Implement methods needed for replication defined in EntryStore traits
#102
Conversation
aquadoggo/src/db/stores/entry.rs
Outdated
| @@ -15,6 +15,20 @@ use p2panda_rs::storage_provider::traits::{AsStorageEntry, EntryStore}; | |||
| use crate::db::models::entry::EntryRow; | |||
| use crate::db::provider::SqlStorage; | |||
|
|
|||
| // Re-write once https://github.com/pietgeursen/lipmaa-link/pull/3 merged in lipma-link | |||
There was a problem hiding this comment.
This is merged and released in 0.2.1 🎉
There was a problem hiding this comment.
And also, yeah it's annoying the fact that technically lipmaa links can go all the way back to 0 but the bamboo spec starts link numbers at 1.
🤔
There was a problem hiding this comment.
Yeh, I did a bit of a double take on that 😂
There was a problem hiding this comment.
I think I might add this to lipmaa link:
/// Get all the lipmaa link numbers back to some value
///
/// ```
/// use lipmaa_link::get_lipmaa_links_back_to;
///
/// let links = get_lipmaa_links_back_to(7, 1);
///
/// assert_eq!(links, vec![6, 5, 4, 1]);
///
/// ```
#[cfg(feature = "std")]
pub fn get_lipmaa_links_back_to(start: u64, lower_inclusive: u64) -> Vec<u64> {
let mut path = Vec::new();
let mut n = start;
while n > lower_inclusive {
n = lipmaa(n);
path.push(n);
}
path
}There was a problem hiding this comment.
Ok yeah, I think it's useful, published in 0.2.2
There was a problem hiding this comment.
Nice! Thanks, I'll use that for sure.
There was a problem hiding this comment.
Oh yeah, I remember some one-off bug we had some time ago 😅
|
Are you being blocked by this not being merged @pietgeursen ? Ideally I'd give it today for the others to review, but seeing as we're going into |
|
If I have more time tonight I'll just rebase on top of your changes.
So, no, not blocked 🙂
…On Tue, 10 May 2022, 5:29 pm Sam Andreae, ***@***.***> wrote:
Are you being blocked by this not being merged @pietgeursen
<https://github.com/pietgeursen> ? Ideally I'd give it today for the
others to review, but seeing as we're going into development anyway I
think it would be ok to push ahead if that helps.
—
Reply to this email directly, view it on GitHub
<#102 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACQRI6XK2QEICNX3Z4IIUI3VJHX3RANCNFSM5VNNEEUA>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
| pub struct StorageEntry { | ||
| entry_signed: EntrySigned, | ||
| operation_encoded: OperationEncoded, | ||
| } |
There was a problem hiding this comment.
I removed this previously because technically we don't need two structs (EntryRow and StorageEntry) to handle entries coming in and out of the db because the storage traits could be implemented directly on EntryRow. However, I decided to reintroduce it as I think it improves understanding of what each struct is for and we do need different structs in most other stores, so it matches and will be recognisable across the repo.
There was a problem hiding this comment.
Related issue here: p2panda/p2panda#320
- update p2panda-rs - restructure modules - introduce new replication methods
sandreae
force-pushed
the
entry-store-updates
branch
from
a1c54e4
to
d784469
Compare
aquadoggo/src/db/models/mod.rs
Outdated
| pub mod entry; | ||
| pub mod log; |
There was a problem hiding this comment.
I quite liked the flattened export as long as these modules only contain one or a handful of exports.
| WHERE | ||
| author = $1 | ||
| AND log_id = $2 | ||
| AND CAST(seq_num AS INTEGER) BETWEEN $3 and $4 |
There was a problem hiding this comment.
While SQLite cannot store u64 it can work with them, but only if they are cast as NUMERIC I think.
SELECT CAST("18446744073709551610" as INTEGER) BETWEEN 18446744073709551609 AND 18446744073709551611; => 0
SELECT CAST("18446744073709551610" as NUMERIC) BETWEEN 18446744073709551609 AND 18446744073709551611; => 1
| async fn entry_at_seq_num( | ||
| /// Get an entry from storage by it's hash id. | ||
| /// | ||
| /// Returns a result containing the entry wrapped in an option if it was |
There was a problem hiding this comment.
a result containing the entry wrapped in an option
I like having docstrings that explain the possible return values. I think writing out the full type like that is a bit verbose though.
📋 Checklist
CHANGELOG.md