Rename regex variable appropriately and allow more complex chaining

Author: snitish

pandas/io/common.py

@@ -71,7 +71,7 @@
 
 _VALID_URLS = set(uses_relative + uses_netloc + uses_params)
 _VALID_URLS.discard("")
-_RFC_3986_PATTERN = re.compile(r"^[A-Za-z][A-Za-z0-9+\-+.]*(::[A-Za-z0-9+\-+.]+)?://")
+_FSSPEC_URL_PATTERN = re.compile(r"^[A-Za-z][A-Za-z0-9+\-+.]*(::[A-Za-z0-9+\-+.]+)*://")
 
 BaseBufferT = TypeVar("BaseBufferT", bound=BaseBuffer)
 
@@ -291,7 +291,7 @@ def is_fsspec_url(url: FilePath | BaseBuffer) -> bool:
     """
     return (
         isinstance(url, str)
-        and bool(_RFC_3986_PATTERN.match(url))
+        and bool(_FSSPEC_URL_PATTERN.match(url))
         and not url.startswith(("http://", "https://"))
     )
 

pandas/tests/io/test_common.py

@@ -505,6 +505,9 @@ def test_is_fsspec_url_chained():
     # GH#48978 Support chained fsspec URLs
     # See https://filesystem-spec.readthedocs.io/en/latest/features.html#url-chaining.
     assert icom.is_fsspec_url("filecache::s3://pandas/test.csv")
+    assert icom.is_fsspec_url("zip://test.csv::filecache::gcs://bucket/afile.zip")
+    assert icom.is_fsspec_url("filecache::zip://test.csv::gcs://bucket/afile.zip")
+    assert icom.is_fsspec_url("filecache::dask::s3://pandas/test.csv")
     assert not icom.is_fsspec_url("filecache:s3://pandas/test.csv")
     assert not icom.is_fsspec_url("filecache:::s3://pandas/test.csv")
     assert not icom.is_fsspec_url("filecache::://pandas/test.csv")