feat: Add wasm support

[augustocdias]

This adds wasm support to OpenMls. I've added new workflow jobs to run tests on wasm target. They'll be tested in chrome and firefox.

Some library changes were needed in order to make this work:

• fluvio-wasm-timer was added in order to provide the std::time structs. This is basically a wrapper that will use either std::time or the web-sys crate.
• ed25519-dalek was upgraded because it seems the underlying rand-core is not compatible with wasm
• rayon and criterion will be available just for non wasm targets.

About rayon, at the moment for wasm it will use a regular iterator instead of the parallel. There's the option to add wasm-bindgen-rayon. This will provide parallel iterators for wasm but we could re-export the function init_thread_pool to OpenMls consumers as it needs to be called before using it (only in wasm targets) or we would have to document that they have to add this as dependency and call it. If you're willing to have it I could add it behind a feature so consumers can choose what fits best for them.

What still needs to be done is add the macro #[wasm_bindgen_test] to all tests and the call in each test module to wasm_bindgen_test::wasm_bindgen_test_configure!(run_in_browser). Tests that don't have this will be ignored. At the moment I've added in just one so you can see what kinds of changes need to be done in the tests to run on wasm. If you're okay with it I could adapt and make sure all tests run on wasm. (I can do it in a separate PR if you prefer)

I've noticed that the SingatureKeyPair struct needs a random generator, shouldn't it be using OpenMlsCrypto instead of re-implementing it?

[augustocdias]

There's one test failing. After debugging I found out that in the json file: test_vectors/treekem.json the private keys are with 64 bytes (which is actually the keypair). This is causing this test to fail as the SigningKey is expected to have 32 bytes. If I read the SigningKey as a keypair, the failing test works but it breaks several other tests that are passing 32 bytes in the private key in the SignatureKeypair struct.

What should I do in this case?

[raphaelrobert]

Closing because this is now replaced by #1483.