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.

Sign up for GitHub

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

Feat/transaction scan #19

Merged
merged 2 commits into from
Jul 24, 2024
Merged

Feat/transaction scan #19

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7ae0efa
feat: impl `Transaction::scan`
KKould Jul 24, 2024
3812cb4
fix: projection miss _null and _ts
KKould Jul 24, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
192 changes: 159 additions & 33 deletions src/lib.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -208,25 +208,26 @@ where

async fn get<'get>(
&'get self,
version: &'get Version<R, FP>,
key: &'get R::Key,
ts: Timestamp,
projection: Projection,
) -> Result<Option<Entry<'get, R>>, ParquetError> {
let mut scan = self.scan(Bound::Included(key), Bound::Unbounded, ts);
) -> Result<Option<Entry<'get, R>>, WriteError<R>>
where
FP: FileProvider,
{
let mut scan = Scan::new(
self,
(Bound::Included(key), Bound::Unbounded),
ts,
version,
vec![],
);

if let Projection::Parts(projection) = projection {
scan = scan.projection(projection)
if let Projection::Parts(mask) = projection {
scan = scan.projection(mask);
}
scan.take().await?.next().await.transpose()
}

fn scan<'scan>(
&'scan self,
lower: Bound<&'scan R::Key>,
uppwer: Bound<&'scan R::Key>,
ts: Timestamp,
) -> Scan<'scan, R, FP> {
Scan::new(self, lower, uppwer, ts)
Ok(scan.take().await?.next().await.transpose()?)
}

fn check_conflict(&self, key: &R::Key, ts: Timestamp) -> bool {
Expand All @@ -251,9 +252,13 @@ where
{
schema: &'scan Schema<R, FP>,
lower: Bound<&'scan R::Key>,
uppwer: Bound<&'scan R::Key>,
upper: Bound<&'scan R::Key>,
ts: Timestamp,

version: &'scan Version<R, FP>,
streams: Vec<ScanStream<'scan, R, FP>>,

limit: Option<usize>,
projection: ProjectionMask,
}

Expand All @@ -264,25 +269,33 @@ where
{
fn new(
schema: &'scan Schema<R, FP>,
lower: Bound<&'scan R::Key>,
uppwer: Bound<&'scan R::Key>,
(lower, upper): (Bound<&'scan R::Key>, Bound<&'scan R::Key>),
ts: Timestamp,
version: &'scan Version<R, FP>,
streams: Vec<ScanStream<'scan, R, FP>>,
) -> Self {
Self {
schema,
lower,
uppwer,
upper,
ts,
version,
streams,
limit: None,
projection: ProjectionMask::all(),
}
}

pub fn limit(self, limit: Option<usize>) -> Self {
Self { limit, ..self }
}

pub fn projection(self, mut projection: Vec<usize>) -> Self {
// skip two columns: _null and _ts
for p in &mut projection {
*p += 2;
}

projection.extend([0, 1, 2]);
let mask = ProjectionMask::roots(
&arrow_to_parquet_schema(R::arrow_schema()).unwrap(),
projection,
Expand All @@ -295,25 +308,32 @@ where
}

pub async fn take(
self,
) -> Result<impl Stream<Item = Result<Entry<'scan, R>, ParquetError>>, ParquetError> {
let mut streams = Vec::<ScanStream<R, FP>>::with_capacity(self.schema.immutables.len() + 1);
streams.push(
mut self,
) -> Result<impl Stream<Item = Result<Entry<'scan, R>, ParquetError>>, WriteError<R>> {
self.streams.push(
self.schema
.mutable
.scan((self.lower, self.uppwer), self.ts)
.scan((self.lower, self.upper), self.ts)
.into(),
);
for immutable in &self.schema.immutables {
streams.push(
self.streams.push(
immutable
.scan((self.lower, self.uppwer), self.ts, self.projection.clone())
.scan((self.lower, self.upper), self.ts, self.projection.clone())
.into(),
);
}
// TODO: sstable scan

MergeStream::from_vec(streams).await
self.version
.streams(
&mut self.streams,
(self.lower, self.upper),
self.ts,
self.limit,
self.projection,
)
.await?;

Ok(MergeStream::from_vec(self.streams).await?)
}
}

Expand All @@ -326,24 +346,32 @@ where
Io(#[from] io::Error),
#[error("write version error: {0}")]
Version(#[from] VersionError<R>),
#[error("write parquet error: {0}")]
Parquet(#[from] ParquetError),
}

#[cfg(test)]
pub(crate) mod tests {
use std::sync::Arc;
use std::{collections::VecDeque, sync::Arc};

use arrow::{
array::{Array, AsArray, RecordBatch},
datatypes::{DataType, Field, Schema, UInt32Type},
};
use async_lock::RwLock;
use once_cell::sync::Lazy;
use parquet::arrow::ProjectionMask;
use tracing::error;

use crate::{
executor::Executor,
inmem::immutable::tests::TestImmutableArrays,
executor::{tokio::TokioExecutor, Executor},
inmem::{
immutable::{tests::TestImmutableArrays, Immutable},
mutable::Mutable,
},
record::{internal::InternalRecordRef, RecordRef},
DbOption, Record, DB,
version::{cleaner::Cleaner, set::tests::build_version_set, Version},
DbOption, Record, WriteError, DB,
};

#[derive(Debug, PartialEq, Eq)]
Expand Down Expand Up @@ -438,6 +466,7 @@ pub(crate) mod tests {
if !vbool_array.is_null(offset) {
vbool = Some(vbool_array.value(offset));
}
column_i += 1;
}

let record = TestRef {
Expand Down Expand Up @@ -482,4 +511,101 @@ pub(crate) mod tests {

schema.immutables[0].as_record_batch().clone()
}

pub(crate) async fn build_schema() -> crate::Schema<Test, TokioExecutor> {
let mutable = Mutable::new();

mutable.insert(
Test {
vstring: "alice".to_string(),
vu32: 1,
vobool: Some(true),
},
1_u32.into(),
);
mutable.insert(
Test {
vstring: "ben".to_string(),
vu32: 2,
vobool: Some(true),
},
1_u32.into(),
);
mutable.insert(
Test {
vstring: "carl".to_string(),
vu32: 3,
vobool: Some(true),
},
1_u32.into(),
);

let immutables = {
let mutable = Mutable::new();

mutable.insert(
Test {
vstring: "dice".to_string(),
vu32: 4,
vobool: Some(true),
},
1_u32.into(),
);
mutable.insert(
Test {
vstring: "erika".to_string(),
vu32: 5,
vobool: Some(true),
},
1_u32.into(),
);
mutable.insert(
Test {
vstring: "funk".to_string(),
vu32: 6,
vobool: Some(true),
},
1_u32.into(),
);

VecDeque::from(vec![Immutable::from(mutable)])
};

crate::Schema {
mutable,
immutables,
_marker: Default::default(),
}
}

pub(crate) async fn build_db<R, E>(
option: Arc<DbOption>,
executor: E,
schema: crate::Schema<R, E>,
version: Version<R, E>,
) -> Result<DB<R, E>, WriteError<R>>
where
R: Record,
E: Executor,
{
E::create_dir_all(&option.path).await?;

let schema = Arc::new(RwLock::new(schema));

let (mut cleaner, clean_sender) = Cleaner::new(option.clone());
let version_set = build_version_set(version, clean_sender, option.clone()).await?;

executor.spawn(async move {
if let Err(err) = cleaner.listen().await {
error!("[Cleaner Error]: {}", err)
}
});

Ok(DB {
schema,
version_set,
lock_map: Arc::new(Default::default()),
_p: Default::default(),
})
}
}
4 changes: 2 additions & 2 deletions src/record/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
pub(crate) mod internal;
mod str;

use std::{hash::Hash, sync::Arc};
use std::{fmt::Debug, hash::Hash, sync::Arc};

use arrow::{
array::{Datum, RecordBatch},
Expand All @@ -16,7 +16,7 @@ use crate::{
};

pub trait Key: 'static + Encode + Decode + Ord + Clone + Send + Hash + std::fmt::Debug {
type Ref<'r>: KeyRef<'r, Key = Self> + Copy
type Ref<'r>: KeyRef<'r, Key = Self> + Copy + Debug
where
Self: 'r;

Expand Down
2 changes: 1 addition & 1 deletion src/stream/merge.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ use super::{Entry, ScanStream};
use crate::{fs::FileProvider, record::Record};

pin_project! {
pub(crate) struct MergeStream<'merge, R, FP>
pub struct MergeStream<'merge, R, FP>
where
R: Record,
FP: FileProvider,
Expand Down
Loading
Loading