server.py

"""
LearnMap Server — API + 静态文件服务

启动:
    python server.py

访问:
    http://localhost:8080

API:
    POST /api/extract  导入 Markdown/文本，返回知识图谱 JSON
"""
import json, os, sys
from pathlib import Path
import http.server
import urllib.parse

# ── 路径 ──
PROJECT_ROOT = Path(__file__).resolve().parent
sys.path.insert(0, str(PROJECT_ROOT))

from scripts.extract_graph import extract


class Handler(http.server.SimpleHTTPRequestHandler):

    def __init__(self, *args, **kwargs):
        super().__init__(*args, directory=str(PROJECT_ROOT), **kwargs)

    # ── GET API ───────────────────────────────────────
    def do_GET(self):
        parsed = urllib.parse.urlparse(self.path)

        if parsed.path == "/api/results":
            result_dir = PROJECT_ROOT / "doc" / "result"
            files = []
            if result_dir.exists():
                for f in sorted(result_dir.glob("*.json")):
                    if f.name != "_raw_response.txt":
                        files.append({
                            "name": f.name,
                            "path": f"doc/result/{f.name}",
                            "size": f.stat().st_size,
                            "mtime": f.stat().st_mtime,
                        })
            self._json(200, {"files": files})
            return

        if parsed.path == "/api/knowledge-base":
            kb_path = PROJECT_ROOT / "database" / "knowledge_base.json"
            if kb_path.exists():
                data = json.loads(kb_path.read_text(encoding="utf-8"))
                self._json(200, data)
            else:
                self._json(200, {"nodes": [], "relations": []})
            return

        if parsed.path == "/api/kb-stats":
            kb_path = PROJECT_ROOT / "database" / "knowledge_base.json"
            nodes = 0
            if kb_path.exists():
                data = json.loads(kb_path.read_text(encoding="utf-8"))
                nodes = len(data.get("nodes", []))
            self._json(200, {"node_count": nodes})
            return

        # 其他 GET 请求 → 静态文件
        return super().do_GET()

    # ── POST API ──────────────────────────────────────
    def do_POST(self):
        parsed = urllib.parse.urlparse(self.path)

        # ── 合并到知识库 ──
        if parsed.path == "/api/merge":
            body = self._read_body()
            new_nodes = body.get("nodes", [])
            new_rels = body.get("relations", [])
            kb_path = PROJECT_ROOT / "database" / "knowledge_base.json"
            kb = json.loads(kb_path.read_text(encoding="utf-8")) if kb_path.exists() else {"nodes": [], "relations": []}

            existing_ids = {n.get("id") or n.get("name") for n in kb["nodes"]}
            added = 0
            for n in new_nodes:
                nid = n.get("id") or n.get("name")
                if nid and nid not in existing_ids:
                    kb["nodes"].append(n); existing_ids.add(nid); added += 1

            existing_rels = {(r.get("source"), r.get("target"), r.get("relation")) for r in kb["relations"]}
            rel_added = 0
            for r in new_rels:
                key = (r.get("source"), r.get("target"), r.get("relation"))
                if key not in existing_rels:
                    kb["relations"].append(r); existing_rels.add(key); rel_added += 1

            kb["version"] = kb.get("version", 1) + 1
            kb["updated_at"] = "2026-07-25"
            kb_path.write_text(json.dumps(kb, ensure_ascii=False, indent=2), encoding="utf-8")

            self._json(200, {"node_count": len(kb["nodes"]), "added_nodes": added, "added_rels": rel_added})
            return

        # ── 提交知识挑战答案 ──
        if parsed.path == "/api/challenge/submit":
            body = self._read_body()
            kb_path = PROJECT_ROOT / "database" / "knowledge_base.json"
            kb = json.loads(kb_path.read_text(encoding="utf-8")) if kb_path.exists() else {"nodes": [], "relations": []}
            if "human_contributions" not in kb:
                kb["human_contributions"] = {"descriptions": [], "relations": [], "expansions": []}
            hc = kb["human_contributions"]

            ctype = body.get("type", "description")
            entry = body.get("entry", {})
            entry["votes"] = 1
            entry["created_at"] = "2026-07-25"

            if ctype == "description":
                hc["descriptions"].append(entry)
            elif ctype == "relation":
                hc["relations"].append(entry)
            elif ctype == "expansion":
                hc["expansions"].append(entry)

            kb_path.write_text(json.dumps(kb, ensure_ascii=False, indent=2), encoding="utf-8")
            self._json(200, {"ok": True, "total": len(hc[ctype + "s" if ctype != "expansion" else "expansions"])})
            return

        if parsed.path == "/api/extract":
            body = self._read_body()
            markdown = body.get("markdown", "")
            filename = body.get("filename", "import.md")

            input_dir = PROJECT_ROOT / "doc" / "input"
            input_dir.mkdir(parents=True, exist_ok=True)
            md_path = input_dir / filename
            md_path.write_text(markdown, encoding="utf-8")

            try:
                result = extract(md_path)
                self._json(200, result)
            except Exception as e:
                self._json(500, {"error": str(e)})
            return

        if parsed.path == "/api/extract-stream":
            self.do_POST_stream()
            return

        self._json(404, {"error": "not found"})

    # ── 流式 API（带进度推送） ──────────────────────────
    def _send_event(self, event_type: str, data: dict):
        """发送一条 newline-delimited JSON 事件"""
        msg = json.dumps({"type": event_type, **data}, ensure_ascii=False) + "\n"
        self.wfile.write(msg.encode())
        self.wfile.flush()

    def do_POST_stream(self):
        """处理 /api/extract-stream — 流式进度推送"""
        parsed = urllib.parse.urlparse(self.path)
        if parsed.path != "/api/extract-stream":
            self._json(404, {"error": "not found"})
            return

        body = self._read_body()
        markdown = body.get("markdown", "")
        filename = body.get("filename", "import.md")

        # 流式响应头
        self.send_response(200)
        self.send_header("Content-Type", "application/x-ndjson; charset=utf-8")
        self.send_header("Access-Control-Allow-Origin", "*")
        self.send_header("Cache-Control", "no-cache")
        self.end_headers()

        try:
            # 步骤 1：保存文档
            self._send_event("progress", {"step": 1, "message": "📄 保存文档...", "total": 5})

            input_dir = PROJECT_ROOT / "doc" / "input"
            input_dir.mkdir(parents=True, exist_ok=True)
            md_path = input_dir / filename
            md_path.write_text(markdown, encoding="utf-8")

            # 步骤 2：读取 Prompt
            self._send_event("progress", {"step": 2, "message": "📋 读取 Prompt 模板...", "total": 5})

            # 步骤 3：调用 DeepSeek
            self._send_event("progress", {"step": 3, "message": "🤖 调用 DeepSeek API（约 10~30 秒）...", "total": 5})

            result = extract(md_path)

            # 步骤 4：解析完成
            self._send_event("progress", {"step": 4, "message": f"✅ 解析完成 — {len(result.get('nodes',[]))} 个节点, {len(result.get('relations',[]))} 条关系", "total": 5})

            # 步骤 5：返回结果
            self._send_event("result", {"data": result})

        except Exception as e:
            self._send_event("error", {"message": str(e)})

    # ── 工具方法 ──────────────────────────────────────
    def _read_body(self) -> dict:
        length = int(self.headers.get("Content-Length", 0))
        if length == 0:
            return {}
        return json.loads(self.rfile.read(length))

    def _json(self, status: int, data):
        self.send_response(status)
        self.send_header("Content-Type", "application/json; charset=utf-8")
        self.send_header("Access-Control-Allow-Origin", "*")
        self.end_headers()
        self.wfile.write(json.dumps(data, ensure_ascii=False).encode())

    def log_message(self, fmt, *args):
        # 只记录 API 调用，跳过静态文件 404（如 sw.js）
        if args and isinstance(args[0], str) and (args[0].startswith("POST") or args[0].startswith("GET /api/")):
            print(f"[API] {fmt % args}")


if __name__ == "__main__":
    port = int(sys.argv[1]) if len(sys.argv) > 1 else 8080
    server = http.server.HTTPServer(("0.0.0.0", port), Handler)
    print(f"🚀  LearnMap Server")
    print(f"    打开: http://localhost:{port}")
    print(f"    API:  POST /api/extract")
    server.serve_forever()