feat(datafusion): support metadata tables for Datafusion

Signed-off-by: xxchan <[email protected]>

Cargo.toml

@@ -91,6 +91,7 @@ serde_derive = "1"
 serde_json = "1"
 serde_repr = "0.1.16"
 serde_with = "3.4"
+strum = "0.26"
 tempfile = "3.8"
 tokio = { version = "1", default-features = false }
 typed-builder = "0.20"

crates/iceberg/Cargo.toml

@@ -78,15 +78,16 @@ serde_derive = { workspace = true }
 serde_json = { workspace = true }
 serde_repr = { workspace = true }
 serde_with = { workspace = true }
+strum = { workspace = true, features = ["derive"] }
 tokio = { workspace = true, optional = true }
 typed-builder = { workspace = true }
 url = { workspace = true }
 uuid = { workspace = true }
 zstd = { workspace = true }
+expect-test = { workspace = true }
 
 [dev-dependencies]
 ctor = { workspace = true }
-expect-test = { workspace = true }
 iceberg-catalog-memory = { workspace = true }
 iceberg_test_utils = { path = "../test_utils", features = ["tests"] }
 pretty_assertions = { workspace = true }

crates/iceberg/src/lib.rs

@@ -83,6 +83,7 @@ pub mod transform;
 mod runtime;
 
 pub mod arrow;
+pub mod test_utils;
 mod utils;
 pub mod writer;
 

crates/iceberg/src/metadata_scan.rs

@@ -35,9 +35,46 @@ use crate::Result;
 /// - <https://github.com/apache/iceberg/blob/ac865e334e143dfd9e33011d8cf710b46d91f1e5/core/src/main/java/org/apache/iceberg/MetadataTableType.java#L23-L39>
 /// - <https://iceberg.apache.org/docs/latest/spark-queries/#querying-with-sql>
 /// - <https://py.iceberg.apache.org/api/#inspecting-tables>
-#[derive(Debug)]
+#[derive(Debug, Clone)]
 pub struct MetadataTable(Table);
 
+/// Metadata table type.
+#[derive(Debug, Clone, strum::EnumIter)]
+pub enum MetadataTableType {
+    /// [`SnapshotsTable`]
+    Snapshots,
+    /// [`ManifestsTable`]
+    Manifests,
+}
+
+impl MetadataTableType {
+    /// Returns the string representation of the metadata table type.
+    pub fn as_str(&self) -> &str {
+        match self {
+            MetadataTableType::Snapshots => "snapshots",
+            MetadataTableType::Manifests => "manifests",
+        }
+    }
+
+    /// Returns all the metadata table types.
+    pub fn all_types() -> impl Iterator<Item = Self> {
+        use strum::IntoEnumIterator;
+        Self::iter()
+    }
+}
+
+impl TryFrom<&str> for MetadataTableType {
+    type Error = String;
+
+    fn try_from(value: &str) -> std::result::Result<Self, String> {
+        match value {
+            "snapshots" => Ok(Self::Snapshots),
+            "manifests" => Ok(Self::Manifests),
+            _ => Err(format!("invalid metadata table type: {value}")),
+        }
+    }
+}
+
 impl MetadataTable {
     /// Creates a new metadata scan.
     pub(super) fn new(table: Table) -> Self {
@@ -262,53 +299,7 @@ mod tests {
 
     use super::*;
     use crate::scan::tests::TableTestFixture;
-
-    /// Snapshot testing to check the resulting record batch.
-    ///
-    /// - `expected_schema/data`: put `expect![[""]]` as a placeholder,
-    ///   and then run test with `UPDATE_EXPECT=1 cargo test` to automatically update the result,
-    ///   or use rust-analyzer (see [video](https://github.com/rust-analyzer/expect-test)).
-    ///   Check the doc of [`expect_test`] for more details.
-    /// - `ignore_check_columns`: Some columns are not stable, so we can skip them.
-    /// - `sort_column`: The order of the data might be non-deterministic, so we can sort it by a column.
-    fn check_record_batch(
-        record_batch: RecordBatch,
-        expected_schema: Expect,
-        expected_data: Expect,
-        ignore_check_columns: &[&str],
-        sort_column: Option<&str>,
-    ) {
-        let mut columns = record_batch.columns().to_vec();
-        if let Some(sort_column) = sort_column {
-            let column = record_batch.column_by_name(sort_column).unwrap();
-            let indices = arrow_ord::sort::sort_to_indices(column, None, None).unwrap();
-            columns = columns
-                .iter()
-                .map(|column| arrow_select::take::take(column.as_ref(), &indices, None).unwrap())
-                .collect_vec();
-        }
-
-        expected_schema.assert_eq(&format!(
-            "{}",
-            record_batch.schema().fields().iter().format(",\n")
-        ));
-        expected_data.assert_eq(&format!(
-            "{}",
-            record_batch
-                .schema()
-                .fields()
-                .iter()
-                .zip_eq(columns)
-                .map(|(field, column)| {
-                    if ignore_check_columns.contains(&field.name().as_str()) {
-                        format!("{}: (skipped)", field.name())
-                    } else {
-                        format!("{}: {:?}", field.name(), column)
-                    }
-                })
-                .format(",\n")
-        ));
-    }
+    use crate::test_utils::check_record_batch;
 
     #[test]
     fn test_snapshots_table() {

crates/iceberg/src/test_utils.rs

@@ -0,0 +1,71 @@
+// Licensed to the Apache Software Foundation (ASF) under one
+// or more contributor license agreements.  See the NOTICE file
+// distributed with this work for additional information
+// regarding copyright ownership.  The ASF licenses this file
+// to you under the Apache License, Version 2.0 (the
+// "License"); you may not use this file except in compliance
+// with the License.  You may obtain a copy of the License at
+//
+//   http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing,
+// software distributed under the License is distributed on an
+// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, either express or implied.  See the License for the
+// specific language governing permissions and limitations
+// under the License.
+
+//! Test utilities.
+//! This module is pub just for internal testing.
+//! It is subject to change and is not intended to be used by external users.
+
+use arrow_array::RecordBatch;
+use expect_test::Expect;
+use itertools::Itertools;
+
+/// Snapshot testing to check the resulting record batch.
+///
+/// - `expected_schema/data`: put `expect![[""]]` as a placeholder,
+///   and then run test with `UPDATE_EXPECT=1 cargo test` to automatically update the result,
+///   or use rust-analyzer (see [video](https://github.com/rust-analyzer/expect-test)).
+///   Check the doc of [`expect_test`] for more details.
+/// - `ignore_check_columns`: Some columns are not stable, so we can skip them.
+/// - `sort_column`: The order of the data might be non-deterministic, so we can sort it by a column.
+pub fn check_record_batch(
+    record_batch: RecordBatch,
+    expected_schema: Expect,
+    expected_data: Expect,
+    ignore_check_columns: &[&str],
+    sort_column: Option<&str>,
+) {
+    let mut columns = record_batch.columns().to_vec();
+    if let Some(sort_column) = sort_column {
+        let column = record_batch.column_by_name(sort_column).unwrap();
+        let indices = arrow_ord::sort::sort_to_indices(column, None, None).unwrap();
+        columns = columns
+            .iter()
+            .map(|column| arrow_select::take::take(column.as_ref(), &indices, None).unwrap())
+            .collect_vec();
+    }
+
+    expected_schema.assert_eq(&format!(
+        "{}",
+        record_batch.schema().fields().iter().format(",\n")
+    ));
+    expected_data.assert_eq(&format!(
+        "{}",
+        record_batch
+            .schema()
+            .fields()
+            .iter()
+            .zip_eq(columns)
+            .map(|(field, column)| {
+                if ignore_check_columns.contains(&field.name().as_str()) {
+                    format!("{}: (skipped)", field.name())
+                } else {
+                    format!("{}: {:?}", field.name(), column)
+                }
+            })
+            .format(",\n")
+    ));
+}

crates/integrations/datafusion/Cargo.toml

@@ -37,5 +37,6 @@ iceberg = { workspace = true }
 tokio = { workspace = true }
 
 [dev-dependencies]
+expect-test = { workspace = true }
 iceberg-catalog-memory = { workspace = true }
 tempfile = { workspace = true }

crates/integrations/datafusion/src/physical_plan/metadata_scan.rs

@@ -0,0 +1,91 @@
+// Licensed to the Apache Software Foundation (ASF) under one
+// or more contributor license agreements.  See the NOTICE file
+// distributed with this work for additional information
+// regarding copyright ownership.  The ASF licenses this file
+// to you under the Apache License, Version 2.0 (the
+// "License"); you may not use this file except in compliance
+// with the License.  You may obtain a copy of the License at
+//
+//   http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing,
+// software distributed under the License is distributed on an
+// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, either express or implied.  See the License for the
+// specific language governing permissions and limitations
+// under the License.
+
+use datafusion::catalog::TableProvider;
+use datafusion::physical_expr::EquivalenceProperties;
+use datafusion::physical_plan::execution_plan::{Boundedness, EmissionType};
+use datafusion::physical_plan::stream::RecordBatchStreamAdapter;
+use datafusion::physical_plan::{DisplayAs, ExecutionPlan, Partitioning, PlanProperties};
+
+use crate::metadata_table::IcebergMetadataTableProvider;
+
+#[derive(Debug)]
+pub struct IcebergMetadataScan {
+    provider: IcebergMetadataTableProvider,
+    properties: PlanProperties,
+}
+
+impl IcebergMetadataScan {
+    pub fn new(provider: IcebergMetadataTableProvider) -> Self {
+        let properties = PlanProperties::new(
+            EquivalenceProperties::new(provider.schema()),
+            Partitioning::UnknownPartitioning(1),
+            EmissionType::Incremental,
+            Boundedness::Bounded,
+        );
+        Self {
+            provider,
+            properties,
+        }
+    }
+}
+
+impl DisplayAs for IcebergMetadataScan {
+    fn fmt_as(
+        &self,
+        _t: datafusion::physical_plan::DisplayFormatType,
+        f: &mut std::fmt::Formatter,
+    ) -> std::fmt::Result {
+        write!(f, "IcebergMetadataScan")
+    }
+}
+
+impl ExecutionPlan for IcebergMetadataScan {
+    fn name(&self) -> &str {
+        "IcebergMetadataScan"
+    }
+
+    fn as_any(&self) -> &dyn std::any::Any {
+        self
+    }
+
+    fn properties(&self) -> &PlanProperties {
+        &self.properties
+    }
+
+    fn children(&self) -> Vec<&std::sync::Arc<dyn ExecutionPlan>> {
+        vec![]
+    }
+
+    fn with_new_children(
+        self: std::sync::Arc<Self>,
+        _children: Vec<std::sync::Arc<dyn ExecutionPlan>>,
+    ) -> datafusion::error::Result<std::sync::Arc<dyn ExecutionPlan>> {
+        Ok(self)
+    }
+
+    fn execute(
+        &self,
+        _partition: usize,
+        _context: std::sync::Arc<datafusion::execution::TaskContext>,
+    ) -> datafusion::error::Result<datafusion::execution::SendableRecordBatchStream> {
+        let batch_fut = self.provider.clone().scan();
+        let schema = self.provider.schema();
+        let stream = futures::stream::once(batch_fut);
+        Ok(Box::pin(RecordBatchStreamAdapter::new(schema, stream)))
+    }
+}

crates/integrations/datafusion/src/physical_plan/mod.rs

@@ -16,4 +16,5 @@
 // under the License.
 
 pub(crate) mod expr_to_predicate;
+pub(crate) mod metadata_scan;
 pub(crate) mod scan;