radicle: add cob::cache module

Add the cob::cache module which describes an SQLite store and schema
for caching COBs. It also provides in-memory and no-cache types for
testing purposes.

The intention is for the store to be used alongside the existing COB
types so that a write-thru cache can be implemented and used for
efficient reads.

Signed-off-by: Fintan Halpenny <fintan.halpenny@gmail.com>
X-Clacks-Overhead: GNU Terry Pratchett
This commit is contained in:
Fintan Halpenny 2024-02-08 16:44:18 +00:00 committed by cloudhead
parent f7ef9cc8b3
commit b26842d4d8
No known key found for this signature in database
3 changed files with 285 additions and 0 deletions

View File

@ -1,4 +1,5 @@
#![warn(clippy::unwrap_used)]
pub mod cache;
pub mod common;
pub mod identity;
pub mod issue;

265
radicle/src/cob/cache.rs Normal file
View File

@ -0,0 +1,265 @@
use std::collections::HashMap;
use std::convert::Infallible;
use std::fmt;
use std::marker::PhantomData;
use std::path::Path;
use std::sync::Arc;
use std::time;
use radicle_cob::ObjectId;
use sqlite as sql;
use thiserror::Error;
use crate::prelude::RepoId;
use crate::sql::transaction;
/// File suffix for storing the COBs database file.
pub const COBS_DB_FILE: &str = "cache.db";
/// How long to wait for the database lock to be released before failing a read.
const DB_READ_TIMEOUT: time::Duration = time::Duration::from_secs(3);
/// How long to wait for the database lock to be released before failing a write.
const DB_WRITE_TIMEOUT: time::Duration = time::Duration::from_secs(6);
/// Database migrations.
/// The first migration is the creation of the initial tables.
const MIGRATIONS: &[&str] = &[include_str!("cache/migrations/1.sql")];
#[derive(Error, Debug)]
pub enum Error {
/// An Internal error.
#[error("internal error: {0}")]
Internal(#[from] sql::Error),
/// No rows returned in query result.
#[error("no rows returned")]
NoRows,
}
/// Read-only type witness.
#[derive(Clone)]
pub struct Read;
/// Read-write type witness.
#[derive(Clone)]
pub struct Write;
/// A file-backed database storing information about the network.
#[derive(Clone)]
pub struct Store<T> {
pub(super) db: Arc<sql::ConnectionThreadSafe>,
marker: PhantomData<T>,
}
impl<T> fmt::Debug for Store<T> {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
f.debug_struct("Database").finish()
}
}
impl Store<Read> {
/// Same as [`Self::open`], but in read-only mode. This is useful to have multiple
/// open databases, as no locking is required.
pub fn reader<P: AsRef<Path>>(path: P) -> Result<Self, Error> {
let mut db = sql::Connection::open_thread_safe_with_flags(
path,
sqlite::OpenFlags::new().with_read_only(),
)?;
db.set_busy_timeout(DB_READ_TIMEOUT.as_millis() as usize)?;
Ok(Self {
db: Arc::new(db),
marker: PhantomData,
})
}
/// Create a new in-memory database.
pub fn memory() -> Result<Self, Error> {
let mut db = sql::Connection::open_thread_safe_with_flags(
":memory:",
sqlite::OpenFlags::new().with_read_only(),
)?;
db.set_busy_timeout(DB_READ_TIMEOUT.as_millis() as usize)?;
Ok(Self {
db: Arc::new(db),
marker: PhantomData,
})
}
}
impl Store<Write> {
/// Open a database at the given path. Creates a new database if it
/// doesn't exist.
pub fn open<P: AsRef<Path>>(path: P) -> Result<Self, Error> {
let mut db = sql::Connection::open_thread_safe(path)?;
db.set_busy_timeout(DB_WRITE_TIMEOUT.as_millis() as usize)?;
migrate(&db)?;
Ok(Self {
db: Arc::new(db),
marker: PhantomData,
})
}
/// Create a new in-memory database.
pub fn memory() -> Result<Self, Error> {
let db = Arc::new(sql::Connection::open_thread_safe(":memory:")?);
migrate(&db)?;
Ok(Self {
db,
marker: PhantomData,
})
}
/// Turn this handle into a read-only handle.
pub fn read_only(self) -> Store<Read> {
Store {
db: self.db,
marker: PhantomData,
}
}
/// Perform a raw query on the database handle.
pub fn raw_query<T, E, F>(&self, query: F) -> Result<T, E>
where
F: FnOnce(&sql::Connection) -> Result<T, E>,
E: From<sql::Error>,
{
transaction(&self.db, query)
}
}
impl<T> Store<T> {
/// Get the database version. This is updated on schema changes.
pub fn version(&self) -> Result<usize, Error> {
version(&self.db)
}
}
/// Get the `user_version` value from the database header.
pub fn version(db: &sql::Connection) -> Result<usize, Error> {
let version = db
.prepare("PRAGMA user_version")?
.into_iter()
.next()
.ok_or(Error::NoRows)??
.read::<i64, _>(0);
Ok(version as usize)
}
/// Bump the `user_version` value.
fn bump(db: &sql::Connection) -> Result<usize, Error> {
let old = version(db)?;
let new = old + 1;
db.execute(format!("PRAGMA user_version = {new}"))?;
Ok(new as usize)
}
/// Migrate the database to the latest schema.
fn migrate(db: &sql::Connection) -> Result<usize, Error> {
let mut version = version(db)?;
for (i, migration) in MIGRATIONS.iter().enumerate() {
if i >= version {
transaction(db, |db| {
db.execute(migration)?;
version = bump(db)?;
Ok::<_, Error>(())
})?;
}
}
Ok(version)
}
/// Update a COB object in the cache.
pub trait Update<T> {
/// The output type, if any, for a successful update.
type Out;
type UpdateError: std::error::Error + Send + Sync + 'static;
fn update(
&mut self,
rid: &RepoId,
id: &ObjectId,
object: &T,
) -> Result<Self::Out, Self::UpdateError>;
}
/// Remove a COB object in the cache.
pub trait Remove<T> {
/// The output type, if any, for a successful removal.
type Out;
type RemoveError: std::error::Error + Send + Sync + 'static;
fn remove(&mut self, rid: &RepoId, id: &ObjectId) -> Result<Self::Out, Self::RemoveError>;
}
/// An in-memory cache for storing COB objects.
///
/// The intention is for this to be used in tests that expect cache
/// reads.
#[derive(Clone, Debug)]
pub struct InMemory<T> {
inner: HashMap<RepoId, HashMap<ObjectId, T>>,
}
impl<T> Default for InMemory<T> {
fn default() -> Self {
Self {
inner: HashMap::new(),
}
}
}
impl<T> Update<T> for InMemory<T>
where
T: Clone,
{
type Out = Option<T>;
type UpdateError = Infallible;
fn update(
&mut self,
rid: &RepoId,
id: &ObjectId,
object: &T,
) -> Result<Self::Out, Self::UpdateError> {
let objects = self.inner.entry(*rid).or_default();
Ok(objects.insert(*id, object.clone()))
}
}
/// The `/dev/null` of caches.
///
/// It will ignore any updates, and successfully return on each call
/// of [`Update::update`].
///
/// The intention is for this to be used in tests that do not expect
/// any cache reads.
pub struct NoCache;
impl<T> Update<T> for NoCache {
type Out = ();
type UpdateError = Infallible;
fn update(
&mut self,
_rid: &RepoId,
_id: &ObjectId,
_object: &T,
) -> Result<Self::Out, Self::UpdateError> {
Ok(())
}
}
impl<T> Remove<T> for NoCache {
type Out = ();
type RemoveError = Infallible;
fn remove(&mut self, _rid: &RepoId, _id: &ObjectId) -> Result<Self::Out, Self::RemoveError> {
Ok(())
}
}

19
radicle/src/cob/cache/migrations/1.sql vendored Normal file
View File

@ -0,0 +1,19 @@
-- Issues
create table if not exists "issues" (
-- Issue ID
"id" text primary key not null,
-- Repository ID
"repo" text not null,
-- Issue in JSON format
"issue" text not null
) strict;
-- Patches
create table if not exists "patches" (
-- Patch ID
"id" text primary key not null,
-- Repository ID
"repo" text not null,
-- Patch in JSON format
"patch" text not null
) strict;