[WIP] Non-git mixin

.gitignore

@@ -108,6 +108,7 @@ venv/
 ENV/
 env.bak/
 venv.bak/
+.venv*
 
 # Spyder project settings
 .spyderproject

CONTRIBUTING.md

@@ -258,13 +258,14 @@ repository they can be run with the following:
 
 ```bash
 $ HUGGINGFACE_CO_STAGING=1 python -m pytest -sv ./tests
+```
 
-In fact, that's how `make test` is implemented (sans the `pip install` line)!
+In fact, that's how `make test` is implemented (without the `pip install` line)!
 
 You can specify a smaller set of tests in order to test only the feature
 you're working on.
 
-For example, the following will only run the tests hel in the `test_repository.py` file:
+For example, the following will only run the tests in the `test_repository.py` file:
 
 ```bash
 $ HUGGINGFACE_CO_STAGING=1 python -m pytest -sv ./tests/test_repository.py

docs/source/how-to-manage.mdx

@@ -1,6 +1,6 @@
 # Create and manage a repository
 
-A repository is a space for you to store your model or dataset files. This guide will show you how to:
+A repository is a place where you can store your model or dataset files. This guide will show you how to:
 
 * Create and delete a repository.
 * Adjust repository visibility.

setup.py

@@ -35,6 +35,8 @@ def get_version() -> str:
 
 extras["tensorflow"] = ["tensorflow", "pydot", "graphviz"]
 
+extras["ml"] = extras["torch"] + extras["fastai"] + extras["tensorflow"]
+
 extras["testing"] = [
     "pytest",
     "pytest-cov",

src/huggingface_hub/hf_api.py

@@ -51,6 +51,7 @@
     ModelFilter,
     ModelTags,
     _filter_emissions,
+    _generate_url,
 )
 
 
@@ -59,8 +60,6 @@
 else:
     from typing_extensions import Literal
 
-
-REGEX_DISCUSSION_URL = re.compile(r".*/discussions/(\d+)$")
 USERNAME_PLACEHOLDER = "hf_user"
 
 logger = logging.get_logger(__name__)
@@ -100,9 +99,6 @@ def _validate_repo_id_deprecation(repo_id, name, organization):
     return name, organization
 
 
-logger = logging.get_logger(__name__)
-
-
 def repo_type_and_id_from_hf_id(
     hf_id: str, hub_url: Optional[str] = None
 ) -> Tuple[Optional[str], Optional[str], str]:
@@ -1431,7 +1427,7 @@ def list_repo_files(
         )
         return [f.rfilename for f in repo_info.siblings]
 
-    @_deprecate_positional_args
+    @_deprecate_positional_args(version="0.12")
     def create_repo(
         self,
         repo_id: str = None,
@@ -1887,6 +1883,11 @@ def create_commit(
                 " `CommitOperationDelete`"
             )
 
+        logger.debug(
+            f"About to commit to the hub: {len(additions)} addition(s) and"
+            f" {len(additions)} deletion(s)."
+        )
+
         for addition in additions:
             addition.validate()
 
@@ -2060,20 +2061,16 @@ def upload_file(
             revision=revision,
             create_pr=create_pr,
         )
-        if pr_url is not None:
-            re_match = re.match(REGEX_DISCUSSION_URL, pr_url)
-            if re_match is None:
-                raise RuntimeError(
-                    "Unexpected response from the hub, expected a Pull Request URL but"
-                    f" got: '{pr_url}'"
-                )
-            revision = quote(f"refs/pr/{re_match[1]}", safe="")
 
-        if repo_type in REPO_TYPES_URL_PREFIXES:
-            repo_id = REPO_TYPES_URL_PREFIXES[repo_type] + repo_id
-        revision = revision if revision is not None else DEFAULT_REVISION
-        return f"{self.endpoint}/{repo_id}/blob/{revision}/{path_in_repo}"
-        # ^ Similar to `hf_hub_url` but it's "blob" instead of "resolve"
+        return _generate_url(
+            "as_file",
+            endpoint=self.endpoint,
+            repo_id=repo_id,
+            path_in_repo=path_in_repo,
+            repo_type=repo_type,
+            revision=revision,
+            pr_url=pr_url,
+        )
 
     def upload_folder(
         self,
@@ -2172,25 +2169,8 @@ def upload_folder(
             if commit_message is not None
             else f"Upload {path_in_repo} with huggingface_hub"
         )
-        folder_path = os.path.normpath(os.path.expanduser(folder_path))
-        if not os.path.isdir(folder_path):
-            raise ValueError(f"Provided path: '{folder_path}' is not a directory")
-
-        files_to_add: List[CommitOperationAdd] = []
-        for dirpath, _, filenames in os.walk(folder_path):
-            for filename in filenames:
-                abs_path = os.path.join(dirpath, filename)
-                rel_path = os.path.relpath(abs_path, folder_path)
-                files_to_add.append(
-                    CommitOperationAdd(
-                        path_or_fileobj=abs_path,
-                        path_in_repo=os.path.normpath(
-                            os.path.join(path_in_repo, rel_path)
-                        ).replace(os.sep, "/"),
-                    )
-                )
 
-        logger.debug(f"About to upload / commit {len(files_to_add)} files to the Hub")
+        files_to_add = _prepare_upload_folder_commit(folder_path, path_in_repo)
 
         pr_url = self.create_commit(
             repo_type=repo_type,
@@ -2203,21 +2183,15 @@ def upload_folder(
             create_pr=create_pr,
         )
 
-        if pr_url is not None:
-            re_match = re.match(REGEX_DISCUSSION_URL, pr_url)
-            if re_match is None:
-                raise RuntimeError(
-                    "Unexpected response from the hub, expected a Pull Request URL but"
-                    f" got: '{pr_url}'"
-                )
-            revision = quote(f"refs/pr/{re_match[1]}", safe="")
-
-        if repo_type in REPO_TYPES_URL_PREFIXES:
-            repo_id = REPO_TYPES_URL_PREFIXES[repo_type] + repo_id
-
-        revision = revision if revision is not None else DEFAULT_REVISION
-        return f"{self.endpoint}/{repo_id}/tree/{revision}/{path_in_repo}"
-        # ^ Similar to `hf_hub_url` but it's "tree" instead of "resolve"
+        return _generate_url(
+            "as_folder",
+            endpoint=self.endpoint,
+            repo_id=repo_id,
+            path_in_repo=path_in_repo,
+            repo_type=repo_type,
+            revision=revision,
+            pr_url=pr_url,
+        )
 
     def delete_file(
         self,
@@ -2380,6 +2354,30 @@ def delete_token(cls):
             pass
 
 
+def _prepare_upload_folder_commit(
+    folder_path: str, path_in_repo: str
+) -> List[CommitOperationAdd]:
+    """Generate the list of Add operations for a commit to upload a folder."""
+    folder_path = os.path.normpath(os.path.expanduser(folder_path))
+    if not os.path.isdir(folder_path):
+        raise ValueError(f"Provided path: '{folder_path}' is not a directory")
+
+    files_to_add: List[CommitOperationAdd] = []
+    for dirpath, _, filenames in os.walk(folder_path):
+        for filename in filenames:
+            abs_path = os.path.join(dirpath, filename)
+            rel_path = os.path.relpath(abs_path, folder_path)
+            files_to_add.append(
+                CommitOperationAdd(
+                    path_or_fileobj=abs_path,
+                    path_in_repo=os.path.normpath(
+                        os.path.join(path_in_repo, rel_path)
+                    ).replace(os.sep, "/"),
+                )
+            )
+    return files_to_add
+
+
 api = HfApi()
 
 set_access_token = api.set_access_token