First, when performing an upload using an upload session, data is streamed directly from the input stream to the request stream. (It is no longer read into an intermediate buffer first.)
Hopefully this will help these uploads respond to throttling behavior better.
Second, these backends now check for and respect a Retry-After header if it is present.
On 5xx errors, the existing exponential backoff is used for the intermediate retries.
A few other minor things have also been cleaned up (e.g., using Task.Delay instead of Thread.Sleep)
The SendRequest method performs a comparison between an object of type T
and null. While there are no use cases currently where the generic type
is a value type, we should constrain the generic type to be a class so
that the comparison remains meaningful.
This reverts commit ecb670a2d9.
While there aren't any cases where the generic type is a struct, the
comparison with default(T) is incorrect. If T is int, we wouldn't want
the behavior to differ based on whether the value was zero or nonzero.
This way, the regular backend manager logic can handle and log the delete of a non-existant file.
Also update some of the other operations to throw FileMissingException (rather than the backend specific DriveItemNotFoundException).
This class extends HttpClient and uses an OAuthHttpMessageHandler under the covers to automatically authenticate requests.
Additionally, it automatically respects the global HttpContextSettings for overall timeout, read/write timeout, and SSL certificate validation.
(BufferRequests isn't currently handled, as HttpClient doesn't seem to easily expose that flag, and as of .NET 4.5, seems to not buffer by default.)
For SharePoint, this means instead of giving the site ID as a parameter (and a relative path as the backup URL), the full URL to the folder in SharePoint (e.g., 'sharepoint://{tenant}.sharepoint.com/{PathToSite}/{DocumentBase}/{subfolder}') can be given, and the site ID will be looked up.
(This should match the existing SharePoint backend, including the convention of using '//' to indicate the document root, though I can't explicitly test that since the SharePoint sites I have access to require 2FA, which the old backend doesn't support.)
I've tested both explicit ID and site path methods of specifying the destination, and they appear to be working.
For Office 365 groups, it now supports looking up a group by its email address.
These backends are implemented via the common base class, MicrosoftGraphBackend, with only a few minor tweaks (different id parameters, protocol keys, descriptions, etc.)
All three backends support streaming, quota retrieval, and rename.
Kenneth Skovhede