Shielding parentchain transfers

[coax1d]

closes #1252
Built ontop of #1355

Adding in initial shielding using Alices account on Parentchain.

Involves refactoring:

• introduced ita_parentchain_interface crate where all listeners are moved to now (extrinsic and event triggering).
• this crate does discriminate ParentchainId for future generalization

testing

run node on host:

./target/release/integritee-node --dev --tmp --unsafe-ws-external --rpc-cors all 

run sidechain

SGX_MODE=SW WORKER_MODE=sidechain WORKER_FEATURES=dcap make
cd bin
export RUST_LOG=info,substrate_api_client=warn,ws=warn,mio=warn,its_consensus_common=info,sidechain=info,integritee_service=trace,enclave_runtime=trace,ac_node_api=warn,sp_io=warn,itc_parentchain_indirect_calls_executor=trace,itp_stf_executor=trace,itc_parentchain_light_client=trace,itc_parentchain_block_importer=trace,itp_stf_state_handler=trace,ita_stf=trace,itp-attestation-handler=debug,itc_parentchain_indirect_calls_executor=trace,itp_top_pool=debug,itc_offchain_worker_executor=debug,ita_parentchain_interface=trace
./integritee-service -c -u ws://172.17.0.1 run --skip-ra --dev &> worker.log 

once producing sidechain blocks, use js/apps to send funds from Bob to Alice

• worker log should show shielding for ....
• and ita_stf::trusted_cal[....] balance_shield([...]

check shielded balance for Bob

./integritee-cli -u ws://172.17.0.1 trusted --mrenclave 2MyXuroF4yiR8A6S8t334C4ZvY6qyKxVBKsd8SPnVW3y balance //Bob

should return the amount you sent
q.e.d.

[clangenb]

I think this goes into the right direction, but the executor mustn't be aware of the business logic specifics. :)

[clangenb]

Sorry for the very late review. :)

Nice, this looks quite good now! According to your judgement, what is missing in this PR?

I will merge master into this branch, as the breaking changes with the api-client might confuse you a bit.

[brenzi]

CI M6 fails because choosing //Alice as the vault account disturbs other tests that use //Alice for other purposes. concretely, demo_shielding_unshielding fails because unshielding to //Alice triggers the new shielding logic (which panics, which it shouldn't, but that's for further investigation)

[brenzi]

these cyclic dependencies are a PITA.

• solvable with untangle dependencies on ita-stf #1483, but that's quite a body of work. looking for shortcuts now

[clangenb]

All in all, locks good. With some minor modifications to trait definitions, I think we can make the code more ergonomic in the codebase. As associated types don't have to be mentioned usually, when a trait is put as a trait bound.