refactor(database): 将数据库初始化逻辑统一整合到 database 包

.github/workflows/ci.yml

@@ -21,10 +21,7 @@ jobs:
       - run: python -c "from app.main import app; print('import ok')"
       - name: Init test DB
         run: |
-          cd ..
-          python scripts/init_db.py
-          python scripts/seed_data.py
-          python scripts/compute_baseline.py
+          python -c "from app.database import init_database; from app.main import DB_PATH; init_database(DB_PATH)"
       - run: python -m pytest tests/ -v --tb=short
 
   frontend:

Makefile

@@ -2,9 +2,11 @@
 
 # 一键初始化数据库并生成 baseline
 data:
-	python3 scripts/init_db.py
-	python3 scripts/seed_data.py
-	python3 scripts/compute_baseline.py
+	cd backend && python3 -c "from app.database import init_database; from pathlib import Path; init_database(Path('data') / 'baseline.db')"
+
+# 清空种子和基线后重新初始化
+data-refresh:
+	cd backend && python3 -c "from app.database import init_database; from pathlib import Path; init_database(Path('data') / 'baseline.db', force=True)"
 
 # 安装所有依赖
 install:

backend/app/database/__init__.py

@@ -0,0 +1,19 @@
+import sqlite3
+from pathlib import Path
+
+from app.database.schema import create_tables
+
+
+def init_database(db_path: Path, force: bool = False):
+    """创建数据库表结构（如不存在）；force=True 时清空种子和基线后重写"""
+    db_path.parent.mkdir(parents=True, exist_ok=True)
+    conn = sqlite3.connect(db_path)
+    create_tables(conn)
+    conn.commit()
+    conn.close()
+
+    from app.database.seed_data import seed_data
+    from app.database.baseline import compute_baseline
+
+    seed_data(db_path, force=force)
+    compute_baseline(db_path, force=force)

scripts/compute_baseline.py → backend/app/database/baseline.py

@@ -1,15 +1,7 @@
-"""
-基于 notes 表数据，预计算各垂类的 baseline 统计指标并写入 baseline_stats 表。
-
-Usage:
-    python scripts/compute_baseline.py
-"""
 import sqlite3
 import json
-import os
 from collections import Counter
-
-DB_PATH = os.path.join(os.path.dirname(__file__), "..", "backend", "data", "baseline.db")
+from pathlib import Path
 
 
 def upsert_stat(cursor, category, metric_name, metric_value=None, metric_json=None):
@@ -174,20 +166,22 @@ def compute_for_category(cursor, category):
     print(f"  [{category}] 已计算 baseline 指标（含粉丝分层与标签分桶）")
 
 
-def main():
-    """计算所有垂类的 baseline 统计指标"""
-    conn = sqlite3.connect(DB_PATH)
+def compute_baseline(db_path: Path, force: bool = False):
+    """计算所有垂类的 baseline 统计指标；force=True 时清空后重算"""
+    conn = sqlite3.connect(db_path)
     cursor = conn.cursor()
 
-    cursor.execute("DELETE FROM baseline_stats")
+    if force:
+        cursor.execute("DELETE FROM baseline_stats")
+    else:
+        cursor.execute("SELECT COUNT(*) FROM baseline_stats")
+        if cursor.fetchone()[0] > 0:
+            conn.close()
+            return
 
     for cat in ["food", "fashion", "tech", "travel", "beauty", "fitness", "lifestyle", "home"]:
         compute_for_category(cursor, cat)
 
     conn.commit()
     conn.close()
     print("所有 baseline 统计指标已计算完毕")
-
-
-if __name__ == "__main__":
-    main()

backend/app/database/schema.py

@@ -0,0 +1,84 @@
+import sqlite3
+
+
+def create_tables(conn: sqlite3.Connection):
+    conn.execute("""
+        CREATE TABLE IF NOT EXISTS notes (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            category TEXT NOT NULL,
+            title TEXT NOT NULL,
+            title_length INTEGER,
+            content TEXT,
+            tags TEXT,
+            publish_hour INTEGER,
+            likes INTEGER DEFAULT 0,
+            collects INTEGER DEFAULT 0,
+            comments INTEGER DEFAULT 0,
+            followers INTEGER DEFAULT 0,
+            is_viral INTEGER DEFAULT 0,
+            cover_has_face INTEGER DEFAULT 0,
+            cover_text_ratio REAL DEFAULT 0,
+            cover_saturation REAL DEFAULT 0,
+            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+        )
+    """)
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_notes_category ON notes(category)")
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_notes_viral ON notes(category, is_viral)")
+
+    conn.execute("""
+        CREATE TABLE IF NOT EXISTS baseline_stats (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            category TEXT NOT NULL,
+            metric_name TEXT NOT NULL,
+            metric_value REAL,
+            metric_json TEXT,
+            updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+            UNIQUE(category, metric_name)
+        )
+    """)
+
+    conn.execute("""
+        CREATE TABLE IF NOT EXISTS diagnosis_history (
+            id TEXT PRIMARY KEY,
+            title TEXT NOT NULL,
+            category TEXT NOT NULL,
+            overall_score REAL,
+            grade TEXT,
+            report_json TEXT NOT NULL,
+            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+        )
+    """)
+    conn.execute("""
+        CREATE INDEX IF NOT EXISTS idx_history_created
+        ON diagnosis_history(created_at DESC)
+    """)
+
+    conn.execute("""
+        CREATE TABLE IF NOT EXISTS usage_log (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            ip TEXT NOT NULL,
+            action TEXT NOT NULL DEFAULT 'diagnose',
+            title TEXT DEFAULT '',
+            category TEXT DEFAULT '',
+            total_tokens INTEGER DEFAULT 0,
+            duration_sec REAL DEFAULT 0,
+            status TEXT DEFAULT 'ok',
+            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+        )
+    """)
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_usage_created ON usage_log(created_at DESC)")
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_usage_ip ON usage_log(ip)")
+
+    conn.execute("""
+        CREATE TABLE IF NOT EXISTS visit_log (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            visitor_hash TEXT NOT NULL,
+            user_agent_hash TEXT DEFAULT '',
+            path TEXT NOT NULL,
+            referrer TEXT DEFAULT '',
+            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+        )
+    """)
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_visit_created ON visit_log(created_at DESC)")
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_visit_visitor ON visit_log(visitor_hash)")
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_visit_path ON visit_log(path)")

scripts/seed_data.py → backend/app/database/seed_data.py

@@ -1,16 +1,8 @@
-"""
-生成模拟 baseline 种子数据用于开发和演示。
-实际比赛前应替换为真实采集的小红书笔记数据。
-
-Usage:
-    python scripts/seed_data.py
-"""
 import sqlite3
 import json
 import random
-import os
+from pathlib import Path
 
-DB_PATH = os.path.join(os.path.dirname(__file__), "..", "backend", "data", "baseline.db")
 
 FOOD_TITLES = [
     "手把手教你做日式溏心蛋！零失败！", "一周减脂餐分享｜好吃不胖",
@@ -41,7 +33,7 @@
     "这个APP改变了我的学习方式", "数码产品年度盘点｜好用到哭",
     "iPad学习法｜从学渣到学霸", "耳机横评｜千元内最值得买的5款",
     "NAS入门指南｜打造私人云存储", "手机摄影技巧｜拍出电影质感",
-    "机械键盘入坑指南｜新手必看", "二手数码避坑指南‼️", 
+    "机械键盘入坑指南｜新手必看", "二手数码避坑指南‼️",
     "AI工具合集｜效率提升10倍", "极简桌面布置｜打造高效工作台",
 ]
 
@@ -153,12 +145,18 @@ def generate_notes(category, titles, tags_pool, count=500):
     return notes
 
 
-def seed():
-    """写入种子数据"""
-    conn = sqlite3.connect(DB_PATH)
+def seed_data(db_path: Path, force: bool = False):
+    """若 notes 表为空，则填充种子数据；force=True 时清空后重写"""
+    conn = sqlite3.connect(db_path)
     cursor = conn.cursor()
 
-    cursor.execute("DELETE FROM notes")
+    if force:
+        cursor.execute("DELETE FROM notes")
+    else:
+        cursor.execute("SELECT COUNT(*) FROM notes")
+        if cursor.fetchone()[0] > 0:
+            conn.close()
+            return
 
     all_notes = []
     all_notes.extend(generate_notes("food", FOOD_TITLES, FOOD_TAGS, 500))
@@ -181,7 +179,3 @@ def seed():
     conn.commit()
     print(f"已插入 {len(all_notes)} 条种子数据")
     conn.close()
-
-
-if __name__ == "__main__":
-    seed()

backend/app/local_memory.py

@@ -12,10 +12,11 @@
 import logging
 import os
 from datetime import datetime
+from pathlib import Path
 
 logger = logging.getLogger("noterx.local_memory")
 
-_DATA_ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "data"))
+_DATA_ROOT = Path(__file__).parent.parent / "data"
 WORKSPACE_ROOT = os.path.join(_DATA_ROOT, "noterx_workspace")
 MEMORY_MD = os.path.join(WORKSPACE_ROOT, "MEMORY.md")
 MEMORY_DIR = os.path.join(WORKSPACE_ROOT, "memory")

backend/app/main.py

@@ -3,7 +3,7 @@
 """
 import logging
 import os
-import sqlite3
+from pathlib import Path
 from contextlib import asynccontextmanager
 
 from fastapi import FastAPI
@@ -13,69 +13,17 @@
 
 from app.api.routes import router as api_router
 from app import local_memory
+from app.database import init_database
 
+DB_PATH = Path(__file__).parent.parent / "data" / "baseline.db"
 FRONTEND_DIST = os.path.join(os.path.dirname(__file__), "..", "..", "frontend", "dist")
 
-DB_PATH = os.path.join(os.path.dirname(__file__), "..", "data", "baseline.db")
-
-
-def _ensure_history_table():
-    """启动时自动创建 diagnosis_history 表（如不存在）"""
-    os.makedirs(os.path.dirname(DB_PATH), exist_ok=True)
-    conn = sqlite3.connect(DB_PATH)
-    conn.execute("""
-        CREATE TABLE IF NOT EXISTS diagnosis_history (
-            id TEXT PRIMARY KEY,
-            title TEXT NOT NULL,
-            category TEXT NOT NULL,
-            overall_score REAL,
-            grade TEXT,
-            report_json TEXT NOT NULL,
-            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
-        )
-    """)
-    conn.execute("""
-        CREATE INDEX IF NOT EXISTS idx_history_created
-        ON diagnosis_history(created_at DESC)
-    """)
-    # Usage tracking table
-    conn.execute("""
-        CREATE TABLE IF NOT EXISTS usage_log (
-            id INTEGER PRIMARY KEY AUTOINCREMENT,
-            ip TEXT NOT NULL,
-            action TEXT NOT NULL DEFAULT 'diagnose',
-            title TEXT DEFAULT '',
-            category TEXT DEFAULT '',
-            total_tokens INTEGER DEFAULT 0,
-            duration_sec REAL DEFAULT 0,
-            status TEXT DEFAULT 'ok',
-            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
-        )
-    """)
-    conn.execute("CREATE INDEX IF NOT EXISTS idx_usage_created ON usage_log(created_at DESC)")
-    conn.execute("CREATE INDEX IF NOT EXISTS idx_usage_ip ON usage_log(ip)")
-    conn.execute("""
-        CREATE TABLE IF NOT EXISTS visit_log (
-            id INTEGER PRIMARY KEY AUTOINCREMENT,
-            visitor_hash TEXT NOT NULL,
-            user_agent_hash TEXT DEFAULT '',
-            path TEXT NOT NULL,
-            referrer TEXT DEFAULT '',
-            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
-        )
-    """)
-    conn.execute("CREATE INDEX IF NOT EXISTS idx_visit_created ON visit_log(created_at DESC)")
-    conn.execute("CREATE INDEX IF NOT EXISTS idx_visit_visitor ON visit_log(visitor_hash)")
-    conn.execute("CREATE INDEX IF NOT EXISTS idx_visit_path ON visit_log(path)")
-    conn.commit()
-    conn.close()
-    local_memory.ensure_memory_md()
-
 
 @asynccontextmanager
 async def lifespan(_app: FastAPI):
     """应用生命周期：启动时自动建表"""
-    _ensure_history_table()
+    init_database(DB_PATH)
+    local_memory.ensure_memory_md()
     yield
 
 logging.basicConfig(
@@ -178,12 +126,10 @@ async def serve_app():
 async def health():
     """详细健康检查，含数据库探测"""
     import sqlite3
-    import os
-    db_path = os.path.join(os.path.dirname(__file__), "..", "data", "baseline.db")
     db_ok = False
     note_count = 0
     try:
-        conn = sqlite3.connect(db_path)
+        conn = sqlite3.connect(DB_PATH)
         cur = conn.cursor()
         cur.execute("SELECT COUNT(*) FROM notes")
         note_count = cur.fetchone()[0]

deploy_backend.py

@@ -89,10 +89,7 @@ def run(ssh, cmd, check=True):
 run(ssh, f"{REMOTE_DIR}/backend/venv/bin/pip install -r {REMOTE_DIR}/backend/requirements.txt")
 
 # Init DB
-print("  Initializing database...")
-run(ssh, f"cd {REMOTE_DIR} && {REMOTE_DIR}/backend/venv/bin/python scripts/init_db.py", check=False)
-run(ssh, f"cd {REMOTE_DIR} && {REMOTE_DIR}/backend/venv/bin/python scripts/seed_data.py", check=False)
-run(ssh, f"cd {REMOTE_DIR} && {REMOTE_DIR}/backend/venv/bin/python scripts/compute_baseline.py", check=False)
+print("  Database tables / seed data / baseline will be handled by app startup")
 
 # Upload .env
 print("  Uploading .env...")