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

Jump to bottom

feat: adopt kernel schemas and improve protocol support #1756

Merged
merged 17 commits into from
Nov 6, 2023
Merged

feat: adopt kernel schemas and improve protocol support #1756

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
17 commits
Select commit Hold shift + click to select a range
f99158b
chore: copy over kernel types
roeap Oct 22, 2023
de4d980
feat!: integrate with kernel schema types
roeap Oct 24, 2023
defe3d2
fix: serde issues and tests
roeap Oct 24, 2023
6f5eb86
chore: clippy fix
roeap Oct 24, 2023
8c39016
chore: remove old schema types
roeap Oct 24, 2023
22a4947
chore: migrate python code
roeap Oct 24, 2023
681ac56
fix: schema arrow conversion
roeap Oct 24, 2023
8f36c88
fix: docstrings ad feature gates
roeap Oct 24, 2023
7e21881
fix: parquet2 build
roeap Oct 25, 2023
2268cac
Merge branch 'main' into schema
roeap Oct 25, 2023
b658fc0
fix: create primitive type from python
roeap Oct 25, 2023
5458089
Merge branch 'main' into schema
roeap Nov 5, 2023
0066299
fix: post merge fixes and consolidation with existing changes
roeap Nov 5, 2023
bf6282c
fix: parquet2 imports
roeap Nov 5, 2023
9aa301e
fix: pr comments
roeap Nov 5, 2023
12fa820
fix: python field tests
roeap Nov 5, 2023
997fac1
Merge branch 'main' into schema
roeap Nov 6, 2023
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
4 changes: 4 additions & 0 deletions crates/deltalake-core/Cargo.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,6 +68,8 @@ tokio = { workspace = true, features = [
# other deps (these should be organized and pulled into workspace.dependencies as necessary)
cfg-if = "1"
errno = "0.3"
either = "1.8"
fix-hidden-lifetime-bug = "0.2"
hyper = { version = "0.14", optional = true }
itertools = "0.11"
lazy_static = "1"
Expand All @@ -80,8 +82,10 @@ once_cell = "1.16.0"
parking_lot = "0.12"
parquet2 = { version = "0.17", optional = true }
percent-encoding = "2"
roaring = "0.10.1"
tracing = { version = "0.1", optional = true }
rand = "0.8"
z85 = "3.0.5"

# hdfs
datafusion-objectstore-hdfs = { version = "0.1.3", default-features = false, features = [
Expand Down
58 changes: 24 additions & 34 deletions crates/deltalake-core/src/delta_datafusion/expr.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -338,14 +338,13 @@ impl<'a> fmt::Display for ScalarValueFormat<'a> {

#[cfg(test)]
mod test {
use std::collections::HashMap;

use arrow_schema::DataType;
use arrow_schema::DataType as ArrowDataType;
use datafusion::prelude::SessionContext;
use datafusion_common::{DFSchema, ScalarValue};
use datafusion_expr::{col, decode, lit, substring, Cast, Expr, ExprSchemable};

use crate::{DeltaOps, DeltaTable, Schema, SchemaDataType, SchemaField};
use crate::kernel::{DataType, PrimitiveType, StructField, StructType};
use crate::{DeltaOps, DeltaTable};

use super::fmt_expr_to_sql;

Expand All @@ -366,66 +365,57 @@ mod test {
}

async fn setup_table() -> DeltaTable {
let schema = Schema::new(vec![
SchemaField::new(
let schema = StructType::new(vec![
StructField::new(
"id".to_string(),
SchemaDataType::primitive("string".to_string()),
DataType::Primitive(PrimitiveType::String),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"value".to_string(),
SchemaDataType::primitive("integer".to_string()),
DataType::Primitive(PrimitiveType::Integer),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"value2".to_string(),
SchemaDataType::primitive("integer".to_string()),
DataType::Primitive(PrimitiveType::Integer),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"modified".to_string(),
SchemaDataType::primitive("string".to_string()),
DataType::Primitive(PrimitiveType::String),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"active".to_string(),
SchemaDataType::primitive("boolean".to_string()),
DataType::Primitive(PrimitiveType::Boolean),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"money".to_string(),
SchemaDataType::primitive("decimal(12,2)".to_string()),
DataType::Primitive(PrimitiveType::Decimal(12, 2)),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"_date".to_string(),
SchemaDataType::primitive("date".to_string()),
DataType::Primitive(PrimitiveType::Date),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"_timestamp".to_string(),
SchemaDataType::primitive("timestamp".to_string()),
DataType::Primitive(PrimitiveType::Timestamp),
true,
HashMap::new(),
),
SchemaField::new(
StructField::new(
"_binary".to_string(),
SchemaDataType::primitive("binary".to_string()),
DataType::Primitive(PrimitiveType::Binary),
true,
HashMap::new(),
),
]);

let table = DeltaOps::new_in_memory()
.create()
.with_columns(schema.get_fields().clone())
.with_columns(schema.fields().clone())
.await
.unwrap();
assert_eq!(table.version(), 0);
Expand All @@ -441,7 +431,7 @@ mod test {
simple!(
Expr::Cast(Cast {
expr: Box::new(lit(1_i64)),
data_type: DataType::Int32
data_type: ArrowDataType::Int32
}),
"arrow_cast(1, 'Int32')".to_string()
),
Expand Down
37 changes: 19 additions & 18 deletions crates/deltalake-core/src/delta_datafusion/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,11 +70,12 @@ use serde::{Deserialize, Serialize};
use url::Url;

use crate::errors::{DeltaResult, DeltaTableError};
use crate::protocol::{self, Add};
use crate::kernel::{Add, DataType as DeltaDataType, Invariant, PrimitiveType};
use crate::protocol::{self};
use crate::storage::ObjectStoreRef;
use crate::table::builder::ensure_table_uri;
use crate::table::state::DeltaTableState;
use crate::{open_table, open_table_with_storage_options, DeltaTable, Invariant, SchemaDataType};
use crate::{open_table, open_table_with_storage_options, DeltaTable};

const PATH_COLUMN: &str = "__delta_rs_path";

Expand Down Expand Up @@ -121,7 +122,7 @@ impl DeltaTableState {
min_value: None,
distinct_count: None
};
self.schema().unwrap().get_fields().len()
self.schema().unwrap().fields().len()
]),
is_exact: true,
},
Expand All @@ -139,13 +140,13 @@ impl DeltaTableState {
column_statistics: acc.column_statistics.map(|col_stats| {
self.schema()
.unwrap()
.get_fields()
.fields()
.iter()
.zip(col_stats)
.map(|(field, stats)| {
let null_count = new_stats
.null_count
.get(field.get_name())
.get(field.name())
.and_then(|x| {
let null_count_acc = stats.null_count?;
let null_count = x.as_value()? as usize;
Expand All @@ -155,7 +156,7 @@ impl DeltaTableState {

let max_value = new_stats
.max_values
.get(field.get_name())
.get(field.name())
.and_then(|x| {
let old_stats = stats.clone();
let max_value = to_scalar_value(x.as_value()?);
Expand All @@ -179,7 +180,7 @@ impl DeltaTableState {

let min_value = new_stats
.min_values
.get(field.get_name())
.get(field.name())
.and_then(|x| {
let old_stats = stats.clone();
let min_value = to_scalar_value(x.as_value()?);
Expand Down Expand Up @@ -222,15 +223,15 @@ impl DeltaTableState {
num_rows: stats.num_rows,
total_byte_size: stats.total_byte_size,
column_statistics: stats.column_statistics.map(|col_stats| {
let fields = self.schema().unwrap().get_fields();
let fields = self.schema().unwrap().fields();
col_stats
.iter()
.zip(fields)
.map(|(col_states, field)| {
let dt = self
.arrow_schema()
.unwrap()
.field_with_name(field.get_name())
.field_with_name(field.name())
.unwrap()
.data_type()
.clone();
Expand Down Expand Up @@ -258,16 +259,14 @@ fn get_prune_stats(table: &DeltaTable, column: &Column, get_max: bool) -> Option
let field = table
.get_schema()
.ok()
.map(|s| s.get_field_with_name(&column.name).ok())??;
.map(|s| s.field_with_name(&column.name).ok())??;

// See issue 1214. Binary type does not support natural order which is required for Datafusion to prune
if let SchemaDataType::primitive(t) = &field.get_type() {
if t == "binary" {
return None;
}
if let DeltaDataType::Primitive(PrimitiveType::Binary) = &field.data_type() {
return None;
}

let data_type = field.get_type().try_into().ok()?;
let data_type = field.data_type().try_into().ok()?;
let partition_columns = &table.get_metadata().ok()?.partition_columns;

let values = table.get_state().files().iter().map(|add| {
Expand Down Expand Up @@ -921,7 +920,7 @@ pub(crate) fn get_null_of_arrow_type(t: &ArrowDataType) -> DeltaResult<ScalarVal
}

pub(crate) fn partitioned_file_from_action(
action: &protocol::Add,
action: &Add,
partition_columns: &[String],
schema: &ArrowSchema,
) -> PartitionedFile {
Expand Down Expand Up @@ -1790,7 +1789,7 @@ mod tests {
let mut partition_values = std::collections::HashMap::new();
partition_values.insert("month".to_string(), Some("1".to_string()));
partition_values.insert("year".to_string(), Some("2015".to_string()));
let action = protocol::Add {
let action = Add {
path: "year=2015/month=1/part-00000-4dcb50d3-d017-450c-9df7-a7257dbd3c5d-c000.snappy.parquet".to_string(),
size: 10644,
partition_values,
Expand All @@ -1801,6 +1800,8 @@ mod tests {
deletion_vector: None,
stats_parsed: None,
tags: None,
base_row_id: None,
default_row_commit_version: None,
};
let schema = ArrowSchema::new(vec![
Field::new("year", ArrowDataType::Int64, true),
Expand Down Expand Up @@ -1953,7 +1954,7 @@ mod tests {

let table = crate::DeltaOps::new_in_memory()
.create()
.with_columns(get_delta_schema().get_fields().clone())
.with_columns(get_delta_schema().fields().clone())
.with_partition_columns(["modified", "id"])
.await
.unwrap();
Expand Down
6 changes: 6 additions & 0 deletions crates/deltalake-core/src/errors.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -205,6 +205,12 @@ pub enum DeltaTableError {
/// Source error
source: Box<dyn std::error::Error + Send + Sync + 'static>,
},

#[error("Kernel: {source}")]
Kernel {
#[from]
source: crate::kernel::Error,
},
}

impl From<object_store::path::Error> for DeltaTableError {
Expand Down
Loading
Loading