[BUG] Fix 404 NOT FOUND issue caused by endpoint tail slash (#2721)

* Fix 404 NOT FOUND BUG caused by tail slash

Signed-off-by: Mingqi Hu <[email protected]>

* Update src/huggingface_hub/constants.py

Co-authored-by: Célina <[email protected]>

* Update src/huggingface_hub/utils/_http.py

Co-authored-by: Célina <[email protected]>

* Update src/huggingface_hub/utils/_http.py

Co-authored-by: Célina <[email protected]>

* Update src/huggingface_hub/constants.py

Co-authored-by: Célina <[email protected]>

* run style

---------

Signed-off-by: Mingqi Hu <[email protected]>
Co-authored-by: Célina <[email protected]>

Author: Mingqi2

src/huggingface_hub/constants.py

@@ -2,6 +2,7 @@
 import re
 import typing
 from typing import Literal, Optional, Tuple
+from urllib.parse import urljoin
 
 
 # Possible values for env variables
@@ -63,9 +64,11 @@ def _as_int(value: Optional[str]) -> Optional[int]:
 
 _HF_DEFAULT_ENDPOINT = "https://huggingface.co"
 _HF_DEFAULT_STAGING_ENDPOINT = "https://hub-ci.huggingface.co"
-ENDPOINT = os.getenv("HF_ENDPOINT") or (_HF_DEFAULT_STAGING_ENDPOINT if _staging_mode else _HF_DEFAULT_ENDPOINT)
+ENDPOINT = os.getenv("HF_ENDPOINT", "").rstrip("/") or (
+    _HF_DEFAULT_STAGING_ENDPOINT if _staging_mode else _HF_DEFAULT_ENDPOINT
+)
 
-HUGGINGFACE_CO_URL_TEMPLATE = ENDPOINT + "/{repo_id}/resolve/{revision}/{filename}"
+HUGGINGFACE_CO_URL_TEMPLATE = urljoin(ENDPOINT, "/{repo_id}/resolve/{revision}/{filename}")
 HUGGINGFACE_HEADER_X_REPO_COMMIT = "X-Repo-Commit"
 HUGGINGFACE_HEADER_X_LINKED_ETAG = "X-Linked-Etag"
 HUGGINGFACE_HEADER_X_LINKED_SIZE = "X-Linked-Size"

src/huggingface_hub/utils/_http.py

@@ -338,9 +338,9 @@ def fix_hf_endpoint_in_url(url: str, endpoint: Optional[str]) -> str:
 
     This is useful when using a proxy and the Hugging Face Hub returns a URL with the default endpoint.
     """
-    endpoint = endpoint or constants.ENDPOINT
+    endpoint = endpoint.rstrip("/") if endpoint else constants.ENDPOINT
     # check if a proxy has been set => if yes, update the returned URL to use the proxy
-    if endpoint not in (None, constants._HF_DEFAULT_ENDPOINT, constants._HF_DEFAULT_STAGING_ENDPOINT):
+    if endpoint not in (constants._HF_DEFAULT_ENDPOINT, constants._HF_DEFAULT_STAGING_ENDPOINT):
         url = url.replace(constants._HF_DEFAULT_ENDPOINT, endpoint)
         url = url.replace(constants._HF_DEFAULT_STAGING_ENDPOINT, endpoint)
     return url