WIP: Init With Root #6441
WIP: Init With Root #6441
Conversation
| "code": 6009, | ||
| "name": "ProofIndexOutOfBounds", | ||
| "msg": "Proof index of is out of bounds" | ||
| }, | ||
| { | ||
| "code": 6010, | ||
| "name": "InvalidProofBuffer", | ||
| "msg": "Invalid proof buffer" |
There was a problem hiding this comment.
Looks like IDL and SDK need to be regenerated after latest changes
| export const changeLogEventBeet = beet.dataEnum<ChangeLogEventRecord>([ | ||
| [ | ||
| 'V1', | ||
| new beet.FixableBeetArgsStruct<ChangeLogEventRecord['V1']>( | ||
| [['fields', beet.tuple([changeLogEventV1Beet])]], | ||
| 'ChangeLogEventRecord["V1"]', | ||
| ), | ||
| ], | ||
| ]) as beet.FixableBeet<ChangeLogEvent>; | ||
| [ | ||
| 'V1', | ||
| new beet.FixableBeetArgsStruct<ChangeLogEventRecord['V1']>( | ||
| [['fields', beet.tuple([changeLogEventV1Beet])]], | ||
| 'ChangeLogEventRecord["V1"]' | ||
| ), | ||
| ], | ||
| ]) as beet.FixableBeet<ChangeLogEvent, ChangeLogEvent> |
There was a problem hiding this comment.
Nit: Has the Solita autogenerated code been run through a different linter? It seems so because afaik Solita generates with commas and semicolons. I would recommend regenerating the Solita TS and not linting it, in order to keep it consistent with other generated Solita code and minimize changed files for the next PR. (I do agree the 2-space indent belongs, and I think Solita does that by default).
There was a problem hiding this comment.
they asked me to remove the lint
There was a problem hiding this comment.
Yeah got it, I think that @ngundotra just didn't want to see all the lint changes as part of this PR, but @StanChe and I were looking into how to actually get the format that is checked in.
It looks like by default the prettierrc.yaml file is not used, so you have to do something like this:
pnpm solita
pnpm lint:fix
pnpm prettier --config prettierrc.yaml --write idl/spl_account_compression.json
I still see some slight differences in the output and had some error messages in the output, but it did work.
@StanChe might have gotten cleaner results but wasn't sure of his local setup.
Having said all this, I think the bigger issue is the missing canopy issue described below in @StanChe 's comment. This could be a blocker for the PR and I think @StanChe is looking into a solution to initialize the canopy on a tree that uses one.
| crate::id(), | ||
| AccountCompressionError::IncorrectAccountOwner | ||
| ); | ||
| let mut merkle_tree_bytes: std::cell::RefMut<'_, &mut [u8]> = |
There was a problem hiding this comment.
Why is the type added for merkle_tree_bytes? Couldn't it just be let mut merkle_tree_bytes = ctx.accounts.merkle_tree..."
| /// Context for initializing a new SPL ConcurrentMerkleTree | ||
| #[derive(Accounts)] | ||
| pub struct InitializeWithRoot<'info> { | ||
| #[account(zero)] | ||
| /// CHECK: This account will be zeroed out, and the size will be validated | ||
| pub merkle_tree: UncheckedAccount<'info>, | ||
|
|
||
| /// Authority that controls write-access to the tree | ||
| /// Typically a program, e.g., the Bubblegum contract validates that leaves are valid NFTs. | ||
| pub authority: Signer<'info>, | ||
|
|
||
| /// Program used to emit changelogs as cpi instruction data. | ||
| pub noop: Program<'info, Noop>, | ||
| } |
There was a problem hiding this comment.
Can we just use Initialize as its a duplicate struct?
| max_depth: u32, | ||
| max_buffer_size: u32, | ||
| root: [u8; 32], | ||
| leaf: [u8; 32], // the first leaf |
There was a problem hiding this comment.
Nit: comment or param name should be rightmost_leaf imo as "first leaf" would only be rightmost if it was empty?
+1 to undoing the lint. In my review I suggested linting in a way that matches Solita output as well (I think this linting differs slightly from Solita generated output, but I could be mistaken) |
|
Unfortunately this change doesn't take into account canopy. If the tree account is created with space for canopy, and the init with root method is called, the proofs provided to modify the assets will be shorter and will only represent the proof under the canopy. Only the last leaf will have the canopy cached properly, while the rest of the canopy will be considered empty. Thus the proofs under any other then the last canopy node will be invalid. I could take over from here and provide an option to init the canopy before the root is provided and also verify the canopy is properly initialized (has all the nodes to the left of the last leaf) before allowing the initialization. |
This pr serves as another attempt to get init with root in. This PR is wip as tests are still being written. But would like to have buy in on the approach.
The approach is as follows, you can init a tree with a root and the proof of the first filled leaf, leaf at index 0. For huuuuuuuuuge trees there is a buffer mechanism to allow you to fill the canopy with more proofs than would fit in a transaction and then you can init.
Im hoping minimize code changes and work within the existing framework, ensure compatibility with indexers.