the earthaccess.EarthAccessFile wrapper need not subclass anything

CHANGELOG.md

@@ -39,6 +39,8 @@
   ([#508](https://github.com/nsidc/earthaccess/issues/508))
 * Create destination directory prior to direct S3 downloads, if it doesn't
   already exist ([#562](https://github.com/nsidc/earthaccess/issues/562))
+* Remove the base class on `EarthAccessFile` to fix method resolution
+  ([#610](https://github.com/nsidc/earthaccess/issues/610))
 
 ## [v0.9.0] 2024-02-28
 

docs/user-reference/store/earthaccessfile.md

@@ -0,0 +1,7 @@
+# Documentation for `EarthAccessFile`
+
+::: earthaccess.store.EarthAccessFile
+    options:
+      inherited_members: true
+    show_root_heading: true
+    show_source: false

earthaccess/api.py

@@ -10,7 +10,7 @@
 from .auth import Auth
 from .results import DataCollection, DataGranule
 from .search import CollectionQuery, DataCollections, DataGranules, GranuleQuery
-from .store import Store
+from .store import EarthAccessFile, Store
 from .system import PROD, System
 from .utils import _validation as validate
 
@@ -211,8 +211,8 @@ def download(
 def open(
     granules: Union[List[str], List[DataGranule]],
     provider: Optional[str] = None,
-) -> List[AbstractFileSystem]:
-    """Returns a list of fsspec file-like objects that can be used to access files
+) -> List[EarthAccessFile]:
+    """Returns a list of file-like objects that can be used to access files
     hosted on S3 or HTTPS by third party libraries like xarray.
 
     Parameters:
@@ -221,7 +221,7 @@ def open(
         provider: e.g. POCLOUD, NSIDC_CPRD, etc.
 
     Returns:
-        a list of s3fs "file pointers" to s3 files.
+        A list of "file pointers" to remote (i.e. s3 or https) files.
     """
     provider = _normalize_location(provider)
     results = earthaccess.__store__.open(granules=granules, provider=provider)

earthaccess/store.py

@@ -25,8 +25,22 @@
 logger = logging.getLogger(__name__)
 
 
-class EarthAccessFile(fsspec.spec.AbstractBufferedFile):
-    def __init__(self, f: fsspec.AbstractFileSystem, granule: DataGranule) -> None:
+class EarthAccessFile:
+    """Handle for a file-like object pointing to an on-prem or Earthdata Cloud granule."""
+
+    def __init__(
+        self, f: fsspec.spec.AbstractBufferedFile, granule: DataGranule
+    ) -> None:
+        """EarthAccessFile connects an Earthdata search result with an open file-like object.
+
+        No methods exist on the class, which passes all attribute and method calls
+        directly to the file-like object given during initialization. An instance of
+        this class can be treated like that file-like object itself.
+
+        Parameters:
+            f: a file-like object
+            granule: a granule search result
+        """
         self.f = f
         self.granule = granule
 
@@ -42,14 +56,14 @@ def __reduce__(self) -> Any:
         )
 
     def __repr__(self) -> str:
-        return str(self.f)
+        return repr(self.f)
 
 
 def _open_files(
     url_mapping: Mapping[str, Union[DataGranule, None]],
     fs: fsspec.AbstractFileSystem,
     threads: Optional[int] = 8,
-) -> List[fsspec.AbstractFileSystem]:
+) -> List[EarthAccessFile]:
     def multi_thread_open(data: tuple) -> EarthAccessFile:
         urls, granule = data
         return EarthAccessFile(fs.open(urls), granule)
@@ -301,17 +315,17 @@ def open(
         self,
         granules: Union[List[str], List[DataGranule]],
         provider: Optional[str] = None,
-    ) -> List[Any]:
-        """Returns a list of fsspec file-like objects that can be used to access files
+    ) -> List[EarthAccessFile]:
+        """Returns a list of file-like objects that can be used to access files
         hosted on S3 or HTTPS by third party libraries like xarray.
 
         Parameters:
-            granules: a list of granules(DataGranule) instances or list of URLs,
-                e.g. s3://some-granule
-            provider: an option
+            granules: a list of granule instances **or** list of URLs, e.g. `s3://some-granule`.
+                If a list of URLs is passed, we need to specify the data provider.
+            provider: e.g. POCLOUD, NSIDC_CPRD, etc.
 
         Returns:
-            A list of s3fs "file pointers" to s3 files.
+            A list of "file pointers" to remote (i.e. s3 or https) files.
         """
         if len(granules):
             return self._open(granules, provider)
@@ -323,17 +337,6 @@ def _open(
         granules: Union[List[str], List[DataGranule]],
         provider: Optional[str] = None,
     ) -> List[Any]:
-        """Returns a list of fsspec file-like objects that can be used to access files
-        hosted on S3 or HTTPS by third party libraries like xarray.
-
-        Parameters:
-            granules: a list of granules(DataGranule) instances or list of URLs,
-                e.g. s3://some-granule
-            provider: an option
-
-        Returns:
-            A list of s3fs "file pointers" to s3 files.
-        """
         raise NotImplementedError("granules should be a list of DataGranule or URLs")
 
     @_open.register

mkdocs.yml

@@ -97,6 +97,7 @@ nav:
               - "Granule Queries": "user-reference/granules/granules-query.md"
               - "Granule Results": "user-reference/granules/granules.md"
           - Store:
+              - "EarthAccessFile": "user-reference/store/earthaccessfile.md"
               - "Store": "user-reference/store/store.md"
           - Auth:
               - "Auth": "user-reference/auth/auth.md"

tests/unit/test_store.py

@@ -7,6 +7,7 @@
 import responses
 import s3fs
 from earthaccess import Auth, Store
+from earthaccess.store import EarthAccessFile
 
 
 class TestStoreSessions(unittest.TestCase):
@@ -129,3 +130,12 @@ def test_store_can_create_s3_fsspec_session(self):
             store.get_s3fs_session()
 
         return None
+
+
+def test_earthaccess_file_getattr():
+    fs = fsspec.filesystem("memory")
+    with fs.open("/foo", "wb") as f:
+        earthaccess_file = EarthAccessFile(f, granule="foo")
+        assert f.tell() == earthaccess_file.tell()
+    # cleanup
+    fs.store.clear()