FromBase64Transform-sample uses InputBlockSize instead of hard-coded 4 #1098
Conversation
There was a problem hiding this comment.
Thanks for updating the example, @gfoidl. There are also Visual Basic and C++ examples (at <https://github.com/dotnet/samples/snippets/visualbasic/VS_Snippets_CLR_System//system.Security.Cryptography.FromBase64Transform Example/vb/class1.vb> and <https://github.com/dotnet/samples/snippets/cpp/VS_Snippets_CLR_System//system.Security.Cryptography.FromBase64Transform Example/vb/class1.cpp>). Could you make the corresponding changes to those files as well?
Also, code running on .NET Framework and earlier versions of .NET Core will now use 1 as the input block size. It would be good to expand the example description (in the FromBase64Transform.xml file in the dotnet/dotnet-api-docs repo) to note that on .NET implementations with a FromBase64Transform.InputBlockSize that is not 4, a constant value of 4 should be substituted for the property value.
|
For api-docs -> dotnet/dotnet-api-docs#2895 |
|
Thank you, @gfoidl, for modifying the Visual Basic and C++ examples as well. I'll merge this PR and review dotnet/dotnet-api-docs#2895 shortly. |
Depends on dotnet/corefx#39599
myTransform.InputBlockSizereturned1, so the sample code used4explicitely, as base64 decoding transforms 4 input bytes to 3 output bytes.dotnet/corefx#39599 fixes this, so the sample code can use the
InputBlockSizeprovided by the transform.Additional did unify the code comments in the sample.
And removed the redundant
Clear, as this is called byDispose.