[data] Fix HTTP streaming file download by trying both open_input_file and open_input_stream

Signed-off-by: xyuzh <xinyzng@gmail.com>

Author: xyuzh

python/ray/data/_internal/planner/plan_download_op.py

@@ -189,12 +189,27 @@ def download_bytes_threaded(
         def load_uri_bytes(uri_path_iterator):
             """Function that takes an iterator of URI paths and yields downloaded bytes for each."""
             for uri_path in uri_path_iterator:
-                try:
-                    with fs.open_input_file(uri_path) as f:
-                        yield f.read()
-                except OSError as e:
+                last_error = None
+                # Handle both file and stream for uri download
+                for uri_open_func in [fs.open_input_file, fs.open_input_stream]:
+                    try:
+                        with uri_open_func(uri_path) as f:
+                            yield f.read()
+                        break
+                    except ValueError as e:
+                        last_error = e
+                        continue
+                    except OSError as e:
+                        logger.debug(f"OSError: '{uri_path}' from column '{uri_column_name}' with error: {e}")
+                        yield None
+                        break
+                    except Exception as e:
+                        logger.error(f"Unexpected error in load_uri_bytes: {e}")
+                        yield None
+                        break
+                else:
                     logger.debug(
-                        f"Failed to download URI '{uri_path}' from column '{uri_column_name}' with error: {e}"
+                        f"Failed to download URI '{uri_path}' from column '{uri_column_name}' with error: {last_error}"
                     )
                     yield None