In the offline documentation, ReadAsync is documented as:
buffer: The buffer into which the asynchronous read operation places the bytes that are read
count: The number of bytes that is less than or equal to the Capacity
However, in the online documentation, it now says:
(http://msdn.microsoft.com/en-us/library/windows.storage.streams.iinputstream.readasync.aspx)
buffer: A buffer that may be used to return the bytes that are read. The return value contains the buffer that holds the results.
and includes an extra remarks section:
Remarks
Always read data from the buffer returned in the
IAsyncOperationWithProgress(IBuffer, UInt32). Don't assume that the input buffer contains the data. Depending on the implementation, the data that's read might be placed into the input buffer, or it might be returned in a different buffer.
So having allocated all of that memory to read, and created a buffer, the function can just return something else, and make me copy it. This is so inefficient for an edge case, as it appears that the buffer returned is the one passed in, but now I
need to check, or worse, copy the bytes.
Is checking the value of the return IBuffer is equal to the one passed in sufficient to know that a copy isn't necessary, or is some other checking necessary?
Anthony Wieser | Wieser Software Ltd |
www.wieser-software.com
