Add demo seed data for Hugging Face Space.

Store sample data as seed/demo.sql (text) and import on first startup, avoiding HF binary file restrictions while keeping the public demo populated.

Co-authored-by: Cursor <cursoragent@cursor.com>

.dockerignore

@@ -7,4 +7,5 @@ __pycache__
 .mypy_cache
 data/
 *.db
+seed/*.db
 .pre-commit-config.yaml

.gitignore

@@ -18,8 +18,9 @@ wheels/
 # 本地数据库目录
 data/
 
-# 数据库文件
+# 数据库文件（Demo 使用 seed/demo.sql 文本导入）
 *.db
+seed/*.db
 
 # 环境变量
 .env

Dockerfile

@@ -7,6 +7,7 @@ RUN mkdir -p /data
 COPY pyproject.toml README.md ./
 COPY qa_annotate/ qa_annotate/
 COPY scripts/ scripts/
+COPY seed/demo.sql seed/demo.sql
 
 RUN pip install --no-cache-dir .
 

scripts/space_init.py

@@ -1,19 +1,53 @@
 #!/usr/bin/env python
-"""Hugging Face Space 启动初始化：建表并可选创建管理员账号。"""
+"""Hugging Face Space 启动初始化：导入 Demo 数据、建表并创建管理员账号。"""
 
 import os
+import sqlite3
 import sys
 from pathlib import Path
 
 project_root = Path(__file__).parent.parent
 sys.path.insert(0, str(project_root))
 
+from qa_annotate.config import settings  # noqa: E402
 from qa_annotate.database.base import SessionLocal, init_db  # noqa: E402
 from qa_annotate.database.crud import UserCRUD  # noqa: E402
 from qa_annotate.schema.user import UserCreate, UserUpdate  # noqa: E402
 from qa_annotate.utils.password import hash_password  # noqa: E402
 
 
+def _project_count(db_path: Path) -> int:
+    if not db_path.exists():
+        return 0
+    with sqlite3.connect(db_path) as conn:
+        try:
+            return conn.execute("SELECT COUNT(*) FROM projects").fetchone()[0]
+        except sqlite3.OperationalError:
+            return 0
+
+
+def seed_demo_database() -> None:
+    if os.environ.get("SEED_DEMO_DATA", "true").lower() not in ("1", "true", "yes"):
+        print("SEED_DEMO_DATA 已关闭，跳过示例数据导入")
+        return
+
+    seed_path = project_root / "seed" / "demo.sql"
+    if not seed_path.exists():
+        print("未找到 seed/demo.sql，跳过示例数据导入")
+        return
+
+    db_path = settings.db_path
+    db_path.parent.mkdir(parents=True, exist_ok=True)
+    if _project_count(db_path) > 0:
+        print("数据库已有项目数据，跳过示例数据导入")
+        return
+
+    sql = seed_path.read_text(encoding="utf-8")
+    with sqlite3.connect(db_path) as conn:
+        conn.executescript(sql)
+    print(f"已从 seed/demo.sql 导入 Demo 示例数据: {db_path}")
+
+
 def ensure_superuser() -> None:
     username = os.environ.get("ADMIN_USERNAME", "admin").strip()
     password = os.environ.get("ADMIN_PASSWORD", "123456")
@@ -49,6 +83,7 @@ def ensure_superuser() -> None:
 
 
 def main() -> None:
+    seed_demo_database()
     init_db()
     ensure_superuser()