mcai-knowledge 0.2.0__py3-none-any.whl

mcai/__init__.py

@@ -0,0 +1,3 @@
+"""元意识知识资产引擎（MCAI Engine）。"""
+
+__version__ = "0.2.0"

mcai/__main__.py

@@ -0,0 +1,7 @@
+"""python -m mcai 入口。"""
+import sys
+
+from .cli import main
+
+if __name__ == "__main__":
+    sys.exit(main())

mcai/autopilot.py

@@ -0,0 +1,229 @@
+"""自动驾驶：让知识库在无人参与下保持「活着、准着、可被调用」。
+
+一轮循环做四件事：
+1. 检查语料/配置指纹，变了就重建；
+2. 记录变更日志（新增/删除/改写的术语），让每次变化可追溯；
+3. 健康自检（知识库、站点、服务端口、接口响应）；
+4. 把状态写到 dist/status.json 与 data/status.json，供调用方判断是否健康。
+
+设计为「一次性执行」：由系统计划任务每 5 分钟调用一次，比常驻进程更抗崩溃。
+"""
+from __future__ import annotations
+
+import hashlib
+import json
+import socket
+import subprocess
+import sys
+import time
+import urllib.request
+from datetime import datetime
+from pathlib import Path
+
+from . import dashboard, knowledge, pipeline, quality, selftest, snapshot
+
+DEFAULT_PORT = 8787
+
+
+class Autopilot:
+    def __init__(self, root: str | Path, port: int = DEFAULT_PORT, ensure_server: bool = True):
+        self.root = Path(root).resolve()
+        self.port = port
+        self.ensure_server = ensure_server
+        self.data = self.root / "data"
+        self.data.mkdir(parents=True, exist_ok=True)
+
+    # ---------- 指纹 ----------
+    def fingerprint(self) -> str:
+        h = hashlib.sha256()
+        cfg_path = self.root / "config" / "corpus.json"
+        for path in sorted((self.root / "config").glob("*.json")):
+            h.update(path.name.encode("utf-8"))
+            h.update(path.read_bytes())
+        try:
+            cfg = json.loads(cfg_path.read_text(encoding="utf-8-sig"))
+        except Exception:  # noqa: BLE001
+            cfg = {}
+        for spec in cfg.get("corpus", []):
+            raw = Path(spec.get("path", ""))
+            path = raw if raw.is_absolute() else self.root / raw
+            if path.exists():
+                stat = path.stat()
+                h.update(f"{spec.get('id')}:{stat.st_size}:{int(stat.st_mtime)}".encode("utf-8"))
+            else:
+                h.update(f"{spec.get('id')}:missing".encode("utf-8"))
+        return h.hexdigest()
+
+    # ---------- 快照与差异 ----------
+    def _snapshot(self) -> dict:
+        path = self.data / "kb.json"
+        if not path.exists():
+            return {}
+        try:
+            kb = json.loads(path.read_text(encoding="utf-8"))
+        except json.JSONDecodeError:
+            return {}
+        return {
+            c["term"]: hashlib.md5(c.get("definition", "").encode("utf-8")).hexdigest()
+            for c in kb.get("concepts", [])
+        }
+
+    @staticmethod
+    def _diff(before: dict, after: dict) -> dict:
+        added = sorted(set(after) - set(before))
+        removed = sorted(set(before) - set(after))
+        changed = sorted(t for t in set(before) & set(after) if before[t] != after[t])
+        return {"added": added, "removed": removed, "changed": changed}
+
+    # ---------- 服务检查 ----------
+    def _port_open(self) -> bool:
+        try:
+            with socket.create_connection(("127.0.0.1", self.port), timeout=0.6):
+                return True
+        except OSError:
+            return False
+
+    def _server_healthy(self) -> bool:
+        try:
+            with urllib.request.urlopen(f"http://127.0.0.1:{self.port}/api/v1/health", timeout=2) as resp:
+                return resp.status == 200
+        except Exception:  # noqa: BLE001
+            return False
+
+    def _start_server(self) -> bool:
+        if not self.ensure_server or self._port_open():
+            return False
+        flags = 0
+        if sys.platform == "win32":
+            flags = 0x00000008 | 0x00000200  # DETACHED_PROCESS | CREATE_NEW_PROCESS_GROUP
+        try:
+            subprocess.Popen(
+                [sys.executable, "-m", "mcai", "serve", "--home", str(self.root), "--port", str(self.port)],
+                cwd=str(self.root),
+                stdin=subprocess.DEVNULL,
+                stdout=subprocess.DEVNULL,
+                stderr=subprocess.DEVNULL,
+                creationflags=flags,
+            )
+            for _ in range(10):
+                time.sleep(0.3)
+                if self._port_open():
+                    return True
+        except Exception:  # noqa: BLE001
+            return False
+        return False
+
+    # ---------- 主循环 ----------
+    def run_once(self) -> dict:
+        started = time.time()
+        state_path = self.data / "autopilot-state.json"
+        state = {}
+        if state_path.exists():
+            try:
+                state = json.loads(state_path.read_text(encoding="utf-8"))
+            except json.JSONDecodeError:
+                state = {}
+
+        before = self._snapshot()
+        fingerprint = self.fingerprint()
+        rebuilt = False
+        error = ""
+        result = {}
+
+        needs_build = (fingerprint != state.get("fingerprint")) or not (self.data / "kb.json").exists()
+        if needs_build:
+            try:
+                result = pipeline.rebuild(self.root)
+                rebuilt = True
+            except Exception as exc:  # noqa: BLE001
+                error = f"{type(exc).__name__}: {exc}"
+
+        after = self._snapshot()
+        diff = self._diff(before, after) if rebuilt else {"added": [], "removed": [], "changed": []}
+
+        started_server = self._start_server()
+        server_ok = self._server_healthy()
+
+        kb = {}
+        try:
+            kb = knowledge.load(self.root).meta.get("stats", {})
+        except Exception:  # noqa: BLE001
+            pass
+
+        try:
+            q = quality.inspect(self.root)
+        except Exception as exc:  # noqa: BLE001
+            q = {"ok": False, "counts": {"error": 1, "warn": 0},
+                 "errors": [{"detail": f"{type(exc).__name__}: {exc}"}]}
+        try:
+            t = selftest.run(self.root)
+        except Exception as exc:  # noqa: BLE001
+            t = {"ok": False, "passed": 0, "total": 0, "failed": [f"{type(exc).__name__}: {exc}"]}
+
+        now = datetime.now().astimezone().isoformat(timespec="seconds")
+        status = {
+            "ok": (not error and bool(kb) and (server_ok if self.ensure_server else True)
+                   and q.get("ok", False) and t.get("ok", False)),
+            "quality": {"ok": q.get("ok"), "errors": q.get("counts", {}).get("error", 0),
+                        "warnings": q.get("counts", {}).get("warn", 0)},
+            "selftest": {"ok": t.get("ok"), "passed": t.get("passed"), "total": t.get("total"),
+                         "failed": t.get("failed", [])},
+            "checked_at": now,
+            "built_at": result.get("built_at") or kb.get("built_at"),
+            "rebuilt": rebuilt,
+            "started_server": started_server,
+            "server_ok": server_ok,
+            "concepts": kb.get("concepts"),
+            "faqs": kb.get("faqs"),
+            "relations": kb.get("relations"),
+            "evidence": kb.get("evidence"),
+            "error": error,
+            "duration_ms": int((time.time() - started) * 1000),
+        }
+
+        # 写状态：data 里永久保留，dist 里给外部读（重建会清空 dist，所以最后写）
+        (self.data / "status.json").write_text(json.dumps(status, ensure_ascii=False, indent=2), encoding="utf-8")
+        dist = self.root / "dist"
+        if dist.exists():
+            (dist / "status.json").write_text(json.dumps(status, ensure_ascii=False, indent=2), encoding="utf-8")
+
+        # 变更日志
+        if rebuilt and before and (diff["added"] or diff["removed"] or diff["changed"]):
+            entry = {"ts": now, **diff, "concepts": status["concepts"], "faqs": status["faqs"]}
+            with (self.data / "changelog.jsonl").open("a", encoding="utf-8") as fh:
+                fh.write(json.dumps(entry, ensure_ascii=False) + "\n")
+
+        with (self.data / "autopilot.jsonl").open("a", encoding="utf-8") as fh:
+            fh.write(json.dumps(status, ensure_ascii=False) + "\n")
+
+        try:
+            dashboard.render(self.root)
+        except Exception:  # noqa: BLE001 - 看板失败不影响巡检
+            pass
+
+        if rebuilt:
+            try:
+                snap = snapshot.create(self.root, note="自动快照：语料变更", keep=8)
+                status["snapshot"] = Path(snap["archive"]).name
+            except Exception:  # noqa: BLE001 - 快照失败不影响巡检
+                pass
+            state["fingerprint"] = fingerprint
+            state["last_build"] = now
+            state_path.write_text(json.dumps(state, ensure_ascii=False, indent=2), encoding="utf-8")
+
+        return {"status": status, "diff": diff, "rebuilt": rebuilt}
+
+    def loop(self, interval: int = 60) -> None:
+        print(f"自动驾驶已启动（每 {interval} 秒检查一次，Ctrl+C 停止）")
+        while True:
+            try:
+                outcome = self.run_once()
+                s = outcome["status"]
+                flag = "重建" if outcome["rebuilt"] else "无变化"
+                print(f"[{s['checked_at']}] {flag} · {s['concepts']} 概念 · 服务 "
+                      f"{'在线' if s['server_ok'] else '离线'} · {s['duration_ms']}ms")
+            except KeyboardInterrupt:
+                raise
+            except Exception as exc:  # noqa: BLE001
+                print(f"[{datetime.now().isoformat(timespec='seconds')}] 异常：{type(exc).__name__}: {exc}")
+            time.sleep(max(5, interval))