Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Migrate all zcash_client_sqlite pool tests to zcash_client_backend and make them generic over backend implementation #1533

Closed
wants to merge 10 commits into from

Conversation

willemolding
Copy link
Contributor

Moves the tests across, modifying where required and makes some additional changes to the data_api trait methods to support the test use-cases

@willemolding willemolding marked this pull request as draft September 10, 2024 21:38
/// The number of ephemeral addresses that can be safely reserved without observing any
/// of them to be mined. This is the same as the gap limit in Bitcoin.
#[cfg(feature = "transparent-inputs")]
pub(crate) const GAP_LIMIT: u32 = 20;
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Does this belong elsewhere maybe?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes; it can't live under zcash_client_backend::data_api_testing as that's only accessible with the test-dependencies feature flag.

);
// previously it was hard-coded to a SqlClientError which could check the index
// not sure how best to handle this in a generic way
// Err(SqliteClientError::ReachedGapLimit(acct, bad_index))
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Any thoughts on how to handle checking for specific errors like this in the back-end agnostic case?

@willemolding willemolding marked this pull request as ready for review September 11, 2024 18:31

/// Return the notes with a given value
#[cfg(any(test, feature = "test-dependencies"))]
fn get_notes(
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In #1539 I'm moving these to a dedicated WalletTest trait so we can freely expose whatever internals are needed for testing purposes. (I'll handle any conflicts.)

Copy link
Contributor

@str4d str4d left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The individual commits are structured as if they should be preserved, but they are currently unreviewable. Please fix the commits in the PR to be reviewable (which may amount to squashing all the commits together and then optionally splitting the resulting commit into functional parts).

@@ -3194,6 +3194,7 @@ pub mod testing {

let results = stmt
.query_and_then::<TransactionSummary<AccountId>, SqliteClientError, _, _>([], |row| {
let raw_tx: Vec<u8> = row.get("raw")?;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This will break (either loudly, or silently with an empty Vec which is a bad failure mode), as the raw column is nullable.

Suggested change
let raw_tx: Vec<u8> = row.get("raw")?;
let raw_tx: Option<Vec<u8>> = row.get("raw")?;

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

And indeed this was reverted in a later commit as broken. I cannot tell from the current commits why this was initially considered necessary, and why it is no longer needed.

@@ -570,1247 +570,127 @@ pub(crate) fn change_note_spends_succeed<T: ShieldedPoolTester>() {
)
}

pub(crate) fn external_address_change_spends_detected_in_restore_from_seed<
Copy link
Contributor

@str4d str4d Sep 12, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please fix this third commit (ad30ad6):

  • Commit message is remove the raw field on get_tx_history as it broke tests but the raw field is not removed.
  • A bunch of test content is deleted, with no visible move.

In order to review this PR, I need the code moves to be in the same commit, so that git show COMMIT --color-moved=zebra --color-moved-ws=allow-indentation-change works. Ideally the commit is as much a pure move as possible, with a fixup commit afterwards, but if the fixups are trivial then doing both at once is okayish.

@@ -100,7 +100,6 @@ pub struct TransactionSummary<AccountId> {
memo_count: usize,
expired_unmined: bool,
is_shielding: bool,
raw: Vec<u8>,
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The raw removal is in the fourth commit (bb4dd30) instead of the third, but the fix is probably to rewrite the fourth commit's message (and separately fixup or remove the third commit per my previous comment).

// Now scan the block in which we received the note that was spent.
st.scan_cached_blocks(received_height, 1);

// Account balance should be the same.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This commit (bb4dd30) appears to be uncommenting a large amount of test code? I cannot figure out whether any of the uncommented code is part of the fix issue querying notes that the commit message says is going on.

Given that there are thousands of lines changed as a result of doing so, please uncomment the code in one commit, and then in a second commit make changes to it.

@@ -238,3 +240,2302 @@ pub fn send_single_step_proposed_transfer<T: ShieldedPoolTester>(
Ok(_)
);
}

// #[cfg(feature = "transparent-inputs")]
// pub(crate) fn send_multi_step_proposed_transfer<T: ShieldedPoolTester>() {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Aha. Is what you did copy across all of the code, and comment it out at the destination in the same commit? Please do not do that, as the current approach makes it impossible to review the move.

Instead, either copy all the code as-is in one move (creating a commit that fails to compile) and then in subsequent commits apply the fixes, or move individual test functions over in one commit at a time.

Copy link
Contributor

@nuttycom nuttycom left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Flushing a couple of comments from the review I started before @str4d's; I will re-review once the commit history is cleaned up to make it easier to see the moves.

.block_height(),
h
);
assert_eq!(self.get_spendable_balance(account_id, 1), value);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This assertion being embedded here is a bit surprising, because a caller could reasonably expect to be able to call this in locations other than just at the start of a test.

@@ -238,3 +240,2302 @@ pub fn send_single_step_proposed_transfer<T: ShieldedPoolTester>(
Ok(_)
);
}

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

As I'm reviewing this commit-by-commit, I'm surprised to see this copied in commented out here. I presume that it will be uncommented in a later commit, but it would b easier to evaluate the diff if the move were done in a single commit.

OvkPolicy::Sender,
&frobbed_proposal,
);
// TODO: Figure out how we want to handle different kinds of errors
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This makes it clear that some of our error infrastructure needs to be refactored such that the persistence-specific errors are separate from the domain errors; right now SqliteClientError mixes the two. This is a somewhat invasive change though. It might make sense to just assert that there is an error for now, with the goal of later being able to uncomment this code once that refactoring has been performed; these TODOs can serve as motivation to get that done - maybe upgrade them to FIXMEs.

@str4d
Copy link
Contributor

str4d commented Sep 20, 2024

Replaced by #1543.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants