BUG: AttributeError when read_sas used with GCS (#33070)

* BUG: AttributeError when read_sas used with GCS

With GCSFS (and possibly other connectors), the output from they file
read is `bytes`, not a Unicode `str`. The `encode` call is unnecessary
in this case.

* remove unnecessary try/catch for Python 3

* fix: remove unnecessary BytesIO conversion

* cln: remove unused bytesio import

Author: tswast

doc/source/whatsnew/v1.1.0.rst

@@ -426,6 +426,7 @@ I/O
 - Bug in :meth:`read_csv` was causing a segfault when there were blank lines between the header and data rows (:issue:`28071`)
 - Bug in :meth:`read_csv` was raising a misleading exception on a permissions issue (:issue:`23784`)
 - Bug in :meth:`read_csv` was raising an ``IndexError`` when header=None and 2 extra data columns
+- Bug in :meth:`read_sas` was raising an ``AttributeError`` when reading files from Google Cloud Storage (issue:`33069`)
 - Bug in :meth:`DataFrame.to_sql` where an ``AttributeError`` was raised when saving an out of bounds date (:issue:`26761`)
 - Bug in :meth:`read_excel` did not correctly handle multiple embedded spaces in OpenDocument text cells. (:issue:`32207`)
 

pandas/io/sas/sas_xport.py

@@ -9,7 +9,6 @@
 """
 from collections import abc
 from datetime import datetime
-from io import BytesIO
 import struct
 import warnings
 
@@ -263,13 +262,9 @@ def __init__(
         if isinstance(filepath_or_buffer, (str, bytes)):
             self.filepath_or_buffer = open(filepath_or_buffer, "rb")
         else:
-            # Copy to BytesIO, and ensure no encoding
-            contents = filepath_or_buffer.read()
-            try:
-                contents = contents.encode(self._encoding)
-            except UnicodeEncodeError:
-                pass
-            self.filepath_or_buffer = BytesIO(contents)
+            # Since xport files include non-text byte sequences, xport files
+            # should already be opened in binary mode in Python 3.
+            self.filepath_or_buffer = filepath_or_buffer
 
         self._read_header()
 

pandas/tests/io/sas/test_xport.py

@@ -26,6 +26,7 @@ def setup_method(self, datapath):
         self.dirpath = datapath("io", "sas", "data")
         self.file01 = os.path.join(self.dirpath, "DEMO_G.xpt")
         self.file02 = os.path.join(self.dirpath, "SSHSV1_A.xpt")
+        self.file02b = open(os.path.join(self.dirpath, "SSHSV1_A.xpt"), "rb")
         self.file03 = os.path.join(self.dirpath, "DRXFCD_G.xpt")
         self.file04 = os.path.join(self.dirpath, "paxraw_d_short.xpt")
 
@@ -119,6 +120,16 @@ def test2(self):
         data = read_sas(self.file02)
         tm.assert_frame_equal(data, data_csv)
 
+    def test2_binary(self):
+        # Test with SSHSV1_A.xpt, read as a binary file
+
+        # Compare to this
+        data_csv = pd.read_csv(self.file02.replace(".xpt", ".csv"))
+        numeric_as_float(data_csv)
+
+        data = read_sas(self.file02b, format="xport")
+        tm.assert_frame_equal(data, data_csv)
+
     def test_multiple_types(self):
         # Test with DRXFCD_G.xpt (contains text and numeric variables)