Mz/move strings

[mayeul-zama]

No description provided.

[tmontaigu]

are we sure that we want to move strings into the hlapi module ?

[tmontaigu]

needs to be removed or add a target to run string tests and make sure it runs in the CI

[tmontaigu]

Hum, thing is that I would provably move the, is the create_trivial_ascii from the ClientKey to the ServerKey as its what exists for integers, imo makes a bit more sense as the user could need the ability to create trivial data, and might prevent user from mistaking trivial as a real encryption

[mayeul-zama]

Hum, thing is that I would provably move the, is the create_trivial_ascii from the ClientKey to the ServerKey as its what exists for integers, imo makes a bit more sense as the user could need the ability to create trivial data, and might prevent user from mistaking trivial as a real encryption

I agree it makes more sense to have it on the ServerKey.
But in this case, when switching from encryption to trivial encryption in tests, we must pass a ServerKey instead of a ClientKey which may need changes inmany places.
Maybe we could have this method on the ServerKey, and on the ClientKey behind #[cfg(test)] (applies to all APIs levels)

What do you think?

[tmontaigu]

Hum, thing is that I would provably move the, is the create_trivial_ascii from the ClientKey to the ServerKey as its what exists for integers, imo makes a bit more sense as the user could need the ability to create trivial data, and might prevent user from mistaking trivial as a real encryption

I agree it makes more sense to have it on the ServerKey.

But in this case, when switching from encryption to trivial encryption in tests, we must pass a ServerKey instead of a ClientKey which may need changes inmany places.

Maybe we could have this method on the ServerKey, and on the ClientKey behind #[cfg(test)] (applies to all APIs levels)

What do you think?

That's ok i think

[tmontaigu]

Looks good, needs to fix CI