Add file pruning test for predicate using primitive fields in ROW column in Delta Lake Connector

plugin/trino-delta-lake/src/test/java/io/trino/plugin/deltalake/TestSplitPruning.java

@@ -47,6 +47,7 @@ public class TestSplitPruning
             "float_nan",
             "float_inf",
             "no_stats",
+            "nested_fields",
             "timestamp",
             "test_partitioning",
             "parquet_struct_statistics",
@@ -358,6 +359,32 @@ public void testParquetStatisticsPruning()
         testCountQuery("SELECT count(*) FROM parquet_struct_statistics WHERE row = ROW(2, 'b')", 3, 9);
     }
 
+    @Test
+    public void testPrimitiveFieldsInsideRowColumnPruning()
+    {
+        assertResultAndSplitCount(
+                "SELECT grandparent.parent1.child1 FROM nested_fields WHERE id > 6",
+                Set.of(70.99, 80.99, 90.99, 100.99),
+                1);
+
+        assertResultAndSplitCount(
+                "SELECT grandparent.parent1.child1 FROM nested_fields WHERE id > 10",
+                Set.of(),
+                0);
+
+        // TODO pruning does not work on primitive fields inside a struct, expected splits should be 1 after file pruning (https://github.com/trinodb/trino/issues/17164)
+        assertResultAndSplitCount(
+                "SELECT grandparent.parent1.child1 FROM nested_fields WHERE parent.child1 > 600",
+                Set.of(70.99, 80.99, 90.99, 100.99),
+                2);
+
+        // TODO pruning does not work on primitive fields inside a struct, expected splits should be 0 after file pruning (https://github.com/trinodb/trino/issues/17164)
+        assertResultAndSplitCount(
+                "SELECT grandparent.parent1.child1 FROM nested_fields WHERE parent.child1 > 1000",
+                Set.of(),
+                2);
+    }
+
     @Test
     public void testJsonStatisticsPruningUppercaseColumn()
     {

plugin/trino-delta-lake/src/test/resources/databricks/pruning/nested_fields/README.md

@@ -0,0 +1,17 @@
+Data generated using Databricks 10.4 LTS:
+
+```sql
+CREATE TABLE nested_fields (id INT, parent STRUCT<child1: BIGINT, child2: STRING>, grandparent STRUCT<parent1 STRUCT<child1: DOUBLE, child2: BOOLEAN>, parent2 STRUCT<child1: DATE>>) USING DELTA LOCATION 's3://starburst-test/nested_fields';
+
+INSERT INTO nested_fields VALUES
+    (1, struct('100', 'INDIA'), struct(struct(10.99, true), struct('2023-01-01'))),
+    (2, struct('200', 'POLAND'), struct(struct(20.99, false), struct('2023-02-01'))),
+    (3, struct('300', 'USA'), struct(struct(30.99, true), struct('2023-03-01'))),
+    (4, struct('400', 'AUSTRIA'), struct(struct(40.99, false), struct('2023-04-01'))),
+    (5, struct('500', 'JAPAN'), struct(struct(50.99, true), struct('2023-05-01'))),
+    (6, struct('600', 'UK'), struct(struct(60.99, false), struct('2023-06-01'))),
+    (7, struct('700', 'FRANCE'), struct(struct(70.99, true), struct('2023-07-01'))),
+    (8, struct('800', 'BRAZIL'), struct(struct(80.99, false), struct('2023-08-01'))),
+    (9, struct('900', 'NAMIBIA'), struct(struct(90.99, true), struct('2023-09-01'))),
+    (10, struct('1000', 'RSA'), struct(struct(100.99, false), struct('2023-10-01')));
+```

plugin/trino-delta-lake/src/test/resources/databricks/pruning/nested_fields/_delta_log/00000000000000000000.json

@@ -0,0 +1,3 @@
+{"commitInfo":{"timestamp":1681993589387,"userId":"7499806201888255","userName":"[email protected]","operation":"CREATE TABLE","operationParameters":{"isManaged":"false","description":null,"partitionBy":"[]","properties":"{}"},"notebook":{"notebookId":"4432842678029071"},"clusterId":"1118-145353-2w37w56t","isolationLevel":"WriteSerializable","isBlindAppend":true,"operationMetrics":{},"engineInfo":"Databricks-Runtime/10.4.x-scala2.12","txnId":"63491529-eec9-485b-9aca-662bac90be67"}}
+{"protocol":{"minReaderVersion":1,"minWriterVersion":2}}
+{"metaData":{"id":"735e2c32-06a1-4c03-a5a8-5d0b92c58f85","format":{"provider":"parquet","options":{}},"schemaString":"{\"type\":\"struct\",\"fields\":[{\"name\":\"id\",\"type\":\"integer\",\"nullable\":true,\"metadata\":{}},{\"name\":\"parent\",\"type\":{\"type\":\"struct\",\"fields\":[{\"name\":\"child1\",\"type\":\"long\",\"nullable\":true,\"metadata\":{}},{\"name\":\"child2\",\"type\":\"string\",\"nullable\":true,\"metadata\":{}}]},\"nullable\":true,\"metadata\":{}},{\"name\":\"grandparent\",\"type\":{\"type\":\"struct\",\"fields\":[{\"name\":\"parent1\",\"type\":{\"type\":\"struct\",\"fields\":[{\"name\":\"child1\",\"type\":\"double\",\"nullable\":true,\"metadata\":{}},{\"name\":\"child2\",\"type\":\"boolean\",\"nullable\":true,\"metadata\":{}}]},\"nullable\":true,\"metadata\":{}},{\"name\":\"parent2\",\"type\":{\"type\":\"struct\",\"fields\":[{\"name\":\"child1\",\"type\":\"date\",\"nullable\":true,\"metadata\":{}}]},\"nullable\":true,\"metadata\":{}}]},\"nullable\":true,\"metadata\":{}}]}","partitionColumns":[],"configuration":{},"createdTime":1681993589155}}

plugin/trino-delta-lake/src/test/resources/databricks/pruning/nested_fields/_delta_log/00000000000000000001.json

@@ -0,0 +1,3 @@
+{"commitInfo":{"timestamp":1681993595946,"userId":"7499806201888255","userName":"[email protected]","operation":"WRITE","operationParameters":{"mode":"Append","partitionBy":"[]"},"notebook":{"notebookId":"4432842678029071"},"clusterId":"1118-145353-2w37w56t","readVersion":0,"isolationLevel":"WriteSerializable","isBlindAppend":true,"operationMetrics":{"numFiles":"2","numOutputRows":"10","numOutputBytes":"4731"},"engineInfo":"Databricks-Runtime/10.4.x-scala2.12","txnId":"cc74cc50-8684-4179-bddc-248ef4350d7c"}}
+{"add":{"path":"part-00000-8e3f6979-c368-4f48-b389-9b4dbf46de93-c000.snappy.parquet","partitionValues":{},"size":2366,"modificationTime":1681993596000,"dataChange":true,"stats":"{\"numRecords\":5,\"minValues\":{\"id\":1,\"parent\":{\"child1\":100,\"child2\":\"AUSTRIA\"},\"grandparent\":{\"parent1\":{\"child1\":10.99},\"parent2\":{\"child1\":\"2023-01-01\"}}},\"maxValues\":{\"id\":5,\"parent\":{\"child1\":500,\"child2\":\"USA\"},\"grandparent\":{\"parent1\":{\"child1\":50.99},\"parent2\":{\"child1\":\"2023-05-01\"}}},\"nullCount\":{\"id\":0,\"parent\":{\"child1\":0,\"child2\":0},\"grandparent\":{\"parent1\":{\"child1\":0,\"child2\":0},\"parent2\":{\"child1\":0}}}}","tags":{"INSERTION_TIME":"1681993596000000","OPTIMIZE_TARGET_SIZE":"268435456"}}}
+{"add":{"path":"part-00001-dd006b0f-93c6-4f87-801f-a8b7153a72ae-c000.snappy.parquet","partitionValues":{},"size":2365,"modificationTime":1681993596000,"dataChange":true,"stats":"{\"numRecords\":5,\"minValues\":{\"id\":6,\"parent\":{\"child1\":600,\"child2\":\"BRAZIL\"},\"grandparent\":{\"parent1\":{\"child1\":60.99},\"parent2\":{\"child1\":\"2023-06-01\"}}},\"maxValues\":{\"id\":10,\"parent\":{\"child1\":1000,\"child2\":\"UK\"},\"grandparent\":{\"parent1\":{\"child1\":100.99},\"parent2\":{\"child1\":\"2023-10-01\"}}},\"nullCount\":{\"id\":0,\"parent\":{\"child1\":0,\"child2\":0},\"grandparent\":{\"parent1\":{\"child1\":0,\"child2\":0},\"parent2\":{\"child1\":0}}}}","tags":{"INSERTION_TIME":"1681993596000001","OPTIMIZE_TARGET_SIZE":"268435456"}}}