Update HPKE encryption of matchkeys according to the latest spec

[akoshelev]

patcg-individual-drafts/ipa#56 asks us to authenticate match key encryption with trigger bit, which we don't do currently.

The entry point into match key encryption is hpke module.

• First thing that must be updated is the HPKE Info struct that must now carry a trigger event indicator. I am not a huge fan of boolean values, but we don't have an enum for event type (yet), so it is probably the right time to add it.

ipa/src/hpke/info.rs

Line 16 in cc48eba

pub struct Info<'a> {

• into_bytes method of Info struct must correctly serialize this indicator as a single byte integer, according to the spec.

ipa/src/hpke/info.rs

Line 76 in cc48eba

pub(super) fn into_bytes(self) -> Box<[u8]> {

After this, we need to update unit tests and property tests to validate the new AAD tag implementation.

• seal and open test methods must take trigger/source event indicator as an parameter

ipa/src/hpke/mod.rs

Line 197 in bc2e7e1

pub fn seal(

ipa/src/hpke/mod.rs

Line 214 in bc2e7e1

pub fn open(

• update all unit tests to use the new signature of seal and open
• update ipa_info_serialize test to validate that binary format of Info includes trigger byte
• update arbitrary_info_corruption test to validate that decryption fails if source event is camouflaged as trigger

@martinthomson the Info struct now represents a single event context because of this trigger bit. I don't think it is a big deal because we can make it cheaply cloneable, so we can follow up on that. @bmcase wants to get his hands dirty with Rust, so this may be a good issue for him to get familiar with our code base.

[martinthomson]

Surely we can reuse Info for all events that share a site, epoch, keyid, and source/trigger bit.

[bmcase]

Created a PR to be reviewed, @akoshelev #598

[akoshelev]

fixed in #598