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

[YAML] Implement basic java mapping operations. #28657

Merged
merged 10 commits into from
Oct 31, 2023
Merged
Original file line number Diff line number Diff line change
@@ -0,0 +1,253 @@
/*
* 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.beam.sdk.schemas.transforms.providers;

import com.google.auto.service.AutoService;
import com.google.auto.value.AutoValue;
import java.net.MalformedURLException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import java.util.Map;
import javax.annotation.Nullable;
import org.apache.beam.sdk.schemas.AutoValueSchema;
import org.apache.beam.sdk.schemas.Schema;
import org.apache.beam.sdk.schemas.annotations.DefaultSchema;
import org.apache.beam.sdk.schemas.annotations.SchemaFieldDescription;
import org.apache.beam.sdk.schemas.transforms.SchemaTransform;
import org.apache.beam.sdk.schemas.transforms.SchemaTransformProvider;
import org.apache.beam.sdk.schemas.transforms.TypedSchemaTransformProvider;
import org.apache.beam.sdk.transforms.DoFn;
import org.apache.beam.sdk.transforms.ParDo;
import org.apache.beam.sdk.values.PCollectionRowTuple;
import org.apache.beam.sdk.values.PCollectionTuple;
import org.apache.beam.sdk.values.Row;
import org.apache.beam.sdk.values.TupleTag;
import org.apache.beam.sdk.values.TupleTagList;

/**
* An implementation of {@link TypedSchemaTransformProvider} for MapToFields for the java language.
*
* <p><b>Internal only:</b> This class is actively being worked on, and it will likely change. We
* provide no backwards compatibility guarantees, and it should not be implemented outside the Beam
* repository.
*/
@SuppressWarnings({
"nullness" // TODO(https://github.com/apache/beam/issues/20497)
})
@AutoService(SchemaTransformProvider.class)
public class JavaMapToFieldsTransformProvider
extends TypedSchemaTransformProvider<JavaMapToFieldsTransformProvider.Configuration> {
protected static final String INPUT_ROWS_TAG = "input";
protected static final String OUTPUT_ROWS_TAG = "output";

@Override
protected Class<Configuration> configurationClass() {
return Configuration.class;
}

@Override
protected SchemaTransform from(Configuration configuration) {
return new JavaMapToFieldsTransform(configuration);
}

@Override
public String identifier() {
return String.format("beam:schematransform:org.apache.beam:yaml:map_to_fields-java:v1");
}

@Override
public List<String> inputCollectionNames() {
return Collections.singletonList(INPUT_ROWS_TAG);
}

@Override
public List<String> outputCollectionNames() {
return Collections.singletonList(OUTPUT_ROWS_TAG);
}

@DefaultSchema(AutoValueSchema.class)
@AutoValue
public abstract static class Configuration {
@Nullable
public abstract String getLanguage();

@Nullable
public abstract Boolean getAppend();

@Nullable
public abstract List<String> getDrop();

public abstract Map<String, JavaRowUdf.Configuration> getFields();

@Nullable
@SuppressWarnings("all")
public abstract ErrorHandling getError_handling();
Copy link
Contributor

Choose a reason for hiding this comment

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

Is there a reason for naming these with underscores when they can be remapped by the YAML provider?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

No, I can go ahead and fix this.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Done.


public static Builder builder() {
return new AutoValue_JavaMapToFieldsTransformProvider_Configuration.Builder();
}

@AutoValue.Builder
public abstract static class Builder {

public abstract Builder setLanguage(String language);

public abstract Builder setAppend(Boolean append);

public abstract Builder setDrop(List<String> drop);

public abstract Builder setFields(Map<String, JavaRowUdf.Configuration> fields);

@SuppressWarnings("all")
public abstract Builder setError_handling(ErrorHandling error_handling);

public abstract Configuration build();
}

@AutoValue
public abstract static class ErrorHandling {
@SchemaFieldDescription("The name of the output PCollection containing failed writes.")
public abstract String getOutput();

public static Builder builder() {
return new AutoValue_JavaMapToFieldsTransformProvider_Configuration_ErrorHandling.Builder();
}

@AutoValue.Builder
public abstract static class Builder {
public abstract Builder setOutput(String output);

public abstract ErrorHandling build();
}
}
}

/** A {@link SchemaTransform} for MapToFields-java. */
protected static class JavaMapToFieldsTransform extends SchemaTransform {

private final Configuration configuration;

JavaMapToFieldsTransform(Configuration configuration) {
this.configuration = configuration;
}

@Override
public PCollectionRowTuple expand(PCollectionRowTuple input) {
Schema inputSchema = input.get(INPUT_ROWS_TAG).getSchema();
Schema.Builder outputSchemaBuilder = new Schema.Builder();
// TODO(yaml): Consider allowing the full java schema naming syntax
// (perhaps as a different dialect/language).
boolean append = configuration.getAppend() != null && configuration.getAppend();
List<String> toDrop =
configuration.getDrop() == null ? Collections.emptyList() : configuration.getDrop();
List<JavaRowUdf> udfs = new ArrayList<>();
if (append) {
for (Schema.Field field : inputSchema.getFields()) {
if (!toDrop.contains(field.getName())) {
try {
udfs.add(
new JavaRowUdf(
JavaRowUdf.Configuration.builder().setExpression(field.getName()).build(),
inputSchema));
} catch (MalformedURLException
| ReflectiveOperationException
| StringCompiler.CompileException exn) {
throw new RuntimeException(exn);
}
outputSchemaBuilder = outputSchemaBuilder.addField(field);
}
}
}
for (Map.Entry<String, JavaRowUdf.Configuration> entry :
configuration.getFields().entrySet()) {
if (!"java".equals(configuration.getLanguage())) {
String expr = entry.getValue().getExpression();
if (expr == null || !inputSchema.hasField(expr)) {
throw new IllegalArgumentException(
"Unknown field or missing language specification for '" + entry.getKey() + "'");
}
}
try {
JavaRowUdf udf = new JavaRowUdf(entry.getValue(), inputSchema);
udfs.add(udf);
outputSchemaBuilder = outputSchemaBuilder.addField(entry.getKey(), udf.getOutputType());
} catch (MalformedURLException
| ReflectiveOperationException
| StringCompiler.CompileException exn) {
throw new RuntimeException(exn);
}
}
Schema outputSchema = outputSchemaBuilder.build();
boolean handleErrors =
configuration.getError_handling() != null
&& configuration.getError_handling().getOutput() != null;
Schema errorSchema =
Schema.of(
Schema.Field.of("failed_row", Schema.FieldType.row(inputSchema)),
Schema.Field.of("error_message", Schema.FieldType.STRING));

PCollectionTuple pcolls =
input
.get(INPUT_ROWS_TAG)
.apply(
"MapToFields",
ParDo.of(createDoFn(udfs, outputSchema, errorSchema, handleErrors))
.withOutputTags(mappedValues, TupleTagList.of(errorValues)));
pcolls.get(mappedValues).setRowSchema(outputSchema);
pcolls.get(errorValues).setRowSchema(errorSchema);

PCollectionRowTuple result =
PCollectionRowTuple.of(OUTPUT_ROWS_TAG, pcolls.get(mappedValues));
if (handleErrors) {
result = result.and(configuration.getError_handling().getOutput(), pcolls.get(errorValues));
}
return result;
}

private static final TupleTag<Row> mappedValues = new TupleTag<Row>() {};
private static final TupleTag<Row> errorValues = new TupleTag<Row>() {};

private static DoFn<Row, Row> createDoFn(
List<JavaRowUdf> udfs, Schema outputSchema, Schema errorSchema, boolean handleErrors) {
return new DoFn<Row, Row>() {
@ProcessElement
public void processElement(@Element Row inputRow, MultiOutputReceiver out) {
try {
Row.Builder outputRow = Row.withSchema(outputSchema);
for (JavaRowUdf udf : udfs) {
outputRow.addValue(udf.getFunction().apply(inputRow));
}
out.get(mappedValues).output(outputRow.build());
} catch (Exception exn) {
if (handleErrors) {
out.get(errorValues)
.output(
Row.withSchema(errorSchema)
.withFieldValue("failed_row", inputRow)
.withFieldValue("error_message", exn.getMessage())
.build());
} else {
throw new RuntimeException(exn);
}
}
}
};
}
}
}
Loading
Loading