voonhous commented on code in PR #19405: URL: https://github.com/apache/hudi/pull/19405#discussion_r3690243997
########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/TestHoodieCatalystExpressionUtils.scala: ########## @@ -0,0 +1,94 @@ +/* + * 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. + */ + +package org.apache.hudi + +import org.apache.spark.sql.catalyst.expressions.{Add, AttributeReference, BitwiseOr, Cast, DateAdd, DateSub, Divide, Exp, Expression, Literal, Log, Lower, Multiply, ShiftLeft, Sqrt, Upper} +import org.apache.spark.sql.types.{DateType, DoubleType, IntegerType, LongType, StringType} +import org.junit.jupiter.api.Assertions.assertEquals +import org.junit.jupiter.api.Test + +/** + * Branch coverage for the order-preserving transformation matcher in + * [[org.apache.spark.sql.BaseHoodieCatalystExpressionUtils]]. This is what lets data skipping map a + * transformed column reference back to its source attribute, so each case is pinned to the exact + * source [[AttributeReference]] it must recover, and non-order-preserving shapes must not match. + */ +class TestCatalystExpressionOrderPreserving extends SparkAdapterSupport { + + private val intAttr = AttributeReference("i", IntegerType)() + private val strAttr = AttributeReference("s", StringType)() + private val dblAttr = AttributeReference("d", DoubleType)() + private val dateAttr = AttributeReference("dt", DateType)() + + private def matched(expr: Expression): Option[AttributeReference] = + sparkAdapter.getCatalystExpressionUtils.tryMatchAttributeOrderingPreservingTransformation(expr) + + @Test + def testIdentityAttributeMatches(): Unit = { + assertEquals(Some(intAttr), matched(intAttr)) + } + + @Test + def testArithmeticTransformationsPreserveOrdering(): Unit = { + assertEquals(Some(intAttr), matched(Add(intAttr, Literal(1)))) + assertEquals(Some(intAttr), matched(Add(Literal(1), intAttr))) + assertEquals(Some(intAttr), matched(Multiply(intAttr, Literal(2)))) + assertEquals(Some(intAttr), matched(Multiply(Literal(2), intAttr))) + assertEquals(Some(intAttr), matched(Divide(intAttr, Literal(2)))) + assertEquals(Some(intAttr), matched(BitwiseOr(intAttr, Literal(1)))) + assertEquals(Some(intAttr), matched(BitwiseOr(Literal(1), intAttr))) + assertEquals(Some(intAttr), matched(ShiftLeft(intAttr, Literal(1)))) + } + + @Test + def testUnaryMathAndStringTransformationsPreserveOrdering(): Unit = { + assertEquals(Some(dblAttr), matched(Exp(dblAttr))) + assertEquals(Some(dblAttr), matched(Log(dblAttr))) + assertEquals(Some(strAttr), matched(Upper(strAttr))) + assertEquals(Some(strAttr), matched(Lower(strAttr))) + } + + @Test + def testDateTransformationsPreserveOrdering(): Unit = { + assertEquals(Some(dateAttr), matched(DateAdd(dateAttr, Literal(1)))) + assertEquals(Some(dateAttr), matched(DateSub(dateAttr, Literal(1)))) + } + + @Test + def testUpCastPreservesOrderingButNumericToStringDoesNot(): Unit = { + // Widening a numeric column preserves ordering, so the source attribute is recovered. + assertEquals(Some(intAttr), matched(Cast(intAttr, LongType))) + // Casting a numeric column to string can reorder values, so it must not match. + assertEquals(None, matched(Cast(intAttr, StringType))) Review Comment: Pinned `Cast(long -> int)` as-is with a TODO pointing at #19445. Fixing `isCastPreservingOrdering` would pull `DataSkippingUtils` changes into a test-only PR, so the fix rides with the issue. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/TestHoodieCatalystExpressionUtils.scala: ########## @@ -0,0 +1,94 @@ +/* + * 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. + */ + +package org.apache.hudi + +import org.apache.spark.sql.catalyst.expressions.{Add, AttributeReference, BitwiseOr, Cast, DateAdd, DateSub, Divide, Exp, Expression, Literal, Log, Lower, Multiply, ShiftLeft, Sqrt, Upper} +import org.apache.spark.sql.types.{DateType, DoubleType, IntegerType, LongType, StringType} +import org.junit.jupiter.api.Assertions.assertEquals +import org.junit.jupiter.api.Test + +/** + * Branch coverage for the order-preserving transformation matcher in + * [[org.apache.spark.sql.BaseHoodieCatalystExpressionUtils]]. This is what lets data skipping map a + * transformed column reference back to its source attribute, so each case is pinned to the exact + * source [[AttributeReference]] it must recover, and non-order-preserving shapes must not match. + */ +class TestCatalystExpressionOrderPreserving extends SparkAdapterSupport { + + private val intAttr = AttributeReference("i", IntegerType)() + private val strAttr = AttributeReference("s", StringType)() + private val dblAttr = AttributeReference("d", DoubleType)() + private val dateAttr = AttributeReference("dt", DateType)() + + private def matched(expr: Expression): Option[AttributeReference] = + sparkAdapter.getCatalystExpressionUtils.tryMatchAttributeOrderingPreservingTransformation(expr) + + @Test + def testIdentityAttributeMatches(): Unit = { + assertEquals(Some(intAttr), matched(intAttr)) + } + + @Test + def testArithmeticTransformationsPreserveOrdering(): Unit = { + assertEquals(Some(intAttr), matched(Add(intAttr, Literal(1)))) + assertEquals(Some(intAttr), matched(Add(Literal(1), intAttr))) + assertEquals(Some(intAttr), matched(Multiply(intAttr, Literal(2)))) + assertEquals(Some(intAttr), matched(Multiply(Literal(2), intAttr))) + assertEquals(Some(intAttr), matched(Divide(intAttr, Literal(2)))) + assertEquals(Some(intAttr), matched(BitwiseOr(intAttr, Literal(1)))) + assertEquals(Some(intAttr), matched(BitwiseOr(Literal(1), intAttr))) + assertEquals(Some(intAttr), matched(ShiftLeft(intAttr, Literal(1)))) + } + + @Test + def testUnaryMathAndStringTransformationsPreserveOrdering(): Unit = { + assertEquals(Some(dblAttr), matched(Exp(dblAttr))) + assertEquals(Some(dblAttr), matched(Log(dblAttr))) + assertEquals(Some(strAttr), matched(Upper(strAttr))) + assertEquals(Some(strAttr), matched(Lower(strAttr))) + } + + @Test + def testDateTransformationsPreserveOrdering(): Unit = { Review Comment: Added a `ParseToDate` case. Went with the 1-arg auxiliary constructor instead of the adapter round-trip; `javap` shows it stable on 3.3.1 through 4.1.1 and it keeps `functions` out of the test. Reachability question filed as #19446. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/TestCatalystExpressionOrderPreserving.scala: ########## @@ -0,0 +1,94 @@ +/* + * 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. + */ + +package org.apache.hudi + +import org.apache.spark.sql.catalyst.expressions.{Add, AttributeReference, BitwiseOr, Cast, DateAdd, DateSub, Divide, Exp, Expression, Literal, Log, Lower, Multiply, ShiftLeft, Sqrt, Upper} +import org.apache.spark.sql.types.{DateType, DoubleType, IntegerType, LongType, StringType} +import org.junit.jupiter.api.Assertions.assertEquals +import org.junit.jupiter.api.Test + +/** + * Branch coverage for the order-preserving transformation matcher in + * [[org.apache.spark.sql.BaseHoodieCatalystExpressionUtils]]. This is what lets data skipping map a + * transformed column reference back to its source attribute, so each case is pinned to the exact + * source [[AttributeReference]] it must recover, and non-order-preserving shapes must not match. + */ +class TestCatalystExpressionOrderPreserving extends SparkAdapterSupport { Review Comment: Renamed to `TestHoodieCatalystExpressionUtils`. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/TestFileFormatUtilsForFileGroupReader.scala: ########## @@ -0,0 +1,96 @@ +/* + * 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. + */ + +package org.apache.spark.sql + +import org.apache.spark.sql.catalyst.expressions.{And, Attribute, AttributeReference, Contains, EndsWith, EqualNullSafe, EqualTo, Expression, GreaterThan, In, IsNotNull, IsNull, LessThanOrEqual, Literal, Not, Or, StartsWith} +import org.apache.spark.sql.catalyst.plans.logical.{Filter, LocalRelation} +import org.apache.spark.sql.types.{BooleanType, IntegerType, StringType, StructType} +import org.junit.jupiter.api.Assertions.{assertEquals, assertSame} +import org.junit.jupiter.api.Test + +/** + * Coverage for [[FileFormatUtilsForFileGroupReader.applyFiltersToPlan]], which lowers pushed-down + * data-source [[org.apache.spark.sql.sources.Filter]]s into a Catalyst [[Filter]] on the plan. Each + * case pins the exact translated Catalyst expression so a wrong mapping would fail, and the + * no-filter case must return the input plan untouched. + */ +class TestFileFormatUtilsForFileGroupReader { Review Comment: Description updated. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/TestFileFormatUtilsForFileGroupReader.scala: ########## @@ -0,0 +1,96 @@ +/* + * 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. + */ + +package org.apache.spark.sql + +import org.apache.spark.sql.catalyst.expressions.{And, Attribute, AttributeReference, Contains, EndsWith, EqualNullSafe, EqualTo, Expression, GreaterThan, In, IsNotNull, IsNull, LessThanOrEqual, Literal, Not, Or, StartsWith} +import org.apache.spark.sql.catalyst.plans.logical.{Filter, LocalRelation} +import org.apache.spark.sql.types.{BooleanType, IntegerType, StringType, StructType} +import org.junit.jupiter.api.Assertions.{assertEquals, assertSame} +import org.junit.jupiter.api.Test + +/** + * Coverage for [[FileFormatUtilsForFileGroupReader.applyFiltersToPlan]], which lowers pushed-down + * data-source [[org.apache.spark.sql.sources.Filter]]s into a Catalyst [[Filter]] on the plan. Each + * case pins the exact translated Catalyst expression so a wrong mapping would fail, and the + * no-filter case must return the input plan untouched. + */ +class TestFileFormatUtilsForFileGroupReader { + + private val a: Attribute = AttributeReference("a", IntegerType)() + private val b: Attribute = AttributeReference("b", StringType)() + private val tableSchema: StructType = new StructType().add("a", IntegerType).add("b", StringType) + private val resolved: Seq[Attribute] = Seq(a, b) + private val plan: LocalRelation = LocalRelation(a, b) + + private def condOf(filters: Seq[sources.Filter]): Expression = + FileFormatUtilsForFileGroupReader.applyFiltersToPlan(plan, tableSchema, resolved, filters) match { + case Filter(cond, child) => + assertSame(plan, child) + cond + case other => throw new AssertionError(s"expected a Filter, got $other") + } + + @Test + def testComparisonAndNullFilters(): Unit = { Review Comment: Added `GreaterThanOrEqual`, `LessThan`, a top-level `sources.And`, and the unknown-column `NoSuchElementException` pin. Left the bare `.get` alone; a named error is a production change. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/TestFileFormatUtilsForFileGroupReader.scala: ########## @@ -0,0 +1,96 @@ +/* + * 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. + */ + +package org.apache.spark.sql + +import org.apache.spark.sql.catalyst.expressions.{And, Attribute, AttributeReference, Contains, EndsWith, EqualNullSafe, EqualTo, Expression, GreaterThan, In, IsNotNull, IsNull, LessThanOrEqual, Literal, Not, Or, StartsWith} +import org.apache.spark.sql.catalyst.plans.logical.{Filter, LocalRelation} +import org.apache.spark.sql.types.{BooleanType, IntegerType, StringType, StructType} +import org.junit.jupiter.api.Assertions.{assertEquals, assertSame} +import org.junit.jupiter.api.Test + +/** + * Coverage for [[FileFormatUtilsForFileGroupReader.applyFiltersToPlan]], which lowers pushed-down + * data-source [[org.apache.spark.sql.sources.Filter]]s into a Catalyst [[Filter]] on the plan. Each + * case pins the exact translated Catalyst expression so a wrong mapping would fail, and the + * no-filter case must return the input plan untouched. + */ +class TestFileFormatUtilsForFileGroupReader { + + private val a: Attribute = AttributeReference("a", IntegerType)() + private val b: Attribute = AttributeReference("b", StringType)() + private val tableSchema: StructType = new StructType().add("a", IntegerType).add("b", StringType) + private val resolved: Seq[Attribute] = Seq(a, b) + private val plan: LocalRelation = LocalRelation(a, b) + + private def condOf(filters: Seq[sources.Filter]): Expression = + FileFormatUtilsForFileGroupReader.applyFiltersToPlan(plan, tableSchema, resolved, filters) match { + case Filter(cond, child) => + assertSame(plan, child) + cond + case other => throw new AssertionError(s"expected a Filter, got $other") + } + + @Test + def testComparisonAndNullFilters(): Unit = { + assertEquals(EqualTo(a, Literal(1)), condOf(Seq(sources.EqualTo("a", 1)))) + assertEquals(EqualNullSafe(a, Literal(1)), condOf(Seq(sources.EqualNullSafe("a", 1)))) + assertEquals(GreaterThan(a, Literal(5)), condOf(Seq(sources.GreaterThan("a", 5)))) + assertEquals(LessThanOrEqual(a, Literal(5)), condOf(Seq(sources.LessThanOrEqual("a", 5)))) + assertEquals(IsNull(b), condOf(Seq(sources.IsNull("b")))) + assertEquals(IsNotNull(b), condOf(Seq(sources.IsNotNull("b")))) + } + + @Test + def testStringAndInFilters(): Unit = { + assertEquals(Contains(b, Literal("x")), condOf(Seq(sources.StringContains("b", "x")))) + assertEquals(StartsWith(b, Literal("x")), condOf(Seq(sources.StringStartsWith("b", "x")))) + assertEquals(EndsWith(b, Literal("x")), condOf(Seq(sources.StringEndsWith("b", "x")))) + assertEquals( + In(a, Seq(Literal(1), Literal(2), Literal(3))), + condOf(Seq(sources.In("a", Array[Any](1, 2, 3))))) + } + + @Test + def testConstantFilters(): Unit = { + assertEquals(Literal(true, BooleanType), condOf(Seq(sources.AlwaysTrue()))) + assertEquals(Literal(false, BooleanType), condOf(Seq(sources.AlwaysFalse()))) + } + + @Test + def testCompositeFilters(): Unit = { + // A nested and/or/not tree is lowered structurally. + assertEquals( + Or(EqualTo(a, Literal(1)), Not(IsNull(b))), + condOf(Seq(sources.Or(sources.EqualTo("a", 1), sources.Not(sources.IsNull("b")))))) + } + + @Test + def testMultipleFiltersAreAndedInOrder(): Unit = { + // Several top-level filters combine left-to-right via And. + assertEquals( + And(GreaterThan(a, Literal(5)), IsNotNull(b)), + condOf(Seq(sources.GreaterThan("a", 5), sources.IsNotNull("b")))) Review Comment: Applied, using the incremental-read shape. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/avro/TestAvroUtils.scala: ########## @@ -0,0 +1,171 @@ +/* + * 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. + */ + +package org.apache.spark.sql.avro + +import org.apache.avro.Schema +import org.apache.spark.sql.types._ +import org.junit.jupiter.api.Assertions.{assertEquals, assertFalse, assertThrows, assertTrue} +import org.junit.jupiter.api.Test + +/** + * Direct coverage for [[AvroUtils]] and its [[AvroUtils.AvroSchemaHelper]], which are otherwise only + * exercised indirectly through the Avro serializers. Focuses on the type-support predicate and the + * schema-matching/validation error paths (extra Catalyst fields, extra required Avro fields, + * positional vs by-name matching, ambiguous by-name lookup), pinning the raised exception messages. + */ +class TestAvroUtils { + + private def parse(json: String): Schema = new Schema.Parser().parse(json) + + @Test + def testSupportsDataType(): Unit = { + assertTrue(AvroUtils.supportsDataType(IntegerType)) + assertTrue(AvroUtils.supportsDataType(StringType)) + assertTrue(AvroUtils.supportsDataType(NullType)) + assertTrue(AvroUtils.supportsDataType(ArrayType(LongType))) + assertTrue(AvroUtils.supportsDataType(MapType(StringType, IntegerType))) + assertTrue(AvroUtils.supportsDataType( + new StructType().add("a", IntegerType).add("b", ArrayType(StringType)))) + // CalendarInterval is not representable in Avro, so every wrapper around it is unsupported too. + assertFalse(AvroUtils.supportsDataType(CalendarIntervalType)) + assertFalse(AvroUtils.supportsDataType(ArrayType(CalendarIntervalType))) + assertFalse(AvroUtils.supportsDataType(MapType(StringType, CalendarIntervalType))) + assertFalse(AvroUtils.supportsDataType(new StructType().add("a", CalendarIntervalType))) + } + + @Test + def testToFieldStr(): Unit = { + assertEquals("top-level record", AvroUtils.toFieldStr(Seq.empty)) + assertEquals("field 'foo'", AvroUtils.toFieldStr(Seq("foo"))) + assertEquals("field 'foo.bar'", AvroUtils.toFieldStr(Seq("foo", "bar"))) + } + + @Test + def testIsNullable(): Unit = { + val record = parse( + """{"type":"record","name":"r","fields":[ + | {"name":"req","type":"int"}, + | {"name":"opt","type":["null","int"]}, + | {"name":"uni","type":["int","long"]} + |]}""".stripMargin) + assertFalse(AvroUtils.isNullable(record.getField("req"))) + assertTrue(AvroUtils.isNullable(record.getField("opt"))) + // A union without a NULL branch is not nullable. + assertFalse(AvroUtils.isNullable(record.getField("uni"))) + } + + @Test + def testSchemaHelperRejectsNonRecordSchema(): Unit = { + val ex = assertThrows(classOf[IncompatibleSchemaException], + () => new AvroUtils.AvroSchemaHelper( + Schema.create(Schema.Type.INT), new StructType(), Seq.empty, Seq.empty, false)) + assertTrue(ex.getMessage.contains("as a RECORD")) + } + + @Test + def testMatchedFieldsAndGetAvroFieldByName(): Unit = { + val avro = parse( + """{"type":"record","name":"r","fields":[ + | {"name":"id","type":"int"}, + | {"name":"name","type":["null","string"]} + |]}""".stripMargin) Review Comment: Applied. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
