delta_kernel 0.21.0

Core crate providing a Delta/Deltalake implementation focused on interoperability with a wide range of query engines.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214

//! End-to-end integration tests for clustered tables.
//!
//! These tests exercise the full lifecycle: create table with clustering columns,
//! write data, commit, checkpoint, and verify the data and clustering metadata
//! are preserved throughout.

use std::collections::HashMap;
use std::sync::Arc;

use delta_kernel::arrow::array::{ArrayRef, Int32Array};
use delta_kernel::committer::FileSystemCommitter;
use delta_kernel::expressions::ColumnName;
use delta_kernel::schema::{DataType, StructField, StructType};
use delta_kernel::snapshot::Snapshot;
use delta_kernel::transaction::create_table::create_table;
use delta_kernel::transaction::data_layout::DataLayout;
use delta_kernel::transaction::CommitResult;
use rstest::rstest;

use test_utils::{
    generate_batch, read_add_infos, read_scan, test_table_setup_mt, write_batch_to_table, IntoArray,
};

/// Full lifecycle: create a clustered table, write data, verify stats include clustering columns,
/// checkpoint, and verify clustering metadata and data survive. When `use_fresh_snapshot` is true,
/// the write happens via a fresh snapshot (simulating a separate session that did not create the
/// table).
#[rstest]
#[case::post_commit_snapshot(false)]
#[case::fresh_snapshot(true)]
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn test_clustered_table_write_and_checkpoint(
    #[case] use_fresh_snapshot: bool,
) -> Result<(), Box<dyn std::error::Error>> {
    let (_temp_dir, table_path, engine) = test_table_setup_mt()?;
    let schema = Arc::new(
        StructType::try_new(vec![
            StructField::new("id", DataType::INTEGER, false),
            StructField::new("name", DataType::STRING, true),
            StructField::new("city", DataType::STRING, true),
        ])
        .unwrap(),
    );
    let expected_clustering = vec![ColumnName::new(["id"]), ColumnName::new(["city"])];

    // Create table clustered on "id" and "city"
    let create_result = create_table(&table_path, schema, "Test/1.0")
        .with_data_layout(DataLayout::Clustered {
            columns: expected_clustering.clone(),
        })
        .build(engine.as_ref(), Box::new(FileSystemCommitter::new()))?
        .commit(engine.as_ref())?;

    let snapshot = if use_fresh_snapshot {
        // Open a fresh snapshot (as if a different process is writing)
        let table_url = delta_kernel::try_parse_uri(&table_path)?;
        Snapshot::builder_for(table_url).build(engine.as_ref())?
    } else {
        match create_result {
            CommitResult::CommittedTransaction(committed) => committed
                .post_commit_snapshot()
                .expect("post-commit snapshot should exist")
                .clone(),
            other => panic!("Expected CommittedTransaction, got: {other:?}"),
        }
    };

    // First write: 3 rows
    let batch = generate_batch(vec![
        ("id", vec![1, 2, 3].into_array()),
        ("name", vec!["alice", "bob", "charlie"].into_array()),
        ("city", vec!["seattle", "portland", "seattle"].into_array()),
    ])?;
    let snapshot = write_batch_to_table(&snapshot, engine.as_ref(), batch, HashMap::new()).await?;
    assert_eq!(snapshot.version(), 1);

    // Second write: 2 more rows
    let batch = generate_batch(vec![
        ("id", vec![4, 5].into_array()),
        ("name", vec!["dave", "eve"].into_array()),
        ("city", vec!["austin", "portland"].into_array()),
    ])?;
    let snapshot = write_batch_to_table(&snapshot, engine.as_ref(), batch, HashMap::new()).await?;
    assert_eq!(snapshot.version(), 2);

    // Verify stats include all clustering columns
    let add_infos = read_add_infos(&snapshot, engine.as_ref())?;
    assert!(!add_infos.is_empty());
    for info in &add_infos {
        let stats = info.stats.as_ref().expect("Add action should have stats");
        for col in &expected_clustering {
            let col_name = col.to_string();
            assert!(
                stats["minValues"].get(&col_name).is_some(),
                "Stats should include minValues for clustering column '{col_name}'"
            );
            assert!(
                stats["maxValues"].get(&col_name).is_some(),
                "Stats should include maxValues for clustering column '{col_name}'"
            );
        }
    }

    // Verify data readable before checkpoint
    let scan = snapshot.clone().scan_builder().build()?;
    let batches = read_scan(&scan, engine.clone())?;
    let total_rows: usize = batches.iter().map(|b| b.num_rows()).sum();
    assert_eq!(total_rows, 5);

    // Checkpoint
    snapshot.checkpoint(engine.as_ref())?;

    // Load fresh snapshot from checkpoint and verify everything survived
    let table_url = delta_kernel::try_parse_uri(&table_path)?;
    let fresh = Snapshot::builder_for(table_url).build(engine.as_ref())?;
    assert_eq!(fresh.version(), 2);
    let scan = fresh.clone().scan_builder().build()?;
    let batches = read_scan(&scan, engine.clone())?;
    let total_rows: usize = batches.iter().map(|b| b.num_rows()).sum();
    assert_eq!(total_rows, 5);

    // Verify stats still include clustering columns after checkpoint
    let add_infos = read_add_infos(&fresh, engine.as_ref())?;
    assert!(!add_infos.is_empty());
    for info in &add_infos {
        let stats = info.stats.as_ref().expect("Add action should have stats");
        for col in &expected_clustering {
            let col_name = col.to_string();
            assert!(
                stats["minValues"].get(&col_name).is_some(),
                "Stats should include minValues for clustering column '{col_name}' after checkpoint"
            );
            assert!(
                stats["maxValues"].get(&col_name).is_some(),
                "Stats should include maxValues for clustering column '{col_name}' after checkpoint"
            );
        }
    }

    Ok(())
}

/// Regression test: writing a batch where a clustering column has ALL null values should succeed.
///
/// `collect_stats` emits null-valued min/max entries for all-null columns, allowing
/// `StatsVerifier` to find the field and confirm `nullCount == numRecords`. The JSON serializer
/// omits null fields on disk, matching Spark's `ignoreNullFields` behavior.
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn test_clustered_table_write_all_null_clustering_column() {
    let (_temp_dir, table_path, engine) = test_table_setup_mt().unwrap();
    let schema = Arc::new(
        StructType::try_new(vec![
            StructField::new("category", DataType::STRING, false),
            StructField::new("region_id", DataType::INTEGER, true),
        ])
        .unwrap(),
    );

    // Create table clustered on "category" and "region_id"
    let create_result = create_table(&table_path, schema, "Test/1.0")
        .with_data_layout(DataLayout::Clustered {
            columns: vec![
                ColumnName::new(["category"]),
                ColumnName::new(["region_id"]),
            ],
        })
        .build(engine.as_ref(), Box::new(FileSystemCommitter::new()))
        .unwrap()
        .commit(engine.as_ref())
        .unwrap();

    let snapshot = create_result
        .unwrap_committed()
        .post_commit_snapshot()
        .expect("post-commit snapshot should exist")
        .clone();

    // Write a batch where region_id is ALL nulls.
    // This should succeed -- all-null clustering columns are valid.
    let all_null_region: ArrayRef = Arc::new(Int32Array::from(vec![None, None, None]));
    let batch = generate_batch(vec![
        ("category", vec!["a", "b", "c"].into_array()),
        ("region_id", all_null_region),
    ])
    .unwrap();

    let snapshot = write_batch_to_table(&snapshot, engine.as_ref(), batch, HashMap::new())
        .await
        .unwrap();
    assert_eq!(snapshot.version(), 1);

    // Verify data is readable
    let scan = snapshot.clone().scan_builder().build().unwrap();
    let batches = read_scan(&scan, engine.clone()).unwrap();
    let total_rows: usize = batches.iter().map(|b| b.num_rows()).sum();
    assert_eq!(total_rows, 3);

    // Verify stats: region_id should have nullCount=3. The JSON serializer omits null
    // fields (matching Spark's ignoreNullFields), so minValues/maxValues should not
    // contain region_id in the serialized JSON stats.
    let add_infos = read_add_infos(&snapshot, engine.as_ref()).unwrap();
    assert_eq!(add_infos.len(), 1);
    let stats = add_infos[0].stats.as_ref().expect("should have stats");
    assert_eq!(stats["numRecords"], 3);
    assert_eq!(stats["nullCount"]["region_id"], 3);
    assert!(
        stats["minValues"].get("region_id").is_none(),
        "JSON minValues should omit region_id when all values are null"
    );
    assert!(
        stats["maxValues"].get("region_id").is_none(),
        "JSON maxValues should omit region_id when all values are null"
    );
}