app.py

import gradio as gr
from huggingface_hub import create_repo, upload_file, Repository, whoami
import subprocess
import os, shutil

def duplicate(source_repo, dst_repo, token, repo_type):
    # Creating repos has inconsistent API (https://github.com/huggingface/huggingface_hub/issues/47)
    repo_namespace, dst_id = dst_repo.split("/")
    username = whoami(token)["name"]
    org = None
    if repo_namespace != username:
      org = repo_namespace

    # Create the destination repo
    if repo_type in ["space", "dataset"]:
        # For some reason create_repo does not allow repo_type="model"..., even if documentation says
        # that's the default.
        url = create_repo(dst_id, token=token, organization=org, repo_type=repo_type, space_sdk="gradio", private=False)
    else:
        url = create_repo(dst_id, token=token, organization=org, private=False)
    
    # Clone source repo
    endpoint = "huggingface.co/"
    if repo_type in ["space", "dataset"]:
        endpoint += repo_type + "/"
    full_path = f"https://{username}:{token}@{endpoint}{source_repo}"
    local_dir = "hub/" + source_repo
    
    if repo_type in ["space", "dataset"]:
        # Same as above
        repo = Repository(local_dir=local_dir, clone_from=full_path, repo_type=repo_type)
    else:
        repo = Repository(local_dir=local_dir, clone_from=full_path)
    
    for root, dirs, files in  os.walk(local_dir):
      if not root.startswith("."):
        if repo_type == "model":
          repo_type = None
        for f in files:
            if not f.startswith("."):
                if ".git" not in root:
                    # remove hub/namespace/reponame
                    directory_path_in_repo = "/".join(root.split("/")[3:]) 
                    path_in_repo = os.path.join(directory_path_in_repo, f)
                    local_file_path = os.path.join(local_dir, path_in_repo)
                    print("From: ", local_file_path, " to: ", path_in_repo)
                    upload_file(path_or_fileobj=local_file_path, path_in_repo=path_in_repo, repo_id=dst_repo, token=token, repo_type=repo_type)


    # Clean up to be nice with the environment
    for filename in os.listdir(local_dir):
      file_path = os.path.join(local_dir, filename)
      if os.path.isfile(file_path) or os.path.islink(file_path):
          os.unlink(file_path)
      elif os.path.isdir(file_path):
          shutil.rmtree(file_path)
    
    return f"Find your repo <a href='{url}' target=\"_blank\" style=\"text-decoration:underline\">here</a>"

interface = gr.Interface(
    fn=duplicate,
    inputs=[
        gr.inputs.Textbox(placeholder="Source repository (e.g. osanseviero/src)"),
        gr.inputs.Textbox(placeholder="Destination repository (e.g. osanseviero/dst)"),
        gr.inputs.Textbox(placeholder="Write access token"),
        gr.inputs.Dropdown(choices=["model", "dataset", "space"])
    ],
    outputs=["html", "image"] ,
    title="Duplicate your repo!",
    description="Duplicate a Hugging Face repository! You need to specify a write token obtained in https://hf.co/settings/token. This Space is a an experimental demo.",
    article="<p>Find your write token at <a href='https://huggingface.co/settings/token' target='_blank'>token settings</a></p>",
    allow_flagging=False,
    live=False,
)
interface.launch(enable_queue=True)