fs: add acknowledge_abuse parameter

Related #62
Pre-requisite for iterative/dvc#7832

Author: efiop

docs/filemanagement.rst

@@ -239,6 +239,14 @@ is downloaded. Just set the `remove_bom` parameter in `GetContentString()` or
 `GetContentFile()` - see `examples/strip_bom_example.py` in the GitHub
 repository for an example.
 
+Abusive files
+-------------
+
+Files identified as `abusive`_ (malware, etc.) are only downloadable by the owner.
+If you see a
+'This file has been identified as malware or spam and cannot be downloaded'
+error, set 'acknowledge_abuse=True' parameter in `GetContentFile()`. By using
+it you indicate you acknowledge the risks of downloading potential malware.
 
 .. _`GoogleDriveFile`: /PyDrive2/pydrive2/#pydrive2.files.GoogleDriveFile
 .. _`Upload()`: /PyDrive2/pydrive2/#pydrive2.files.GoogleDriveFile.Upload
@@ -251,3 +259,4 @@ repository for an example.
 .. _`GetContentString()`: ./PyDrive2/pydrive2/#pydrive2.files.GoogleDriveFile.GetContentString
 .. _`official documentation`: https://developers.google.com/drive/v2/reference/files#resource-representations
 .. _`known`: https://productforums.google.com/forum/#!topic/docs/BJLimQDGtjQ
+.. _`abusive`: https://support.google.com/docs/answer/148505?visit_id=637940058767312674-3438409752&rd=1

docs/fsspec.rst

@@ -95,6 +95,13 @@ Using json keyfile string:
 
 Use `client_user_email` if you are using `delegation of authority`_.
 
+Abusive files
+-------------
+
+Use `acknowledge_abuse=True` parameter if you see
+'This file has been identified as malware or spam and cannot be downloaded'
+error. See `Abusive files`_ for more info.
+
 Using filesystem
 ----------------
 
@@ -114,4 +121,5 @@ about and manipulating files, refer to fsspec docs on
 .. _`fsspec`: https://filesystem-spec.readthedocs.io/en/latest/
 .. _`GDriveFileSystem`: /PyDrive2/pydrive2/#pydrive2.fs.GDriveFileSystem
 .. _`delegation of authority`: https://developers.google.com/admin-sdk/directory/v1/guides/delegation
+.. _`Abusive files`: /PyDrive2/filemanagement/index.html#abusive-files
 .. _`how to use a filesystem`: https://filesystem-spec.readthedocs.io/en/latest/usage.html#use-a-file-system

pydrive2/files.py

@@ -295,6 +295,7 @@ def GetContentFile(
         remove_bom=False,
         callback=None,
         chunksize=DEFAULT_CHUNK_SIZE,
+        acknowledge_abuse=False,
     ):
         """Save content of this file as a local file.
 
@@ -308,6 +309,8 @@ def GetContentFile(
         :type param: callable
         :param chunksize: chunksize in bytes (standard 100 MB(1024*1024*100))
         :type chunksize: int
+        :param acknowledge_abuse: Allow abusive files.
+        :type acknowledge_abuse: bool
         :raises: ApiRequestError, FileNotUploadedError
         """
         files = self.auth.service.files()
@@ -331,7 +334,12 @@ def download(fd, request):
             # But that would first require a slow call to FetchMetadata().
             # We prefer to try-except for speed.
             try:
-                download(fd, files.get_media(fileId=file_id))
+                download(
+                    fd,
+                    files.get_media(
+                        fileId=file_id, acknowledgeAbuse=acknowledge_abuse
+                    ),
+                )
             except errors.HttpError as error:
                 exc = ApiRequestError(error)
                 if (
@@ -362,6 +370,7 @@ def GetContentIOBuffer(
         encoding=None,
         remove_bom=False,
         chunksize=DEFAULT_CHUNK_SIZE,
+        acknowledge_abuse=False,
     ):
         """Get a file-like object which has a buffered read() method.
 
@@ -373,6 +382,8 @@ def GetContentIOBuffer(
         :type remove_bom: bool
         :param chunksize: default read()/iter() chunksize.
         :type chunksize: int
+        :param acknowledge_abuse: Allow abusive files.
+        :type acknowledge_abuse: bool
         :returns: MediaIoReadable -- file-like object.
         :raises: ApiRequestError, FileNotUploadedError
         """
@@ -386,7 +397,11 @@ def GetContentIOBuffer(
         # But that would first require a slow call to FetchMetadata().
         # We prefer to try-except for speed.
         try:
-            request = self._WrapRequest(files.get_media(fileId=file_id))
+            request = self._WrapRequest(
+                files.get_media(
+                    fileId=file_id, acknowledgeAbuse=acknowledge_abuse
+                )
+            )
             return MediaIoReadable(
                 request, encoding=encoding, chunksize=chunksize
             )

pydrive2/fs/spec.py

@@ -167,6 +167,7 @@ def __init__(
         client_json_file_path=None,
         use_service_account=False,
         profile=None,
+        acknowledge_abuse=False,
         **kwargs,
     ):
         """Create an instance of GDriveFileSystem.
@@ -192,6 +193,8 @@ def __init__(
         :type use_service_account: bool.
         :param profile: Profile name for caching credentials
             (ignored for service account).
+        :param acknowledge_abuse: Allow downloading abusive files.
+        :type acknowledge_abuse: bool
         :type profile: str.
         :raises: GDriveAuthError
         """
@@ -229,6 +232,7 @@ def __init__(
 
         self.client = GoogleDrive(google_auth)
         self._trash_only = trash_only
+        self._acknowledge_abuse = acknowledge_abuse
 
     def split_path(self, path):
         parts = path.replace("//", "/").rstrip("/").split("/", 1)
@@ -549,7 +553,7 @@ def _gdrive_get_file(self, item_id, rpath, callback=None, block_size=None):
         # it does not create a file on the remote
         gdrive_file = self.client.CreateFile(param)
 
-        extra_args = {}
+        extra_args = {"acknowledge_abuse": self._acknowledge_abuse}
         if block_size:
             extra_args["chunksize"] = block_size
 
@@ -577,7 +581,9 @@ def _gdrive_open_file(self, item_id):
         param = {"id": item_id}
         # it does not create a file on the remote
         gdrive_file = self.client.CreateFile(param)
-        fd = gdrive_file.GetContentIOBuffer()
+        fd = gdrive_file.GetContentIOBuffer(
+            acknowledge_abuse=self._acknowledge_abuse
+        )
         return IterStream(iter(fd))
 
     def rm_file(self, path):