Use upstream cpp-base64 git url

[jadh4v]

set(cpp_base64_GIT_REPOSITORY "https://github.com/CodeFinder2/cpp-base64.git")

Ideally improvement to dependency should be contributed upstream and if these contribution are under review and waiting for their integration is not possible, a more permanent fork should be created outside of a user specific account (e.g https://github.com/InsightSoftwareConsortium)

Considering that https://github.com/CodeFinder2/cpp-base64/commit/9144cd53be930b37235ae552a92b5d2aa51e9325 is available in the upstream, I then suggest to use it.

Originally posted by @jcfr in https://github.com/InsightSoftwareConsortium/itk-wasm/pull/670#discussion_r998600613

[thewtex]

Note that CodeFinder2 is not upstream for cpp-base64, it is ReneNyffenegger.

[jcfr]

Good point, we should then fork the upstream into either the KitwareMedical or InsightSoftwareConsortium and use our usual naming convention for staging branch.

That said, since the decode/encode is already available in ITK through Base64.h within the itksys module, I am wondering if we should simply remove that extra dependency.

[thewtex]

Good point, we should then fork the upstream into either the KitwareMedical or InsightSoftwareConsortium and use our usual naming convention for staging branch.

Yes.

Fork created here: https://github.com/InsightSoftwareConsortium/cpp-base64

That said, since the decode/encode is already available in ITK through Base64.h within the itksys module, I am wondering if we should simply remove that extra dependency.

May be worth investigating.

cpp-base64 has the following features:

• C++ interface
• C++17 support
• Performance optimized

That said, it would be nice to remove the dependency if it found that there are not performance regressions and the interface is still usable.