codebee 0.1.62 → 0.1.63

package/tools/wx_reader.py

@@ -1,261 +1,261 @@
-# -*- coding: utf-8 -*-
-"""微信读取 sidecar：由 64 位 Python ≥3.9（配 wechatauto-replica）运行，
-把本机微信 4.x 群消息增量拉成本仓库已有的「导出 txt」格式。
-
-为什么是独立进程：wechatauto 依赖链（winsdk 等）需要 64 位 Python ≥3.9，
-而 CodeBee 服务钉在 3.8-32。sidecar 通过文件与主程序解耦——拉到的新消息
-按「2024-01-01 12:34:56 昵称」行格式追加进监控文件夹，wxdigest 现有的
-解析/增量游标/摘要/桌宠提醒整条链路零改动复用。
-
-用法（两个模式，互斥）：
-  列群（供前端选择）：
-    python wx_reader.py --list-groups --result OUT.json
-  拉增量（每次扫描前调用）：
-    python wx_reader.py --pull --out DIR --state STATE.json \
-                        --groups-file GROUPS.json --result OUT.json
-    GROUPS.json = [{"username": "...@chatroom", "name": "群名"}, ...]
-
-只读保证：只调用 get_groups/get_new_messages/get_group_members/get_self_info，
-不调用任何发送/删除接口。
-结果 JSON 写 --result 文件（stdout 被库日志污染，不作为契约）。
-语法保持 3.8 兼容（测试解释器要 import 本文件的纯函数）。
-"""
-from __future__ import annotations
-
-import argparse
-import json
-import os
-import re
-import sys
-import time
-
-# 群消息 content 自带「<wxid>: 前缀」，剥掉后用成员表翻译成昵称
-_PREFIX_RE = re.compile(r"^([A-Za-z0-9_+@\-.]{3,}):\s*\n?")
-_ILLEGAL_FN = re.compile(r'[\\/:*?"<>|]')
-
-SELF_NAME = "我"
-
-
-def sanitize_filename(name):
-    """群名 → 合法文件名（Windows 非法字符换下划线；空名回落 username）。"""
-    out = _ILLEGAL_FN.sub("_", str(name or "").strip())
-    return out[:80] or "unnamed"
-
-
-def strip_sender_prefix(content, sender_username):
-    """群消息正文剥掉自带的「wxid: 」前缀（前缀与发送者一致才可信地剥）。"""
-    text = str(content or "")
-    m = _PREFIX_RE.match(text)
-    if m and (not sender_username or m.group(1) == sender_username):
-        return text[m.end():].strip()
-    return text.strip()
-
-
-def member_map(members):
-    """get_group_members 结果 → {username: 显示名}。备注优先（微信 UI 同款）。"""
-    out = {}
-    for m in members or []:
-        if not isinstance(m, dict):
-            continue
-        uid = m.get("username") or ""
-        if not uid:
-            continue
-        out[uid] = m.get("remark") or m.get("nick_name") or uid
-    return out
-
-
-def format_line(ts_epoch, sender_name, text):
-    """一条消息 → 导出格式行（时间戳行与正文之间不留空行，与现有解析器对齐）。"""
-    stamp = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(ts_epoch or 0))
-    return "%s %s\n%s\n" % (stamp, sender_name or "?", str(text or "").strip())
-
-
-def type_placeholder(mtype):
-    """非文本消息 → 占位符（摘要时让模型知道有图/语音存在）。None=跳过不入正文。"""
-    t = str(mtype or "")
-    if "文本" in t:
-        return None            # 文本走正文，不是占位符
-    for key, ph in (("图片", "[图片]"), ("语音", "[语音]"), ("视频", "[视频]"),
-                    ("文件", "[文件]"), ("链接", "[链接]"), ("名片", "[名片]"),
-                    ("位置", "[位置]"), ("动画表情", "[表情]"), ("系统", None)):
-        if key in t:
-            return ph
-    return "[%s]" % t if t and len(t) <= 8 else None
-
-
-def _quiet_lib_logs():
-    try:
-        import logging
-        logging.getLogger("wechatauto").setLevel(logging.WARNING)
-    except Exception:
-        pass
-
-
-def _write_result(path, obj):
-    tmp = str(path) + ".tmp"
-    with open(tmp, "w", encoding="utf-8") as f:
-        json.dump(obj, f, ensure_ascii=False)
-    os.replace(tmp, path)
-
-
-def _load_json(path, default):
-    try:
-        with open(path, "r", encoding="utf-8") as f:
-            return json.load(f)
-    except Exception:
-        return default
-
-
-def load_state(path):
-    return _load_json(path, {}) or {}
-
-
-def save_state(path, state):
-    tmp = str(path) + ".tmp"
-    with open(tmp, "w", encoding="utf-8") as f:
-        json.dump(state, f, ensure_ascii=False)
-    os.replace(tmp, path)
-
-
-def cmd_list_groups(result_path):
-    _quiet_lib_logs()
-    import wechatauto
-    db = wechatauto.WeChatDB()
-    groups, seen = [], set()
-    for g in db.get_groups():
-        if isinstance(g, dict) and g.get("username"):
-            groups.append({"username": g["username"],
-                           "name": g.get("name") or g["username"],
-                           "member_count": str(g.get("member_count") or "")})
-            seen.add(g["username"])
-    # 新群兜底（2026-09-21 用户实测「缺一个新建的群」）：刚建的群可能还没
-    # 落进 chat_room/contact，但只要有过消息就在会话列表里——合并补进，
-    # 昵称未同步时先显示 wxid，微信同步后「刷新群列表」即见真名。
-    try:
-        for s in db.get_sessions(200):
-            u = str((s or {}).get("username") or "")
-            if u.endswith("@chatroom") and u not in seen:
-                groups.append({"username": u,
-                               "name": db.get_nickname(u) or u,
-                               "member_count": ""})
-                seen.add(u)
-    except Exception:
-        pass   # 会话库不可用不影响主清单
-    _write_result(result_path, {"ok": True, "groups": groups, "count": len(groups)})
-
-
-def cmd_pull(out_dir, state_path, groups_file, result_path):
-    _quiet_lib_logs()
-    import wechatauto
-    sel = _load_json(groups_file, [])
-    sel = [g for g in (sel or []) if isinstance(g, dict) and g.get("username")]
-    state = load_state(state_path)
-    os.makedirs(out_dir, exist_ok=True)
-    db = wechatauto.WeChatDB()
-    try:
-        self_wxid = db.wxid
-    except Exception:
-        self_wxid = ""
-    pulled, errors = {}, []
-    for g in sel:
-        username, name = g["username"], g.get("name") or g["username"]
-        since = 0
-        try:
-            since = int((state.get(username) or {}).get("since_seq") or 0)
-        except (TypeError, ValueError):
-            since = 0
-        try:
-            msgs = db.get_new_messages(username, since_seq=since, limit=500)
-        except Exception as e:
-            errors.append("%s: %s" % (name, e))
-            continue
-        if not msgs:
-            pulled[username] = 0
-            continue
-        members = {}
-        try:
-            members = member_map(db.get_group_members(username))
-        except Exception:
-            members = {}
-        # 昵称缓存：退群成员不在成员表里，靠 get_nickname 兜底一次后永久缓存
-        gstate = state.get(username) or {}
-        names = dict(gstate.get("names") or {})
-
-        def resolve(uid):
-            if not uid:
-                return "?"
-            if self_wxid and uid == self_wxid:
-                return SELF_NAME
-            if uid in members:
-                return members[uid]
-            if uid in names:
-                return names[uid]
-            try:
-                nick = str(db.get_nickname(uid) or "").strip()
-            except Exception:
-                nick = ""
-            name2 = nick or uid
-            if len(names) < 5000:          # 缓存上限防无限膨胀
-                names[uid] = name2
-            return name2
-
-        fname = os.path.join(out_dir, sanitize_filename(name) + ".txt")
-        header = "【%s】\n" % name
-        need_header = not os.path.exists(fname)
-        lines = []
-        max_seq = since
-        for m in msgs:
-            seq = m.get("sort_seq") or 0
-            if seq <= since:
-                continue
-            max_seq = max(max_seq, seq)
-            uid = m.get("sender_username") or ""
-            who = resolve(uid)
-            mtype = m.get("type")
-            text = None
-            if "文本" in str(mtype or ""):
-                text = strip_sender_prefix(m.get("content"), uid)
-            else:
-                text = type_placeholder(mtype)
-            if not text:
-                continue
-            lines.append(format_line(m.get("create_time"), who, text))
-        if lines:
-            with open(fname, "a", encoding="utf-8") as f:
-                if need_header:
-                    f.write(header)
-                f.write("\n".join(lines) + "\n")
-        state[username] = {"name": name, "since_seq": max_seq,
-                           "names": names, "t": time.time()}
-        pulled[username] = len(lines)
-    save_state(state_path, state)
-    _write_result(result_path, {"ok": not errors, "pulled": pulled,
-                                "errors": errors, "count": sum(pulled.values())})
-
-
-def main():
-    ap = argparse.ArgumentParser(description="CodeBee 微信群消息只读 sidecar")
-    ap.add_argument("--list-groups", action="store_true")
-    ap.add_argument("--pull", action="store_true")
-    ap.add_argument("--out", default="")
-    ap.add_argument("--state", default="")
-    ap.add_argument("--groups-file", default="")
-    ap.add_argument("--result", required=True, help="结果 JSON 落点（stdout 不可靠）")
-    args = ap.parse_args()
-    try:
-        if args.list_groups:
-            cmd_list_groups(args.result)
-        elif args.pull:
-            cmd_pull(args.out, args.state, args.groups_file, args.result)
-        else:
-            ap.error("需要 --list-groups 或 --pull")
-    except Exception as e:
-        # 顶层异常也落 result：微信未登录/DB 锁住等真实原因必须透传给主程序
-        _write_result(args.result, {"ok": False, "errors": [str(e)],
-                                    "groups": [], "pulled": {}, "count": 0})
-        return 1
-    return 0
-
-
-if __name__ == "__main__":
-    sys.exit(main())
+# -*- coding: utf-8 -*-
+"""微信读取 sidecar：由 64 位 Python ≥3.9（配 wechatauto-replica）运行，
+把本机微信 4.x 群消息增量拉成本仓库已有的「导出 txt」格式。
+
+为什么是独立进程：wechatauto 依赖链（winsdk 等）需要 64 位 Python ≥3.9，
+而 CodeBee 服务钉在 3.8-32。sidecar 通过文件与主程序解耦——拉到的新消息
+按「2024-01-01 12:34:56 昵称」行格式追加进监控文件夹，wxdigest 现有的
+解析/增量游标/摘要/桌宠提醒整条链路零改动复用。
+
+用法（两个模式，互斥）：
+  列群（供前端选择）：
+    python wx_reader.py --list-groups --result OUT.json
+  拉增量（每次扫描前调用）：
+    python wx_reader.py --pull --out DIR --state STATE.json \
+                        --groups-file GROUPS.json --result OUT.json
+    GROUPS.json = [{"username": "...@chatroom", "name": "群名"}, ...]
+
+只读保证：只调用 get_groups/get_new_messages/get_group_members/get_self_info，
+不调用任何发送/删除接口。
+结果 JSON 写 --result 文件（stdout 被库日志污染，不作为契约）。
+语法保持 3.8 兼容（测试解释器要 import 本文件的纯函数）。
+"""
+from __future__ import annotations
+
+import argparse
+import json
+import os
+import re
+import sys
+import time
+
+# 群消息 content 自带「<wxid>: 前缀」，剥掉后用成员表翻译成昵称
+_PREFIX_RE = re.compile(r"^([A-Za-z0-9_+@\-.]{3,}):\s*\n?")
+_ILLEGAL_FN = re.compile(r'[\\/:*?"<>|]')
+
+SELF_NAME = "我"
+
+
+def sanitize_filename(name):
+    """群名 → 合法文件名（Windows 非法字符换下划线；空名回落 username）。"""
+    out = _ILLEGAL_FN.sub("_", str(name or "").strip())
+    return out[:80] or "unnamed"
+
+
+def strip_sender_prefix(content, sender_username):
+    """群消息正文剥掉自带的「wxid: 」前缀（前缀与发送者一致才可信地剥）。"""
+    text = str(content or "")
+    m = _PREFIX_RE.match(text)
+    if m and (not sender_username or m.group(1) == sender_username):
+        return text[m.end():].strip()
+    return text.strip()
+
+
+def member_map(members):
+    """get_group_members 结果 → {username: 显示名}。备注优先（微信 UI 同款）。"""
+    out = {}
+    for m in members or []:
+        if not isinstance(m, dict):
+            continue
+        uid = m.get("username") or ""
+        if not uid:
+            continue
+        out[uid] = m.get("remark") or m.get("nick_name") or uid
+    return out
+
+
+def format_line(ts_epoch, sender_name, text):
+    """一条消息 → 导出格式行（时间戳行与正文之间不留空行，与现有解析器对齐）。"""
+    stamp = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(ts_epoch or 0))
+    return "%s %s\n%s\n" % (stamp, sender_name or "?", str(text or "").strip())
+
+
+def type_placeholder(mtype):
+    """非文本消息 → 占位符（摘要时让模型知道有图/语音存在）。None=跳过不入正文。"""
+    t = str(mtype or "")
+    if "文本" in t:
+        return None            # 文本走正文，不是占位符
+    for key, ph in (("图片", "[图片]"), ("语音", "[语音]"), ("视频", "[视频]"),
+                    ("文件", "[文件]"), ("链接", "[链接]"), ("名片", "[名片]"),
+                    ("位置", "[位置]"), ("动画表情", "[表情]"), ("系统", None)):
+        if key in t:
+            return ph
+    return "[%s]" % t if t and len(t) <= 8 else None
+
+
+def _quiet_lib_logs():
+    try:
+        import logging
+        logging.getLogger("wechatauto").setLevel(logging.WARNING)
+    except Exception:
+        pass
+
+
+def _write_result(path, obj):
+    tmp = str(path) + ".tmp"
+    with open(tmp, "w", encoding="utf-8") as f:
+        json.dump(obj, f, ensure_ascii=False)
+    os.replace(tmp, path)
+
+
+def _load_json(path, default):
+    try:
+        with open(path, "r", encoding="utf-8") as f:
+            return json.load(f)
+    except Exception:
+        return default
+
+
+def load_state(path):
+    return _load_json(path, {}) or {}
+
+
+def save_state(path, state):
+    tmp = str(path) + ".tmp"
+    with open(tmp, "w", encoding="utf-8") as f:
+        json.dump(state, f, ensure_ascii=False)
+    os.replace(tmp, path)
+
+
+def cmd_list_groups(result_path):
+    _quiet_lib_logs()
+    import wechatauto
+    db = wechatauto.WeChatDB()
+    groups, seen = [], set()
+    for g in db.get_groups():
+        if isinstance(g, dict) and g.get("username"):
+            groups.append({"username": g["username"],
+                           "name": g.get("name") or g["username"],
+                           "member_count": str(g.get("member_count") or "")})
+            seen.add(g["username"])
+    # 新群兜底（2026-09-21 用户实测「缺一个新建的群」）：刚建的群可能还没
+    # 落进 chat_room/contact，但只要有过消息就在会话列表里——合并补进，
+    # 昵称未同步时先显示 wxid，微信同步后「刷新群列表」即见真名。
+    try:
+        for s in db.get_sessions(200):
+            u = str((s or {}).get("username") or "")
+            if u.endswith("@chatroom") and u not in seen:
+                groups.append({"username": u,
+                               "name": db.get_nickname(u) or u,
+                               "member_count": ""})
+                seen.add(u)
+    except Exception:
+        pass   # 会话库不可用不影响主清单
+    _write_result(result_path, {"ok": True, "groups": groups, "count": len(groups)})
+
+
+def cmd_pull(out_dir, state_path, groups_file, result_path):
+    _quiet_lib_logs()
+    import wechatauto
+    sel = _load_json(groups_file, [])
+    sel = [g for g in (sel or []) if isinstance(g, dict) and g.get("username")]
+    state = load_state(state_path)
+    os.makedirs(out_dir, exist_ok=True)
+    db = wechatauto.WeChatDB()
+    try:
+        self_wxid = db.wxid
+    except Exception:
+        self_wxid = ""
+    pulled, errors = {}, []
+    for g in sel:
+        username, name = g["username"], g.get("name") or g["username"]
+        since = 0
+        try:
+            since = int((state.get(username) or {}).get("since_seq") or 0)
+        except (TypeError, ValueError):
+            since = 0
+        try:
+            msgs = db.get_new_messages(username, since_seq=since, limit=500)
+        except Exception as e:
+            errors.append("%s: %s" % (name, e))
+            continue
+        if not msgs:
+            pulled[username] = 0
+            continue
+        members = {}
+        try:
+            members = member_map(db.get_group_members(username))
+        except Exception:
+            members = {}
+        # 昵称缓存：退群成员不在成员表里，靠 get_nickname 兜底一次后永久缓存
+        gstate = state.get(username) or {}
+        names = dict(gstate.get("names") or {})
+
+        def resolve(uid):
+            if not uid:
+                return "?"
+            if self_wxid and uid == self_wxid:
+                return SELF_NAME
+            if uid in members:
+                return members[uid]
+            if uid in names:
+                return names[uid]
+            try:
+                nick = str(db.get_nickname(uid) or "").strip()
+            except Exception:
+                nick = ""
+            name2 = nick or uid
+            if len(names) < 5000:          # 缓存上限防无限膨胀
+                names[uid] = name2
+            return name2
+
+        fname = os.path.join(out_dir, sanitize_filename(name) + ".txt")
+        header = "【%s】\n" % name
+        need_header = not os.path.exists(fname)
+        lines = []
+        max_seq = since
+        for m in msgs:
+            seq = m.get("sort_seq") or 0
+            if seq <= since:
+                continue
+            max_seq = max(max_seq, seq)
+            uid = m.get("sender_username") or ""
+            who = resolve(uid)
+            mtype = m.get("type")
+            text = None
+            if "文本" in str(mtype or ""):
+                text = strip_sender_prefix(m.get("content"), uid)
+            else:
+                text = type_placeholder(mtype)
+            if not text:
+                continue
+            lines.append(format_line(m.get("create_time"), who, text))
+        if lines:
+            with open(fname, "a", encoding="utf-8") as f:
+                if need_header:
+                    f.write(header)
+                f.write("\n".join(lines) + "\n")
+        state[username] = {"name": name, "since_seq": max_seq,
+                           "names": names, "t": time.time()}
+        pulled[username] = len(lines)
+    save_state(state_path, state)
+    _write_result(result_path, {"ok": not errors, "pulled": pulled,
+                                "errors": errors, "count": sum(pulled.values())})
+
+
+def main():
+    ap = argparse.ArgumentParser(description="CodeBee 微信群消息只读 sidecar")
+    ap.add_argument("--list-groups", action="store_true")
+    ap.add_argument("--pull", action="store_true")
+    ap.add_argument("--out", default="")
+    ap.add_argument("--state", default="")
+    ap.add_argument("--groups-file", default="")
+    ap.add_argument("--result", required=True, help="结果 JSON 落点（stdout 不可靠）")
+    args = ap.parse_args()
+    try:
+        if args.list_groups:
+            cmd_list_groups(args.result)
+        elif args.pull:
+            cmd_pull(args.out, args.state, args.groups_file, args.result)
+        else:
+            ap.error("需要 --list-groups 或 --pull")
+    except Exception as e:
+        # 顶层异常也落 result：微信未登录/DB 锁住等真实原因必须透传给主程序
+        _write_result(args.result, {"ok": False, "errors": [str(e)],
+                                    "groups": [], "pulled": {}, "count": 0})
+        return 1
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())