有 txt 文件合并为单个文件，同时自动清理图片、语音、视频等消息后附带的二进制乱码：
+
+```bash
+# 合并到默认输出文件（<目录名>_merged.txt）
+python3 merge_and_clean.py exported_personal
+
+# 指定输出文件名
+python3 merge_and_clean.py exported_personal output.txt
+
+# 合并后自动删除原导出目录
+python3 merge_and_clean.py exported_personal --clean
+python3 merge_and_clean.py exported_personal output.txt --clean
 ```
 
-### 5. MCP Server（让 AI 直接查询）
+### 6. MCP Server（让 AI 直接查询）
 
 安装依赖并注册到 Claude Code：
 

export_messages.py

@@ -98,14 +98,14 @@ def resolve_username(chat_name, contacts):
 
 
 def get_all_msg_dbs(decrypted_dir):
-    """Find all message_N.db files (N = 0, 1, 2, ...)."""
+    """Find all message_N.db or biz_message_N.db files (N = 0, 1, 2, ...)."""
     import re
     msg_dir = os.path.join(decrypted_dir, "message")
     if not os.path.isdir(msg_dir):
         return []
     dbs = []
     for f in sorted(os.listdir(msg_dir)):
-        if re.match(r"^message_\d+\.db$", f):
+        if re.match(r"^(biz_)?message_\d+\.db$", f):
             dbs.append(os.path.join(msg_dir, f))
     return dbs
 
@@ -249,7 +249,7 @@ def list_conversations(msg_dbs, session_db_path, contacts):
     return results
 
 
-def export_chat(msg_dbs, username, contacts, limit=None):
+def export_chat(msg_dbs, username, contacts, limit=None, since_ts=None, until_ts=None):
     """Export messages for a specific conversation from all message DBs."""
     table = username_to_table(username)
     is_group = "@chatroom" in username
@@ -263,18 +263,32 @@ def export_chat(msg_dbs, username, contacts, limit=None):
     try:
         total = conn.execute(f"SELECT count(*) FROM [{table}]").fetchone()[0]
 
-        query = (
-            f"SELECT local_id, local_type, create_time, real_sender_id, "
-            f"message_content, source FROM [{table}] ORDER BY create_time ASC"
-        )
-        if limit:
+        where_clauses = []
+        params = []
+        if since_ts is not None:
+            where_clauses.append("create_time >= ?")
+            params.append(since_ts)
+        if until_ts is not None:
+            where_clauses.append("create_time <= ?")
+            params.append(until_ts)
+        where_sql = ("WHERE " + " AND ".join(where_clauses)) if where_clauses else ""
+
+        if limit and not where_clauses:
             query = (
                 f"SELECT * FROM (SELECT local_id, local_type, create_time, "
                 f"real_sender_id, message_content, source FROM [{table}] "
                 f"ORDER BY create_time DESC LIMIT {limit}) ORDER BY create_time ASC"
             )
+            rows = conn.execute(query).fetchall()
+        else:
+            query = (
+                f"SELECT local_id, local_type, create_time, real_sender_id, "
+                f"message_content, source FROM [{table}] {where_sql} ORDER BY create_time ASC"
+            )
+            if limit:
+                query += f" LIMIT {limit}"
+            rows = conn.execute(query, params).fetchall()
 
-        rows = conn.execute(query).fetchall()
         lines = [format_message(r, is_group, contacts) for r in rows]
 
         display_name = contacts.get(username, username)
@@ -297,12 +311,15 @@ def safe_filename(display_name, username):
     return name
 
 
-def export_to_file(msg_dbs, username, output_dir, contacts, limit=None):
+def export_to_file(msg_dbs, username, output_dir, contacts, limit=None, since_ts=None, until_ts=None):
     """Export messages to a text file named by display name."""
-    lines, info = export_chat(msg_dbs, username, contacts, limit)
+    lines, info = export_chat(msg_dbs, username, contacts, limit, since_ts, until_ts)
     if lines is None:
         return False, info
 
+    if not lines:
+        return False, f"skipped (no messages in range) | {info}"
+
     os.makedirs(output_dir, exist_ok=True)
 
     display_name = contacts.get(username, "")
@@ -340,8 +357,31 @@ def main():
     parser.add_argument(
         "-s", "--search", help="Search keyword across all conversations",
     )
+    parser.add_argument(
+        "--since", help="Start date (inclusive), format: YYYY-MM-DD",
+    )
+    parser.add_argument(
+        "--until", help="End date (inclusive), format: YYYY-MM-DD (default: today)",
+    )
+    parser.add_argument(
+        "--personal", action="store_true",
+        help="Only export personal chats and groups (exclude public accounts starting with gh_)",
+    )
     args = parser.parse_args()
 
+    # Parse date range into Unix timestamps
+    since_ts = None
+    until_ts = None
+    if args.since:
+        since_ts = int(datetime.strptime(args.since, "%Y-%m-%d").timestamp())
+    if args.until:
+        from datetime import timedelta
+        until_ts = int((datetime.strptime(args.until, "%Y-%m-%d") + timedelta(days=1)).timestamp()) - 1
+    elif args.since:
+        # if --since given but no --until, default until = end of today
+        from datetime import timedelta
+        until_ts = int((datetime.now().replace(hour=23, minute=59, second=59, microsecond=0)).timestamp())
+
     # Load databases
     msg_dbs = get_all_msg_dbs(args.dir)
     if not msg_dbs:
@@ -371,11 +411,20 @@ def main():
                 ).fetchone()[0]
                 if not exists:
                     continue
+                extra_where = []
+                extra_params = [f"%{args.search}%"]
+                if since_ts is not None:
+                    extra_where.append("create_time >= ?")
+                    extra_params.append(since_ts)
+                if until_ts is not None:
+                    extra_where.append("create_time <= ?")
+                    extra_params.append(until_ts)
+                date_sql = (" AND " + " AND ".join(extra_where)) if extra_where else ""
                 rows = conn.execute(
                     f"SELECT local_id, local_type, create_time, real_sender_id, "
                     f"message_content, source FROM [{table}] "
-                    f"WHERE message_content LIKE ? ORDER BY create_time DESC LIMIT 10",
-                    (f"%{args.search}%",),
+                    f"WHERE message_content LIKE ?{date_sql} ORDER BY create_time DESC LIMIT 10",
+                    extra_params,
                 ).fetchall()
                 if rows:
                     display = contacts.get(username, username)
@@ -400,7 +449,7 @@ def main():
             display = contacts.get(username, username)
             print(f"[*] Matched '{args.chat}' -> {display} ({username})")
 
-        lines, info = export_chat(msg_dbs, username, contacts, args.limit)
+        lines, info = export_chat(msg_dbs, username, contacts, args.limit, since_ts, until_ts)
         if lines is None:
             print(f"[-] {info}")
             sys.exit(1)
@@ -409,23 +458,29 @@ def main():
         for line in lines:
             print(line)
 
-        success, result_info = export_to_file(msg_dbs, username, args.output, contacts, args.limit)
+        success, result_info = export_to_file(msg_dbs, username, args.output, contacts, args.limit, since_ts, until_ts)
         print(f"\n[*] Saved: {result_info}")
 
     elif args.all:
         # Export all conversations
         convos = list_conversations(msg_dbs, session_db, contacts)
         os.makedirs(args.output, exist_ok=True)
         exported = 0
+        skipped = 0
         for c in convos:
             if not c["has_msgs"]:
                 continue
+            if args.personal and c["username"].startswith("gh_"):
+                skipped += 1
+                continue
             success, info = export_to_file(
-                msg_dbs, c["username"], args.output, contacts, args.limit,
+                msg_dbs, c["username"], args.output, contacts, args.limit, since_ts, until_ts,
             )
             if success:
                 print(f"  ✅ {info}")
                 exported += 1
+        if skipped:
+            print(f"[*] Skipped {skipped} public accounts (gh_*)")
         print(f"\n[*] Exported {exported} conversations to {args.output}/")
 
     else:

merge_and_clean.py

@@ -0,0 +1,148 @@
+#!/usr/bin/env python3
+"""
+合并导出的聊天记录 txt 文件并清理乱码。
+
+用法：
+    python3 merge_and_clean.py <导出目录> [输出文件] [--clean]
+
+示例：
+    python3 merge_and_clean.py exported_personal
+    python3 merge_and_clean.py exported_personal merged.txt
+    python3 merge_and_clean.py exported_personal --clean
+    python3 merge_and_clean.py exported_personal merged.txt --clean
+
+选项：
+    --clean    合并完成后删除原导出目录
+"""
+
+import os
+import re
+import shutil
+import sys
+
+TIMESTAMP_RE = re.compile(r'^\[\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}\]')
+
+
+def merge(src_dir, out_file):
+    files = sorted(f for f in os.listdir(src_dir) if f.endswith('.txt'))
+    if not files:
+        print(f"[-] 目录 {src_dir} 中没有找到 txt 文件")
+        sys.exit(1)
+
+    print(f"[*] 合并 {len(files)} 个文件...")
+    with open(out_file, 'wb') as out:
+        for fname in files:
+            with open(os.path.join(src_dir, fname), 'rb') as f:
+                out.write(f.read())
+            out.write(b'\n' + b'=' * 80 + b'\n\n')
+
+    print(f"[*] 已合并到 {out_file}")
+
+
+def clean_image_garbage(content: bytes) -> tuple[bytes, int]:
+    """清理 [image]/[voice]/[video]/[emoji]/[type:xxx] 等标签后的二进制乱码。"""
+    # 先处理 "[tag] (乱码)" 有括号形式
+    cleaned, n1 = re.subn(rb'\[(?!\d{4}-\d{2}-\d{2})([^\]]+)\] \([^\n]*\)', rb'[\1]', content)
+    # 再处理 "[image]乱码到行尾" 无括号形式（[image] 后仍有非换行内容）
+    cleaned, n2 = re.subn(rb'\[image\][^\n]+', b'[image]', cleaned)
+    return cleaned, n1 + n2
+
+
+def clean_garbage_lines(content: str) -> tuple[str, int]:
+    """删除含 Unicode 替换字符（乱码）的整条消息块。"""
+    lines = content.split('\n')
+    output = []
+    i = 0
+    removed = 0
+
+    while i < len(lines):
+        line = lines[i]
+
+        # 保留真正的注释行（# 开头加空格）、纯等号分隔线、空行
+        is_comment = line.startswith('# ')
+        is_separator = line.startswith('=') and all(c == '=' for c in line.strip())
+        if is_comment or is_separator or line.strip() == '':
+            output.append(line)
+            i += 1
+            continue
+
+        # 消息块：以时间戳开头，收集到下一条消息/注释/分隔线为止
+        if TIMESTAMP_RE.match(line):
+            block = [line]
+            j = i + 1
+            while j < len(lines):
+                nxt = lines[j]
+                if (TIMESTAMP_RE.match(nxt) or
+                        nxt.startswith('# ') or
+                        (nxt.startswith('=') and all(c == '=' for c in nxt.strip()))):
+                    break
+                block.append(nxt)
+                j += 1
+
+            block_text = '\n'.join(block)
+            if '\ufffd' in block_text:
+                removed += 1
+            else:
+                output.append(block_text)
+            i = j
+            continue
+
+        # 其他孤立行（不以时间戳开头）
+        if '\ufffd' in line:
+            removed += 1
+        else:
+            output.append(line)
+        i += 1
+
+    return '\n'.join(output), removed
+
+
+def main():
+    args = sys.argv[1:]
+    if not args:
+        print(__doc__)
+        sys.exit(1)
+
+    src_dir = args[0]
+    if not os.path.isdir(src_dir):
+        print(f"[-] 目录不存在：{src_dir}")
+        sys.exit(1)
+
+    do_clean = '--clean' in args
+    positional = [a for a in args[1:] if not a.startswith('--')]
+    out_file = positional[0] if positional else src_dir.rstrip('/') + '_merged.txt'
+
+    # 步骤 1：合并
+    merge(src_dir, out_file)
+
+    # 步骤 2：清理标签后的二进制乱码
+    with open(out_file, 'rb') as f:
+        content = f.read()
+
+    cleaned_bytes, n_tag = clean_image_garbage(content)
+
+    with open(out_file, 'wb') as f:
+        f.write(cleaned_bytes)
+
+    print(f"[*] 清理标签乱码：{n_tag} 处")
+
+    # 步骤 3：清理含替换字符的消息块
+    with open(out_file, 'r', encoding='utf-8', errors='replace') as f:
+        text = f.read()
+
+    cleaned_text, n_msg = clean_garbage_lines(text)
+
+    with open(out_file, 'w', encoding='utf-8') as f:
+        f.write(cleaned_text)
+
+    print(f"[*] 清理乱码消息块：{n_msg} 条")
+
+    if do_clean:
+        shutil.rmtree(src_dir)
+        print(f"[*] 已删除原目录：{src_dir}")
+
+    print(f"[✓] 完成：{out_file}")
+
+
+if __name__ == '__main__':
+    main()