Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[HUDI-3523] Introduce AddPrimitiveColumnSchemaPostProcessor to support add new primitive column to the end of a schema #5031

Merged
merged 1 commit into from
Sep 23, 2022
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
@@ -0,0 +1,172 @@
/*
* 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.utilities.schema;

import org.apache.hudi.common.config.ConfigProperty;
import org.apache.hudi.common.config.TypedProperties;
import org.apache.hudi.common.model.HoodieRecord;
import org.apache.hudi.utilities.exception.HoodieSchemaPostProcessException;

import org.apache.avro.LogicalTypes;
import org.apache.avro.Schema;
import org.apache.log4j.LogManager;
import org.apache.log4j.Logger;
import org.apache.spark.api.java.JavaSparkContext;

import java.util.ArrayList;
import java.util.List;
import java.util.Locale;

/**
* A {@link SchemaPostProcessor} use to add column to given schema. Currently. only supports adding one column at a time.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: use -> used

* Users can specify the position of new column by config {@link Config#SCHEMA_POST_PROCESSOR_ADD_COLUMN_NEXT_PROP},
* the new column will be added before this column.
* <p>
* Currently supported types : bytes, string, int, long, float, double, boolean, decimal
*/
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I understand this is schema post processor. But can we add a line here in documentation that i hudi, adding a new column in the middle may not be advisable. users might run into issues. So, its always a good practice to append new columns to the end.

public class AddColumnSchemaPostProcessor extends SchemaPostProcessor {

private static final Logger LOG = LogManager.getLogger(AddColumnSchemaPostProcessor.class);

public AddColumnSchemaPostProcessor(TypedProperties props, JavaSparkContext jssc) {
super(props, jssc);
}

/**
* Configs supported.
*/
public static class Config {
public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_NAME_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.name")
.noDefaultValue()
.withDocumentation("New column's name");

public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_TYPE_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.type")
.noDefaultValue()
.withDocumentation("New column's type");
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can we add the supported types here in documentation as well?


public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_DOC_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.doc")
.noDefaultValue()
.withDocumentation("New column's doc");

public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_DEFAULT_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.default")
.noDefaultValue()
.withDocumentation("New column's default value");

public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_SIZE_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.size")
.noDefaultValue()
.withDocumentation("New column's size, used in decimal type");

public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_PRECISION_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.precision")
.noDefaultValue()
.withDocumentation("New column's precision, used in decimal type");

public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_SCALE_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.scale")
.noDefaultValue()
.withDocumentation("New column's precision, used in decimal type");
wangxianghu marked this conversation as resolved.
Show resolved Hide resolved

public static final ConfigProperty<String> SCHEMA_POST_PROCESSOR_ADD_COLUMN_NEXT_PROP = ConfigProperty
.key("hoodie.deltastreamer.schemaprovider.schema_post_processor.add.column.next")
.defaultValue(HoodieRecord.HOODIE_IS_DELETED)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

there should not be any default. We should just append to the end by defualt

.withDocumentation("Column name which locate next to new column, `_hoodie_is_deleted` by default.");
}

public static final String BYTES = "BYTES";
public static final String STRING = "STRING";
public static final String INT = "INT";
public static final String LONG = "LONG";
public static final String FLOAT = "FLOAT";
public static final String DOUBLE = "DOUBLE";
public static final String BOOLEAN = "BOOLEAN";
public static final String DECIMAL = "DECIMAL";

@Override
public Schema processSchema(Schema schema) {
String newColumnName = this.config.getString(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NAME_PROP.key());

if (schema.getField(newColumnName) != null) {
LOG.warn(String.format("Column %s already exist!", newColumnName));
return schema;
}

List<Schema.Field> sourceFields = schema.getFields();
List<Schema.Field> targetFields = new ArrayList<>(sourceFields.size() + 1);

String nextColumnName = this.config.getString(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NEXT_PROP.key(),
Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NEXT_PROP.defaultValue());

// mark whether the new column is added
boolean isAdded = false;
for (Schema.Field sourceField : sourceFields) {
if (sourceField.name().equals(nextColumnName)) {
targetFields.add(buildNewColumn());
isAdded = true;
}
targetFields.add(new Schema.Field(sourceField.name(), sourceField.schema(), sourceField.doc(), sourceField.defaultVal()));
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

from the documentation, I thought we will add the new column after the column specified in Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NEXT_PROP. But guess, we are adding the new column prior to that. Can we please clarify this in documentation. Or may be give an example.

}

// this would happen when `nextColumn` does not exist. just append the new column to the end
if (!isAdded) {
targetFields.add(buildNewColumn());
}

return Schema.createRecord(schema.getName(), schema.getDoc(), schema.getNamespace(), false, targetFields);
}

private Schema.Field buildNewColumn() {
Schema.Field result;

String columnName = this.config.getString(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NAME_PROP.key());
String type = this.config.getString(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_TYPE_PROP.key()).toUpperCase(Locale.ROOT);
String doc = this.config.getString(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_DOC_PROP.key(), null);
Object defaultValue = this.config.getOrDefault(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_DEFAULT_PROP.key(),
null);

switch (type) {
case STRING:
case BYTES:
case INT:
case LONG:
case FLOAT:
case DOUBLE:
case BOOLEAN:
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

so, looks like we are supporting only primitives here. I don't see arrays or maps etc. Can we call it out in the class documentation. would like to understand why we have such a limitation. why not add all diff types?

result = new Schema.Field(columnName, Schema.create(Schema.Type.valueOf(type)), doc, defaultValue);
break;
case DECIMAL:
int size = this.config.getInteger(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_SIZE_PROP.key(), 10);
int precision = this.config.getInteger(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_PRECISION_PROP.key());
int scale = this.config.getInteger(Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_SCALE_PROP.key());

Schema decimalSchema = Schema.createFixed(null, null, null, size);
LogicalTypes.decimal(precision, scale).addToSchema(decimalSchema);

result = new Schema.Field(columnName, decimalSchema, doc, defaultValue);
break;
default:
throw new HoodieSchemaPostProcessException(String.format("Type %s is not supported", type));
}
return result;
}
}
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@

import org.apache.hudi.common.config.TypedProperties;
import org.apache.hudi.utilities.exception.HoodieSchemaPostProcessException;
import org.apache.hudi.utilities.schema.AddColumnSchemaPostProcessor;
import org.apache.hudi.utilities.schema.DeleteSupportSchemaPostProcessor;
import org.apache.hudi.utilities.schema.DropColumnSchemaPostProcessor;
import org.apache.hudi.utilities.schema.SchemaPostProcessor;
Expand All @@ -33,10 +34,14 @@
import org.apache.avro.Schema.Type;
import org.junit.jupiter.api.Assertions;
import org.junit.jupiter.api.Test;
import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.Arguments;
import org.junit.jupiter.params.provider.MethodSource;

import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import java.util.stream.Stream;

import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertNotNull;
Expand All @@ -55,13 +60,18 @@ public class TestSchemaPostProcessor extends UtilitiesTestBase {
+ "{\"name\":\"_row_key\",\"type\":\"string\"},{\"name\":\"rider\",\"type\":\"string\"},{\"name\":\"driver\","
+ "\"type\":\"string\"},{\"name\":\"fare\",\"type\":\"double\"}]}";

private static Stream<Arguments> configParams() {
String[] types = {"bytes", "string", "int", "long", "float", "double", "boolean"};
return Stream.of(types).map(Arguments::of);
}

@Test
public void testPostProcessor() throws IOException {
properties.put(Config.SCHEMA_POST_PROCESSOR_PROP, DummySchemaPostProcessor.class.getName());
SchemaProvider provider =
UtilHelpers.wrapSchemaProviderWithPostProcessor(
UtilHelpers.createSchemaProvider(DummySchemaProvider.class.getName(), properties, jsc),
properties, jsc,null);
UtilHelpers.createSchemaProvider(DummySchemaProvider.class.getName(), properties, jsc),
properties, jsc, null);

Schema schema = provider.getSourceSchema();
assertEquals(schema.getType(), Type.RECORD);
Expand All @@ -76,9 +86,9 @@ public void testSparkAvro() throws IOException {
transformerClassNames.add(FlatteningTransformer.class.getName());

SchemaProvider provider =
UtilHelpers.wrapSchemaProviderWithPostProcessor(
UtilHelpers.createSchemaProvider(SparkAvroSchemaProvider.class.getName(), properties, jsc),
properties, jsc, transformerClassNames);
UtilHelpers.wrapSchemaProviderWithPostProcessor(
UtilHelpers.createSchemaProvider(SparkAvroSchemaProvider.class.getName(), properties, jsc),
properties, jsc, transformerClassNames);

Schema schema = provider.getSourceSchema();
assertEquals(schema.getType(), Type.RECORD);
Expand Down Expand Up @@ -144,6 +154,48 @@ public void testDeleteColumnThrows() {
Assertions.assertThrows(HoodieSchemaPostProcessException.class, () -> processor.processSchema(schema));
}

@ParameterizedTest
@MethodSource("configParams")
public void testAddPrimitiveTypeColumn(String type) {
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NAME_PROP.key(), "primitive_column");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_TYPE_PROP.key(), type);
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NEXT_PROP.key(), "fare");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_DOC_PROP.key(), "primitive column test");

AddColumnSchemaPostProcessor processor = new AddColumnSchemaPostProcessor(properties, null);
Schema schema = new Schema.Parser().parse(ORIGINAL_SCHEMA);
Schema targetSchema = processor.processSchema(schema);

Schema.Field newColumn = targetSchema.getField("primitive_column");
Schema.Field nextColumn = targetSchema.getField("fare");

assertNotNull(newColumn);
assertEquals("primitive column test", newColumn.doc());
assertEquals(type, newColumn.schema().getType().getName());
assertEquals(nextColumn.pos(), newColumn.pos() + 1);
}

@Test
public void testAddDecimalColumn() {
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_NAME_PROP.key(), "decimal_column");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_TYPE_PROP.key(), "decimal");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_DOC_PROP.key(), "decimal column test");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_DEFAULT_PROP.key(), "0.75");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_PRECISION_PROP.key(), "8");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_SCALE_PROP.key(), "6");
properties.put(AddColumnSchemaPostProcessor.Config.SCHEMA_POST_PROCESSOR_ADD_COLUMN_SIZE_PROP.key(), "8");

AddColumnSchemaPostProcessor processor = new AddColumnSchemaPostProcessor(properties, null);
Schema schema = new Schema.Parser().parse(ORIGINAL_SCHEMA);
Schema targetSchema = processor.processSchema(schema);

Schema.Field newColumn = targetSchema.getField("decimal_column");

assertNotNull(newColumn);
assertEquals("decimal", newColumn.schema().getLogicalType().getName());
assertEquals(5, newColumn.pos());
}

@Test
public void testSparkAvroSchema() throws IOException {
SparkAvroPostProcessor processor = new SparkAvroPostProcessor(properties, null);
Expand Down