test: Enhance Iceberg V3 table tests with additional scenarios and assertions

Author: Joe-Abraham

presto-iceberg/src/test/java/com/facebook/presto/iceberg/TestIcebergV3.java

@@ -122,10 +122,14 @@ public void testInsertIntoV3Table()
     @Test
     public void testDeleteOnV3TableNotSupported()
     {
-        String tableName = "test_delete_v3_table";
+        String tableName = "test_v3_delete";
         try {
-            assertUpdate("CREATE TABLE " + tableName + " (id integer, value varchar) WITH (\"format-version\" = '3', \"write.delete.mode\" = 'merge-on-read')");
-            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'one'), (2, 'two')", 2);
+            assertUpdate("CREATE TABLE " + tableName
+                    + " (id INTEGER, name VARCHAR, value DOUBLE) WITH (\"format-version\" = '3', \"write.delete.mode\" = 'merge-on-read')");
+            assertUpdate("INSERT INTO " + tableName
+                    + " VALUES (1, 'Alice', 100.0), (2, 'Bob', 200.0), (3, 'Charlie', 300.0)", 3);
+            assertQuery("SELECT * FROM " + tableName + " ORDER BY id",
+                    "VALUES (1, 'Alice', 100.0), (2, 'Bob', 200.0), (3, 'Charlie', 300.0)");
             assertThatThrownBy(() -> getQueryRunner().execute("DELETE FROM " + tableName + " WHERE id = 1"))
                     .hasMessageContaining("Iceberg table updates for format version 3 are not supported yet");
         }
@@ -137,11 +141,17 @@ public void testDeleteOnV3TableNotSupported()
     @Test
     public void testUpdateOnV3TableNotSupported()
     {
-        String tableName = "test_update_v3_table";
+        String tableName = "test_v3_update";
         try {
-            assertUpdate("CREATE TABLE " + tableName + " (id integer, value varchar) WITH (\"format-version\" = '3', \"write.update.mode\" = 'merge-on-read')");
-            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'one'), (2, 'two')", 2);
-            assertThatThrownBy(() -> getQueryRunner().execute("UPDATE " + tableName + " SET value = 'updated' WHERE id = 1"))
+            assertUpdate("CREATE TABLE " + tableName
+                    + " (id INTEGER, name VARCHAR, status VARCHAR, score DOUBLE) WITH (\"format-version\" = '3', \"write.update.mode\" = 'merge-on-read')");
+            assertUpdate("INSERT INTO " + tableName
+                            + " VALUES (1, 'Alice', 'active', 85.5), (2, 'Bob', 'active', 92.0), (3, 'Charlie', 'inactive', 78.3)",
+                    3);
+            assertQuery("SELECT * FROM " + tableName + " ORDER BY id",
+                    "VALUES (1, 'Alice', 'active', 85.5), (2, 'Bob', 'active', 92.0), (3, 'Charlie', 'inactive', 78.3)");
+            assertThatThrownBy(() -> getQueryRunner()
+                    .execute("UPDATE " + tableName + " SET status = 'updated', score = 95.0 WHERE id = 1"))
                     .hasMessageContaining("Iceberg table updates for format version 3 are not supported yet");
         }
         finally {
@@ -152,15 +162,22 @@ public void testUpdateOnV3TableNotSupported()
     @Test
     public void testMergeOnV3TableNotSupported()
     {
-        String tableName = "test_merge_v3_table";
-        String sourceTable = "test_merge_v3_source";
+        String tableName = "test_v3_merge_target";
+        String sourceTable = "test_v3_merge_source";
         try {
-            assertUpdate("CREATE TABLE " + tableName + " (id integer, value varchar) WITH (\"format-version\" = '3', \"write.update.mode\" = 'merge-on-read')");
-            assertUpdate("CREATE TABLE " + sourceTable + " (id integer, value varchar)");
-            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'one')", 1);
-            assertUpdate("INSERT INTO " + sourceTable + " VALUES (1, 'updated')", 1);
+            assertUpdate("CREATE TABLE " + tableName
+                    + " (id INTEGER, name VARCHAR, value DOUBLE) WITH (\"format-version\" = '3', \"write.update.mode\" = 'merge-on-read')");
+            assertUpdate("CREATE TABLE " + sourceTable + " (id INTEGER, name VARCHAR, value DOUBLE)");
+            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'Alice', 100.0), (2, 'Bob', 200.0)", 2);
+            assertUpdate("INSERT INTO " + sourceTable + " VALUES (1, 'Alice Updated', 150.0), (3, 'Charlie', 300.0)",
+                    2);
+            assertQuery("SELECT * FROM " + tableName + " ORDER BY id", "VALUES (1, 'Alice', 100.0), (2, 'Bob', 200.0)");
+            assertQuery("SELECT * FROM " + sourceTable + " ORDER BY id",
+                    "VALUES (1, 'Alice Updated', 150.0), (3, 'Charlie', 300.0)");
             assertThatThrownBy(() -> getQueryRunner().execute(
-                    "MERGE INTO " + tableName + " t USING " + sourceTable + " s ON t.id = s.id WHEN MATCHED THEN UPDATE SET value = s.value"))
+                    "MERGE INTO " + tableName + " t USING " + sourceTable + " s ON t.id = s.id " +
+                            "WHEN MATCHED THEN UPDATE SET name = s.name, value = s.value " +
+                            "WHEN NOT MATCHED THEN INSERT (id, name, value) VALUES (s.id, s.name, s.value)"))
                     .hasMessageContaining("Iceberg table updates for format version 3 are not supported yet");
         }
         finally {
@@ -172,30 +189,82 @@ public void testMergeOnV3TableNotSupported()
     @Test
     public void testOptimizeOnV3TableNotSupported()
     {
-        String tableName = "test_optimize_v3_table";
+        String tableName = "test_v3_optimize";
         try {
-            assertUpdate("CREATE TABLE " + tableName + " (id integer, value varchar) WITH (\"format-version\" = '3')");
-            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'one')", 1);
-            assertUpdate("INSERT INTO " + tableName + " VALUES (2, 'two')", 1);
-            assertThatThrownBy(() -> getQueryRunner().execute(format("CALL system.rewrite_data_files('%s', '%s')", TEST_SCHEMA, tableName)))
+            assertUpdate("CREATE TABLE " + tableName
+                    + " (id INTEGER, category VARCHAR, value DOUBLE) WITH (\"format-version\" = '3')");
+            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'A', 100.0)", 1);
+            assertUpdate("INSERT INTO " + tableName + " VALUES (2, 'B', 200.0)", 1);
+            assertUpdate("INSERT INTO " + tableName + " VALUES (3, 'A', 150.0)", 1);
+            assertUpdate("INSERT INTO " + tableName + " VALUES (4, 'C', 300.0)", 1);
+            assertQuery("SELECT * FROM " + tableName + " ORDER BY id",
+                    "VALUES (1, 'A', 100.0), (2, 'B', 200.0), (3, 'A', 150.0), (4, 'C', 300.0)");
+
+            assertThatThrownBy(() -> getQueryRunner()
+                    .execute(format("CALL system.rewrite_data_files('%s', '%s')", TEST_SCHEMA, tableName)))
                     .hasMessageContaining("OPTIMIZE is not supported for Iceberg table format version > 2");
         }
         finally {
             dropTable(tableName);
         }
     }
 
+    @Test
+    public void testV3SupportedOperations()
+    {
+        String tableName = "test_v3_supported";
+        try {
+            assertUpdate("CREATE TABLE " + tableName
+                    + " (id INTEGER, name VARCHAR, created_date DATE, amount DECIMAL(10,2)) WITH (\"format-version\" = '3', partitioning = ARRAY['created_date'])");
+
+            assertUpdate("INSERT INTO " + tableName + " VALUES " +
+                    "(1, 'Transaction A', DATE '2024-01-01', 100.50), " +
+                    "(2, 'Transaction B', DATE '2024-01-02', 250.75), " +
+                    "(3, 'Transaction C', DATE '2024-01-01', 175.00)", 3);
+
+            assertQuery("SELECT * FROM " + tableName + " ORDER BY id",
+                    "VALUES " +
+                            "(1, 'Transaction A', DATE '2024-01-01', 100.50), " +
+                            "(2, 'Transaction B', DATE '2024-01-02', 250.75), " +
+                            "(3, 'Transaction C', DATE '2024-01-01', 175.00)");
+
+            assertQuery(
+                    "SELECT created_date, count(*), sum(amount) FROM " + tableName
+                            + " GROUP BY created_date ORDER BY created_date",
+                    "VALUES " +
+                            "(DATE '2024-01-01', 2, 275.50), " +
+                            "(DATE '2024-01-02', 1, 250.75)");
+
+            assertQuery("SELECT * FROM " + tableName + " WHERE created_date = DATE '2024-01-01' ORDER BY id",
+                    "VALUES " +
+                            "(1, 'Transaction A', DATE '2024-01-01', 100.50), " +
+                            "(3, 'Transaction C', DATE '2024-01-01', 175.00)");
+
+            assertUpdate("INSERT INTO " + tableName + " VALUES (4, 'Transaction D', DATE '2024-01-03', 300.00)", 1);
+
+            assertQuery("SELECT count(*) as total_count FROM " + tableName, "SELECT 4");
+        }
+        finally {
+            dropTable(tableName);
+        }
+    }
+
     @Test
     public void testSelectFromV3TableAfterInsert()
     {
         String tableName = "test_select_v3_table";
         try {
-            assertUpdate("CREATE TABLE " + tableName + " (id integer, name varchar, price decimal(10,2)) WITH (\"format-version\" = '3')");
-            assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'apple', 1.50), (2, 'banana', 0.75), (3, 'cherry', 2.00)", 3);
-            assertQuery("SELECT * FROM " + tableName + " ORDER BY id", "VALUES (1, 'apple', 1.50), (2, 'banana', 0.75), (3, 'cherry', 2.00)");
+            assertUpdate("CREATE TABLE " + tableName
+                    + " (id integer, name varchar, price decimal(10,2)) WITH (\"format-version\" = '3')");
+            assertUpdate(
+                    "INSERT INTO " + tableName + " VALUES (1, 'apple', 1.50), (2, 'banana', 0.75), (3, 'cherry', 2.00)",
+                    3);
+            assertQuery("SELECT * FROM " + tableName + " ORDER BY id",
+                    "VALUES (1, 'apple', 1.50), (2, 'banana', 0.75), (3, 'cherry', 2.00)");
             assertQuery("SELECT count(*) FROM " + tableName, "SELECT 3");
             assertQuery("SELECT sum(price) FROM " + tableName, "SELECT 4.25");
-            assertQuery("SELECT name FROM " + tableName + " WHERE price > 1.00 ORDER BY name", "VALUES ('apple'), ('cherry')");
+            assertQuery("SELECT name FROM " + tableName + " WHERE price > 1.00 ORDER BY name",
+                    "VALUES ('apple'), ('cherry')");
         }
         finally {
             dropTable(tableName);
@@ -210,8 +279,10 @@ public void testV3TableWithPartitioning()
             assertUpdate("CREATE TABLE " + tableName + " (id integer, category varchar, value integer) " +
                     "WITH (\"format-version\" = '3', partitioning = ARRAY['category'])");
             assertUpdate("INSERT INTO " + tableName + " VALUES (1, 'A', 100), (2, 'B', 200), (3, 'A', 150)", 3);
-            assertQuery("SELECT * FROM " + tableName + " WHERE category = 'A' ORDER BY id", "VALUES (1, 'A', 100), (3, 'A', 150)");
-            assertQuery("SELECT category, sum(value) FROM " + tableName + " GROUP BY category ORDER BY category", "VALUES ('A', 250), ('B', 200)");
+            assertQuery("SELECT * FROM " + tableName + " WHERE category = 'A' ORDER BY id",
+                    "VALUES (1, 'A', 100), (3, 'A', 150)");
+            assertQuery("SELECT category, sum(value) FROM " + tableName + " GROUP BY category ORDER BY category",
+                    "VALUES ('A', 250), ('B', 200)");
         }
         finally {
             dropTable(tableName);
@@ -220,7 +291,8 @@ public void testV3TableWithPartitioning()
 
     private Table loadTable(String tableName)
     {
-        Catalog catalog = CatalogUtil.loadCatalog(HadoopCatalog.class.getName(), ICEBERG_CATALOG, getProperties(), new Configuration());
+        Catalog catalog = CatalogUtil.loadCatalog(HadoopCatalog.class.getName(), ICEBERG_CATALOG, getProperties(),
+                new Configuration());
         return catalog.loadTable(TableIdentifier.of(TEST_SCHEMA, tableName));
     }
 
@@ -233,7 +305,8 @@ private Map<String, String> getProperties()
     private File getCatalogDirectory()
     {
         Path dataDirectory = getDistributedQueryRunner().getCoordinator().getDataDirectory();
-        Path catalogDirectory = getIcebergDataDirectoryPath(dataDirectory, HADOOP.name(), new IcebergConfig().getFileFormat(), false);
+        Path catalogDirectory = getIcebergDataDirectoryPath(dataDirectory, HADOOP.name(),
+                new IcebergConfig().getFileFormat(), false);
         return catalogDirectory.toFile();
     }
 }