docs: enhance CLI help with argument descriptions, --config option, and epilog examples

src/cli/main.py

@@ -2,15 +2,15 @@
 from pathlib import Path
 import sys
 import io
+from typing import Annotated
 
 import typer
 import uvicorn
 
-# Windows 终端默认用 GBK, 无法输出 emoji, 强制 UTF-8
+# Windows 终端 GBK → UTF-8
 if sys.stdout.encoding != "utf-8":
     sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
 
-# 确保 src 在路径中
 sys.path.insert(0, str(Path(__file__).parent.parent))
 
 from src.core.config import load_config, DEFAULT_CONFIG_PATH
@@ -19,8 +19,15 @@ from src.core.embedder import create_embedder
 from src.core.ingest import DocumentIngestor
 from src.core.search import Searcher
 
-
+app = typer.Typer(
-app = typer.Typer(name="md-vector-db", help="Markdown 文档向量数据库管理工具")
+    name="md-vector-db",
+    help="Markdown 文档向量数据库 — 入库、嵌入、语义检索",
+    epilog="示例:\n"
+    "  md-vector-db ingest docs/readme.md\n"
+    "  md-vector-db ingest-dir ./md_docs/\n"
+    "  md-vector-db search \"如何配置\" --top-k 5\n"
+    "  md-vector-db serve --port 8000",
+)
 
 
 def _get_components(config_path: str = DEFAULT_CONFIG_PATH):
@@ -33,18 +40,35 @@ def _get_components(config_path: str = DEFAULT_CONFIG_PATH):
     return cfg, searcher, ingestor
 
 
-@app.command()
+# -- 共享选项 --
-def ingest(file_path: str):
+ConfigOption = Annotated[
-    """入库单个 Markdown 文件."""
+    str,
-    _, _, ingestor = _get_components()
+    typer.Option(
+        "--config", "-c",
+        help="配置文件路径 (默认: config.yaml)",
+        show_default="config.yaml",
+    ),
+]
+
+# --- 命令 ---
+
+
+@app.command(help="入库单个 Markdown 文件.")
+def ingest(
+    file_path: Annotated[str, typer.Argument(help="Markdown 文件路径")],
+    config: ConfigOption = DEFAULT_CONFIG_PATH,
+):
+    _, _, ingestor = _get_components(config)
     count = ingestor.ingest_file(file_path)
     typer.echo(f"✅ 已入库: {file_path} ({count} 个 chunks)")
 
 
-@app.command()
+@app.command(help="批量入库目录下所有 .md 文件 (递归).")
-def ingest_dir(dir_path: str):
+def ingest_dir(
-    """入库目录下所有 Markdown 文件."""
+    dir_path: Annotated[str, typer.Argument(help="包含 Markdown 文件的目录路径")],
-    _, _, ingestor = _get_components()
+    config: ConfigOption = DEFAULT_CONFIG_PATH,
+):
+    _, _, ingestor = _get_components(config)
     results = ingestor.ingest_directory(dir_path)
     total = sum(results.values())
     for name, count in results.items():
@@ -52,10 +76,13 @@ def ingest_dir(dir_path: str):
     typer.echo(f"✅ 共入库 {len(results)} 个文件, {total} 个 chunks")
 
 
-@app.command()
+@app.command(help="语义检索已入库的文档.")
-def search(query: str, top_k: int = 10):
+def search(
-    """语义检索."""
+    query: Annotated[str, typer.Argument(help="搜索关键词或自然语言查询")],
-    _, searcher, _ = _get_components()
+    top_k: Annotated[int, typer.Option("--top-k", "-k", help="返回结果数量 (1-100)")] = 10,
+    config: ConfigOption = DEFAULT_CONFIG_PATH,
+):
+    _, searcher, _ = _get_components(config)
     results = searcher.search(query, top_k=top_k)
     if not results:
         typer.echo("未找到匹配结果。")
@@ -69,18 +96,22 @@ def search(query: str, top_k: int = 10):
         typer.echo(preview)
 
 
-@app.command()
+@app.command(help="启动 HTTP API 服务.")
-def serve(port: int = 8000):
+def serve(
-    """启动 HTTP 服务."""
+    port: Annotated[int, typer.Option("--port", "-p", help="监听端口")] = 8000,
+    config: ConfigOption = DEFAULT_CONFIG_PATH,
+):
+    # serve 模式下 config 仅用于校验路径存在，实际由 server 模块自行加载
     typer.echo(f"🚀 启动服务: http://localhost:{port}")
     typer.echo(f"📖 API 文档: http://localhost:{port}/docs")
     uvicorn.run("src.server.app:app", host="0.0.0.0", port=port, reload=False)
 
 
-@app.command()
+@app.command(help="查看向量库统计信息 (collection、chunks 数量、源文件列表).")
-def stats():
+def stats(
-    """查看统计信息."""
+    config: ConfigOption = DEFAULT_CONFIG_PATH,
-    _, searcher, _ = _get_components()
+):
+    _, searcher, _ = _get_components(config)
     info = searcher.get_collection_info()
     sources = searcher.list_sources()
     typer.echo(f"📊 Collection: {info['name']}")