Add Gradio zip downloader app

app.py

@@ -1,7 +1,181 @@
+import os
+import zipfile
+import tempfile
+import requests
 import gradio as gr
+from urllib.parse import quote
+
+
+HF_BASE = os.environ.get("HF_ENDPOINT", "https://huggingface.co")
+HF_TOKEN = os.environ.get("HF_TOKEN")
+
+
+def get_headers():
+    headers = {
+        "User-Agent": "hf-zip-gradio-demo/1.0"
+    }
+
+    if HF_TOKEN:
+        headers["Authorization"] = f"Bearer {HF_TOKEN}"
+
+    return headers
+
+
+def encode_repo(repo: str) -> str:
+    return "/".join(quote(part, safe="") for part in repo.split("/"))
+
+
+def encode_path(path: str) -> str:
+    return "/".join(quote(part, safe="") for part in path.split("/"))
+
+
+def get_repo_api_path(repo_type: str, repo: str) -> str:
+    encoded_repo = encode_repo(repo)
+
+    if repo_type == "dataset":
+        return f"/api/datasets/{encoded_repo}"
+
+    if repo_type == "space":
+        return f"/api/spaces/{encoded_repo}"
+
+    return f"/api/models/{encoded_repo}"
+
+
+def get_repo_resolve_prefix(repo_type: str, repo: str) -> str:
+    encoded_repo = encode_repo(repo)
+
+    if repo_type == "dataset":
+        return f"/datasets/{encoded_repo}"
+
+    if repo_type == "space":
+        return f"/spaces/{encoded_repo}"
+
+    return f"/{encoded_repo}"
+
+
+def list_repo_files(repo: str, revision: str, repo_type: str):
+    api_path = get_repo_api_path(repo_type, repo)
+    url = f"{HF_BASE}{api_path}/tree/{quote(revision, safe='')}?recursive=1"
+
+    response = requests.get(url, headers=get_headers(), timeout=60)
+
+    if not response.ok:
+        raise RuntimeError(f"获取文件列表失败：{response.status_code} {response.text[:500]}")
+
+    items = response.json()
+
+    files = []
+    for item in items:
+        if item.get("type") == "file":
+            files.append({
+                "path": item.get("path"),
+                "size": item.get("size", 0)
+            })
+
+    return files
+
+
+def get_file_download_url(repo: str, repo_type: str, revision: str, path: str) -> str:
+    prefix = get_repo_resolve_prefix(repo_type, repo)
+
+    return (
+        f"{HF_BASE}{prefix}"
+        f"/resolve/{quote(revision, safe='')}"
+        f"/{encode_path(path)}"
+    )
+
+
+def download_repo_as_zip(repo: str, revision: str, repo_type: str, progress=gr.Progress()):
+    repo = repo.strip()
+    revision = revision.strip() or "main"
+
+    if not repo or "/" not in repo:
+        raise gr.Error("repo 参数格式错误，应为 owner/name，例如 sshleifer/tiny-gpt2")
+
+    files = list_repo_files(repo, revision, repo_type)
+
+    if not files:
+        raise gr.Error("没有找到文件")
+
+    safe_repo_name = repo.split("/")[-1].replace("/", "_")
+    zip_path = os.path.join(
+        tempfile.gettempdir(),
+        f"{safe_repo_name}-{revision}.zip"
+    )
+
+    progress(0, desc=f"找到 {len(files)} 个文件，开始打包...")
+
+    with zipfile.ZipFile(zip_path, "w", compression=zipfile.ZIP_STORED, allowZip64=True) as zip_file:
+        for index, file in enumerate(files):
+            file_path = file["path"]
+
+            progress(
+                index / len(files),
+                desc=f"正在下载并写入：{file_path}"
+            )
+
+            url = get_file_download_url(
+                repo=repo,
+                repo_type=repo_type,
+                revision=revision,
+                path=file_path
+            )
+
+            response = requests.get(url, headers=get_headers(), stream=True, timeout=120)
+
+            if not response.ok:
+                raise gr.Error(f"下载文件失败：{file_path}，状态码：{response.status_code}")
+
+            with tempfile.NamedTemporaryFile(delete=False) as temp_file:
+                temp_file_path = temp_file.name
+
+                try:
+                    for chunk in response.iter_content(chunk_size=1024 * 1024):
+                        if chunk:
+                            temp_file.write(chunk)
+
+                    temp_file.flush()
+                    zip_file.write(temp_file_path, arcname=file_path)
+
+                finally:
+                    if os.path.exists(temp_file_path):
+                        os.remove(temp_file_path)
+
+    progress(1, desc="ZIP 生成完成")
+
+    return zip_path
+
+
+with gr.Blocks(title="Hugging Face 仓库 ZIP 下载器") as demo:
+    gr.Markdown("# Hugging Face 仓库 ZIP 下载器")
+    gr.Markdown("输入模型、数据集或 Space 仓库名，生成 ZIP 文件下载。")
+
+    with gr.Row():
+        repo = gr.Textbox(
+            label="仓库名",
+            value="sshleifer/tiny-gpt2",
+            placeholder="例如：Qwen/Qwen2.5-0.5B-Instruct"
+        )
+
+        revision = gr.Textbox(
+            label="分支 / revision",
+            value="main"
+        )
+
+        repo_type = gr.Dropdown(
+            label="仓库类型",
+            choices=["model", "dataset", "space"],
+            value="model"
+        )
+
+    button = gr.Button("生成 ZIP")
+    output_file = gr.File(label="下载 ZIP 文件")
+
+    button.click(
+        fn=download_repo_as_zip,
+        inputs=[repo, revision, repo_type],
+        outputs=output_file
+    )
 
-def greet(name):
-    return "Hello " + name + "!!"
 
-demo = gr.Interface(fn=greet, inputs="text", outputs="text")
 demo.launch()

requirements.txt

@@ -0,0 +1 @@
+requests