fix(chat): 群聊发送者解析支持 alias 并避免覆盖 sender_username

- _split_group_sender_prefix 支持传入已知 sender（username/alias），弱特征场景避免误拆前缀

- 群聊消息按需读取 contact.db 的 alias 辅助解析昵称:\n前缀（带缓存）

- 仅在 sender 缺失时回填前缀，避免昵称覆盖 sender id

src/wechat_decrypt_tool/chat_export_service.py

@@ -833,15 +833,6 @@ def _iter_rows_for_conversation(
                     raw_text = _decode_message_content(r["compress_content"], r["message_content"]).strip()
 
                     is_group = bool(conv_username.endswith("@chatroom"))
-                    sender_prefix = ""
-                    if is_group and raw_text and (not raw_text.startswith("<")) and (not raw_text.startswith('"<')):
-                        sender_prefix, raw_text = _split_group_sender_prefix(raw_text)
-                    if is_group and sender_prefix:
-                        sender_username = sender_prefix
-                    if is_group and raw_text and (raw_text.startswith("<") or raw_text.startswith('"<')):
-                        xml_sender = _extract_sender_from_group_xml(raw_text)
-                        if xml_sender:
-                            sender_username = xml_sender
 
                     if is_sent:
                         sender_username = account_wxid
@@ -881,8 +872,21 @@ def _parse_message_for_export(
     is_group: bool,
     resource_conn: Optional[sqlite3.Connection],
     resource_chat_id: Optional[int],
+    sender_alias: str = "",
 ) -> dict[str, Any]:
     raw_text = row.raw_text or ""
+    sender_username = str(row.sender_username or "").strip()
+
+    if is_group and raw_text and (not raw_text.startswith("<")) and (not raw_text.startswith('"<')):
+        sender_prefix, raw_text = _split_group_sender_prefix(raw_text, sender_username, sender_alias)
+        if sender_prefix and (not sender_username):
+            sender_username = sender_prefix
+
+    if is_group and raw_text and (raw_text.startswith("<") or raw_text.startswith('"<')):
+        xml_sender = _extract_sender_from_group_xml(raw_text)
+        if xml_sender:
+            sender_username = xml_sender
+
     local_type = int(row.local_type or 0)
     is_sent = bool(row.is_sent)
 
@@ -1152,7 +1156,7 @@ def _parse_message_for_export(
         "type": local_type,
         "renderType": render_type,
         "isSent": bool(is_sent),
-        "senderUsername": row.sender_username,
+        "senderUsername": sender_username,
         "conversationUsername": conv_username,
         "isGroup": bool(is_group),
         "content": content_text,
@@ -1216,6 +1220,38 @@ def _write_conversation_json(
     arcname = f"{conv_dir}/messages.json"
     exported = 0
 
+    contact_conn: Optional[sqlite3.Connection] = None
+    alias_cache: dict[str, str] = {}
+    if conv_is_group:
+        try:
+            contact_db_path = account_dir / "contact.db"
+            if contact_db_path.exists():
+                contact_conn = sqlite3.connect(str(contact_db_path))
+        except Exception:
+            contact_conn = None
+
+    def lookup_alias(username: str) -> str:
+        u = str(username or "").strip()
+        if not u or contact_conn is None:
+            return ""
+        if u in alias_cache:
+            return alias_cache[u]
+
+        alias = ""
+        try:
+            r = contact_conn.execute("SELECT alias FROM contact WHERE username = ? LIMIT 1", (u,)).fetchone()
+            if r is not None and r[0] is not None:
+                alias = str(r[0] or "").strip()
+            if not alias:
+                r = contact_conn.execute("SELECT alias FROM stranger WHERE username = ? LIMIT 1", (u,)).fetchone()
+                if r is not None and r[0] is not None:
+                    alias = str(r[0] or "").strip()
+        except Exception:
+            alias = ""
+
+        alias_cache[u] = alias
+        return alias
+
     # NOTE: Do not keep an entry handle opened while also writing other entries (avatars/media).
     # zipfile forbids interleaving writes; stream to a temp file then add it to zip at the end.
     with tempfile.TemporaryDirectory(prefix="wechat_chat_export_") as tmp_dir:
@@ -1263,12 +1299,28 @@ def _write_conversation_json(
                 local_types=local_types,
             ):
                 scanned += 1
+
+                sender_alias = ""
+                if conv_is_group and row.raw_text and (not row.raw_text.startswith("<")) and (not row.raw_text.startswith('"<')):
+                    sep = row.raw_text.find(":\n")
+                    if sep > 0:
+                        prefix = row.raw_text[:sep].strip()
+                        su = str(row.sender_username or "").strip()
+                        if prefix and su and prefix != su:
+                            strong_hint = prefix.startswith("wxid_") or prefix.endswith("@chatroom") or "@" in prefix
+                            if not strong_hint:
+                                body_probe = row.raw_text[sep + 2 :].lstrip("\n").lstrip()
+                                body_is_xml = body_probe.startswith("<") or body_probe.startswith('"<')
+                                if not body_is_xml:
+                                    sender_alias = lookup_alias(su)
+
                 msg = _parse_message_for_export(
                     row=row,
                     conv_username=conv_username,
                     is_group=conv_is_group,
                     resource_conn=resource_conn,
                     resource_chat_id=resource_chat_id,
+                    sender_alias=sender_alias,
                 )
                 if want_types:
                     rt_key = _normalize_render_type_key(msg.get("renderType"))
@@ -1326,6 +1378,12 @@ def _write_conversation_json(
             tw.flush()
 
         zf.write(str(tmp_path), arcname)
+    if contact_conn is not None:
+        try:
+            contact_conn.close()
+        except Exception:
+            pass
+
     return exported
 
 
@@ -1360,6 +1418,38 @@ def _write_conversation_txt(
     arcname = f"{conv_dir}/messages.txt"
     exported = 0
 
+    contact_conn: Optional[sqlite3.Connection] = None
+    alias_cache: dict[str, str] = {}
+    if conv_is_group:
+        try:
+            contact_db_path = account_dir / "contact.db"
+            if contact_db_path.exists():
+                contact_conn = sqlite3.connect(str(contact_db_path))
+        except Exception:
+            contact_conn = None
+
+    def lookup_alias(username: str) -> str:
+        u = str(username or "").strip()
+        if not u or contact_conn is None:
+            return ""
+        if u in alias_cache:
+            return alias_cache[u]
+
+        alias = ""
+        try:
+            r = contact_conn.execute("SELECT alias FROM contact WHERE username = ? LIMIT 1", (u,)).fetchone()
+            if r is not None and r[0] is not None:
+                alias = str(r[0] or "").strip()
+            if not alias:
+                r = contact_conn.execute("SELECT alias FROM stranger WHERE username = ? LIMIT 1", (u,)).fetchone()
+                if r is not None and r[0] is not None:
+                    alias = str(r[0] or "").strip()
+        except Exception:
+            alias = ""
+
+        alias_cache[u] = alias
+        return alias
+
     # Same as JSON: write to temp file first to avoid zip interleaving writes.
     with tempfile.TemporaryDirectory(prefix="wechat_chat_export_") as tmp_dir:
         tmp_path = Path(tmp_dir) / "messages.txt"
@@ -1391,12 +1481,27 @@ def _write_conversation_txt(
                 local_types=local_types,
             ):
                 scanned += 1
+                sender_alias = ""
+                if conv_is_group and row.raw_text and (not row.raw_text.startswith("<")) and (not row.raw_text.startswith('"<')):
+                    sep = row.raw_text.find(":\n")
+                    if sep > 0:
+                        prefix = row.raw_text[:sep].strip()
+                        su = str(row.sender_username or "").strip()
+                        if prefix and su and prefix != su:
+                            strong_hint = prefix.startswith("wxid_") or prefix.endswith("@chatroom") or "@" in prefix
+                            if not strong_hint:
+                                body_probe = row.raw_text[sep + 2 :].lstrip("\n").lstrip()
+                                body_is_xml = body_probe.startswith("<") or body_probe.startswith('"<')
+                                if not body_is_xml:
+                                    sender_alias = lookup_alias(su)
+
                 msg = _parse_message_for_export(
                     row=row,
                     conv_username=conv_username,
                     is_group=conv_is_group,
                     resource_conn=resource_conn,
                     resource_chat_id=resource_chat_id,
+                    sender_alias=sender_alias,
                 )
                 if want_types:
                     rt_key = _normalize_render_type_key(msg.get("renderType"))
@@ -1449,6 +1554,12 @@ def _write_conversation_txt(
             tw.flush()
 
         zf.write(str(tmp_path), arcname)
+    if contact_conn is not None:
+        try:
+            contact_conn.close()
+        except Exception:
+            pass
+
     return exported
 
 

src/wechat_decrypt_tool/chat_helpers.py

@@ -694,7 +694,11 @@ def _infer_transfer_status_text(
     return "转账"
 
 
-def _split_group_sender_prefix(text: str) -> tuple[str, str]:
+def _split_group_sender_prefix(
+    text: str,
+    known_sender_username: str = "",
+    known_sender_alias: str = "",
+) -> tuple[str, str]:
     if not text:
         return "", text
     sep = text.find(":\n")
@@ -706,7 +710,21 @@ def _split_group_sender_prefix(text: str) -> tuple[str, str]:
         return "", text
     if re.search(r"\s", prefix):
         return "", text
-    if prefix.startswith("wxid_") or prefix.endswith("@chatroom") or "@" in prefix:
+
+    strong_hint = prefix.startswith("wxid_") or prefix.endswith("@chatroom") or "@" in prefix
+    probe = body.lstrip()
+    body_is_xml = probe.startswith("<") or probe.startswith('"<')
+
+    known_values = {str(known_sender_username or "").strip(), str(known_sender_alias or "").strip()}
+    known_values.discard("")
+    if known_values:
+        if prefix in known_values:
+            return prefix, body
+        if strong_hint or body_is_xml:
+            return prefix, body
+        return "", text
+
+    if strong_hint or body_is_xml:
         return prefix, body
     return "", text
 
@@ -1013,7 +1031,7 @@ def _build_latest_message_preview(
     raw_text = (raw_text or "").strip()
     sender_prefix = ""
     if is_group and raw_text and (not raw_text.startswith("<")) and (not raw_text.startswith('"<')):
-        sender_prefix, raw_text = _split_group_sender_prefix(raw_text)
+        sender_prefix, raw_text = _split_group_sender_prefix(raw_text, sender_username)
     if is_group and (not sender_prefix) and sender_username:
         sender_prefix = str(sender_username).strip()
 
@@ -1400,7 +1418,7 @@ def _row_to_search_hit(
 
     sender_prefix = ""
     if is_group and raw_text and (not raw_text.startswith("<")) and (not raw_text.startswith('"<')):
-        sender_prefix, raw_text = _split_group_sender_prefix(raw_text)
+        sender_prefix, raw_text = _split_group_sender_prefix(raw_text, sender_username)
 
     if is_group and sender_prefix:
         sender_username = sender_prefix

src/wechat_decrypt_tool/routers/chat.py

@@ -111,6 +111,33 @@ def _normalize_chat_source(value: Optional[str]) -> str:
     raise HTTPException(status_code=400, detail="Invalid source, use 'decrypted' or 'realtime'.")
 
 
+def _lookup_contact_alias(
+    conn: Optional[sqlite3.Connection],
+    cache: dict[str, str],
+    username: str,
+) -> str:
+    u = str(username or "").strip()
+    if not u or conn is None:
+        return ""
+    if u in cache:
+        return cache[u]
+
+    alias = ""
+    try:
+        r = conn.execute("SELECT alias FROM contact WHERE username = ? LIMIT 1", (u,)).fetchone()
+        if r is not None and r[0] is not None:
+            alias = str(r[0] or "").strip()
+        if not alias:
+            r = conn.execute("SELECT alias FROM stranger WHERE username = ? LIMIT 1", (u,)).fetchone()
+            if r is not None and r[0] is not None:
+                alias = str(r[0] or "").strip()
+    except Exception:
+        alias = ""
+
+    cache[u] = alias
+    return alias
+
+
 def _scan_db_storage_mtime_ns(db_storage_dir: Path) -> int:
     try:
         base = str(db_storage_dir)
@@ -1386,6 +1413,16 @@ def _append_full_messages_from_rows(
     resource_conn: Optional[sqlite3.Connection],
     resource_chat_id: Optional[int],
 ) -> None:
+    contact_conn: Optional[sqlite3.Connection] = None
+    alias_cache: dict[str, str] = {}
+    if is_group:
+        try:
+            contact_db_path = account_dir / "contact.db"
+            if contact_db_path.exists():
+                contact_conn = sqlite3.connect(str(contact_db_path))
+        except Exception:
+            contact_conn = None
+
     for r in rows:
         local_id = int(r["local_id"] or 0)
         create_time = int(r["create_time"] or 0)
@@ -1448,10 +1485,21 @@ def _append_full_messages_from_rows(
         raw_text = raw_text.strip()
 
         sender_prefix = ""
-        if is_group and not raw_text.startswith("<") and not raw_text.startswith('"<'):
+        if is_group and raw_text and (not raw_text.startswith("<")) and (not raw_text.startswith('"<')):
-            sender_prefix, raw_text = _split_group_sender_prefix(raw_text)
+            sender_alias = ""
+            sep = raw_text.find(":\n")
+            if sep > 0:
+                prefix = raw_text[:sep].strip()
+                if prefix and sender_username and prefix != sender_username:
+                    strong_hint = prefix.startswith("wxid_") or prefix.endswith("@chatroom") or "@" in prefix
+                    if not strong_hint:
+                        body_probe = raw_text[sep + 2 :].lstrip("\n").lstrip()
+                        body_is_xml = body_probe.startswith("<") or body_probe.startswith('"<')
+                        if not body_is_xml:
+                            sender_alias = _lookup_contact_alias(contact_conn, alias_cache, sender_username)
+            sender_prefix, raw_text = _split_group_sender_prefix(raw_text, sender_username, sender_alias)
 
-        if is_group and sender_prefix:
+        if is_group and sender_prefix and (not sender_username):
             sender_username = sender_prefix
 
         if is_group and (raw_text.startswith("<") or raw_text.startswith('"<')):
@@ -1812,6 +1860,12 @@ def _append_full_messages_from_rows(
             }
         )
 
+    if contact_conn is not None:
+        try:
+            contact_conn.close()
+        except Exception:
+            pass
+
 
 def _postprocess_full_messages(
     *,
@@ -2301,6 +2355,16 @@ def _collect_chat_messages(
     pat_usernames: set[str] = set()
     has_more_any = False
 
+    contact_conn: Optional[sqlite3.Connection] = None
+    alias_cache: dict[str, str] = {}
+    if is_group:
+        try:
+            contact_db_path = account_dir / "contact.db"
+            if contact_db_path.exists():
+                contact_conn = sqlite3.connect(str(contact_db_path))
+        except Exception:
+            contact_conn = None
+
     for db_path in db_paths:
         conn = sqlite3.connect(str(db_path))
         conn.row_factory = sqlite3.Row
@@ -2384,10 +2448,21 @@ def _collect_chat_messages(
                 raw_text = raw_text.strip()
 
                 sender_prefix = ""
-                if is_group and not raw_text.startswith("<") and not raw_text.startswith('"<'):
+                if is_group and raw_text and (not raw_text.startswith("<")) and (not raw_text.startswith('"<')):
-                    sender_prefix, raw_text = _split_group_sender_prefix(raw_text)
+                    sender_alias = ""
+                    sep = raw_text.find(":\n")
+                    if sep > 0:
+                        prefix = raw_text[:sep].strip()
+                        if prefix and sender_username and prefix != sender_username:
+                            strong_hint = prefix.startswith("wxid_") or prefix.endswith("@chatroom") or "@" in prefix
+                            if not strong_hint:
+                                body_probe = raw_text[sep + 2 :].lstrip("\n").lstrip()
+                                body_is_xml = body_probe.startswith("<") or body_probe.startswith('"<')
+                                if not body_is_xml:
+                                    sender_alias = _lookup_contact_alias(contact_conn, alias_cache, sender_username)
+                    sender_prefix, raw_text = _split_group_sender_prefix(raw_text, sender_username, sender_alias)
 
-                if is_group and sender_prefix:
+                if is_group and sender_prefix and (not sender_username):
                     sender_username = sender_prefix
 
                 if is_group and (raw_text.startswith("<") or raw_text.startswith('"<')):
@@ -2744,6 +2819,12 @@ def _collect_chat_messages(
         finally:
             conn.close()
 
+    if contact_conn is not None:
+        try:
+            contact_conn.close()
+        except Exception:
+            pass
+
     return merged, has_more_any, sender_usernames, quote_usernames, pat_usernames