refactor: Integrate extracted and published near-account-id

[ruseinov]

This PR supersedes #10001

Some context: #10001 (comment)

Part of: near/near-account-id-rs#5

[frol]

@ruseinov Do NOT mark this PR ready for review until near-account-id 1.0 is released. This PR is not going to be merged with a git dependency.

[ruseinov]

@ruseinov Do NOT mark this PR ready for review until near-account-id 1.0 is released. This PR is not going to be merged with a git dependency.

Sounds good!

[frol]

@ruseinov I see that CI still fails, please, take a closer look (I see that some of the recently merged PRs also had some CI checks failed, but fewer than in this PR)

[ruseinov]

@ruseinov I see that CI still fails, please, take a closer look (I see that some of the recently merged PRs also had some CI checks failed, but fewer than in this PR)

Yep, some of those are confusing, looking into it.

[ruseinov]

@frol I've made some fixes, but the workflows need to be approved every single time it seems :/ The upgradable stuff that fails seems to do so, because it's flaky. Sometimes it is green, sometimes it is not.

[frol]

The upgradable stuff that fails seems to do so, because it's flaky. Sometimes it is green, sometimes it is not.

I didn't see "upgradable" succeeding recently for this PR. I re-run it locally:

nearcore/.buildkite/pipeline.yml

Lines 68 to 73 in 7c0b58f

	- label: "upgradable"
	command: |
	source ~/.cargo/env && set -eux
	cd pytest
	pip3 install --user -r requirements.txt
	python3 tests/sanity/upgradable.py

And I get the following error in /tmp/near/upgradable/test1_finished/stderr and /tmp/near/upgradable/test2_finished/stderr (these are 1.35.0 release nodes and this test checks if protocol upgrade will be handled fine once transferred to the node that is built from this PR):

thread 'actix-rt|system:0|arbiter:2' panicked at chain/client/src/client.rs:1694:17:
Can't clear old data, Err(IOErr(Custom { kind: InvalidData, error: "Not all bytes read" }))
stack backtrace:
   0: rust_begin_unwind
             at /rustc/cc66ad468955717ab92600c770da8c1601a4ff33/library/std/src/panicking.rs:595:5
   1: core::panicking::panic_fmt
             at /rustc/cc66ad468955717ab92600c770da8c1601a4ff33/library/core/src/panicking.rs:67:14
   2: near_client::client::Client::on_block_accepted_with_optional_chunk_produce
             at /Users/frol/projects/near/nearcore/chain/client/src/client.rs:1694:17
   3: near_client::client::Client::postprocess_ready_blocks
             at /Users/frol/projects/near/nearcore/chain/client/src/client.rs:1418:13
   4: near_client::client_actor::ClientActor::try_process_unfinished_blocks
             at /Users/frol/projects/near/nearcore/chain/client/src/client_actor.rs:1210:13
   5: <near_client::client_actor::ClientActor as actix::handler::Handler<near_o11y::context::WithSpanContext<near_client::client_actor::ApplyChunksDoneMessage>>>::handle::{{closure}}
             at /Users/frol/projects/near/nearcore/chain/client/src/client_actor.rs:811:9
   6: near_performance_metrics::stats_disabled::measure_performance
             at /Users/frol/projects/near/nearcore/utils/near-performance-metrics/src/stats_disabled.rs:11:5
   7: <near_client::client_actor::ClientActor as actix::handler::Handler<near_o11y::context::WithSpanContext<near_client::client_actor::ApplyChunksDoneMessage>>>::handle
             at /Users/frol/projects/near/nearcore/chain/client/src/client_actor.rs:804:5
   8: <actix::address::envelope::SyncEnvelopeProxy<M> as actix::address::envelope::EnvelopeProxy<A>>::handle
             at /Users/frol/.cargo/registry/src/index.crates.io-6f17d22bba15001f/actix-0.13.0/src/address/envelope.rs:80:23
   9: <actix::address::envelope::Envelope<A> as actix::address::envelope::EnvelopeProxy<A>>::handle

That is borsh deserialization error that happens when there is more data than the struct needs to be fully deserialized. I cannot think of what can be wrong here, but it seems that the root cause is that the node from this PR serializes data that cannot be deserialized by 1.35.0 release. I still fail to see what could have gone wrong here.

[frol]

Hmmm, I tried running the test from master branch and I got exactly the same error, but it passed on CI somehow while CI fails for this PR.

[ruseinov]

Hmmm, I tried running the test from master branch and I got exactly the same error, but it passed on CI somehow while CI fails for this PR.

Will check it out.

[frol]

@ruseinov It seems it was either a glitch or it has been fixed by some other PR merged to master somehow, but "upgradable" CI passes now.

[frol]

The change looks good to me, and it seems that the test that was blocking us is reproducible on master branch as well (the most recent commit to master failed with the same error as this PR used to, and this PR passes that test now with no changes made to address [or cause] that issue).

In order to avoid merge conflicts, I suggest we merge this change sooner rather later, so let's use 1.0.0-alpha.1 version, and once we update near-sdk-rs with it, I will just 1.0.0 release and submit a small PR bumping the version.

[frol]

@nikurt @nagisa Could you review and approve this PR or suggest someone who can do that? I believe this PR is ready to be merged.

[nikurt]

@nikurt @nagisa Could you review and approve this PR or suggest someone who can do that? I believe this PR is ready to be merged.

The author should click the re-request button to re-request a review.

While we're at it, let me do more review begging. Please take a look at near/NEPs#514

[nikurt]

Changes:

• Replace near-account-id with an external crate
• Simplify calls to AccountIdRef::as_ref()
• Simplify parsing of AccountId

All of them look good and improve readability.

[nikurt]

Checked that the crate exists.
https://docs.rs/crate/near-account-id/1.0.0-alpha.1