Skip to content

Fix correctness issue in JsonToRow, RowJsonDeserializer and RowJsonSerializer #40769

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.

Sign up for GitHub

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

Open
sjvanrossum wants to merge 4 commits into
base: master
Choose a base branch
from
+54 −21
Open

Fix correctness issue in JsonToRow, RowJsonDeserializer and RowJsonSerializer #40769

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
4dd2171
Add reproducer test for non-SQL logical types in RowJson
sjvanrossum Oct 24, 2026
248d1f2
Convert non-SQL logical types between input and base values in RowJson
sjvanrossum Oct 24, 2026
9b81765
Add reproducer test for custom dead-letter field names in JsonToRow
sjvanrossum Oct 24, 2026
0c47006
Use configured line and error field names in JsonToRowWithErrFn
sjvanrossum Oct 24, 2026
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
43 changes: 32 additions & 11 deletions sdks/java/core/src/main/java/org/apache/beam/sdk/transforms/JsonToRow.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -261,22 +261,33 @@ public JsonToRowWithErrFn withNullBehavior(NullBehavior nullBehavior) {
return this.toBuilder().setNullBehavior(nullBehavior).build();
}

Schema getErrorSchema() {
if (getExtendedErrorInfo()) {
if (LINE_FIELD_NAME.equals(getLineFieldName())
&& ERROR_FIELD_NAME.equals(getErrorFieldName())) {
return ERROR_ROW_WITH_ERR_MSG_SCHEMA;
}
return Schema.of(
Field.of(getLineFieldName(), FieldType.STRING),
Field.of(getErrorFieldName(), FieldType.STRING));
} else {
if (LINE_FIELD_NAME.equals(getLineFieldName())) {
return ERROR_ROW_SCHEMA;
}
return Schema.of(Field.of(getLineFieldName(), FieldType.STRING));
}
}

@Override
public ParseResult expand(PCollection<String> jsonStrings) {
Schema errorSchema = getErrorSchema();

PCollectionTuple result =
jsonStrings.apply(
ParDo.of(ParseWithError.create(this))
ParDo.of(ParseWithError.create(this, errorSchema))
.withOutputTags(PARSED_LINE, TupleTagList.of(PARSE_ERROR)));

PCollection<Row> failures;

if (getExtendedErrorInfo()) {
failures =
result.get(PARSE_ERROR).setRowSchema(JsonToRowWithErrFn.ERROR_ROW_WITH_ERR_MSG_SCHEMA);
} else {
failures = result.get(PARSE_ERROR).setRowSchema(JsonToRowWithErrFn.ERROR_ROW_SCHEMA);
}
PCollection<Row> failures = result.get(PARSE_ERROR).setRowSchema(errorSchema);

return ParseResult.resultBuilder()
.setCallingPipeline(jsonStrings.getPipeline())
Expand All @@ -292,18 +303,28 @@ protected abstract static class ParseWithError extends DoFn<String, Row> {

public abstract JsonToRowWithErrFn getJsonToRowWithErrFn();

public abstract Schema getErrorSchema();

public abstract Builder toBuilder();

@AutoValue.Builder
public abstract static class Builder {
public abstract Builder setJsonToRowWithErrFn(JsonToRowWithErrFn value);

public abstract Builder setErrorSchema(Schema value);

public abstract ParseWithError build();
}

public static ParseWithError create(JsonToRowWithErrFn jsonToRowWithErrFn) {
return create(jsonToRowWithErrFn, jsonToRowWithErrFn.getErrorSchema());
}

public static ParseWithError create(
JsonToRowWithErrFn jsonToRowWithErrFn, Schema errorSchema) {
return new AutoValue_JsonToRow_JsonToRowWithErrFn_ParseWithError.Builder()
.setJsonToRowWithErrFn(jsonToRowWithErrFn)
.setErrorSchema(errorSchema)
.build();
}

Expand All @@ -317,14 +338,14 @@ public void processElement(@Element String element, MultiOutputReceiver output)
output
.get(PARSE_ERROR)
.output(
Row.withSchema(ERROR_ROW_WITH_ERR_MSG_SCHEMA)
Row.withSchema(getErrorSchema())
.addValue(element)
.addValue(ex.getMessage())
.build());
} else {
output
.get(PARSE_ERROR)
.output(Row.withSchema(ERROR_ROW_SCHEMA).addValue(element).build());
.output(Row.withSchema(getErrorSchema()).addValue(element).build());
}
return;
}
Expand Down
19 changes: 11 additions & 8 deletions sdks/java/core/src/main/java/org/apache/beam/sdk/util/RowJson.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -321,19 +321,20 @@ private Object extractJsonNodeValue(FieldValue fieldValue) {
}

if (fieldValue.typeName().isLogicalType()) {
String identifier = fieldValue.type().getLogicalType().getIdentifier();
Schema.LogicalType<Object, Object> logicalType =
(Schema.LogicalType<Object, Object>) fieldValue.type().getLogicalType();
String identifier = logicalType.getIdentifier();
if (SqlTypes.DATE.getIdentifier().equals(identifier)) {
return dateValueExtractor().extractValue(fieldValue.jsonValue());
} else if (SqlTypes.TIME.getIdentifier().equals(identifier)) {
return timeValueExtractor().extractValue(fieldValue.jsonValue());
} else if (SqlTypes.DATETIME.getIdentifier().equals(identifier)) {
return localDatetimeValueExtractor().extractValue(fieldValue.jsonValue());
} else {
return extractJsonNodeValue(
FieldValue.of(
fieldValue.name(),
fieldValue.type().getLogicalType().getBaseType(),
fieldValue.jsonValue()));
return logicalType.toInputType(
extractJsonNodeValue(
FieldValue.of(
fieldValue.name(), logicalType.getBaseType(), fieldValue.jsonValue())));
}
}

Expand Down Expand Up @@ -602,15 +603,17 @@ private void writeValue(JsonGenerator gen, FieldType type, Object value) throws
gen.writeEndObject();
break;
case LOGICAL_TYPE:
String identifier = type.getLogicalType().getIdentifier();
Schema.LogicalType<Object, Object> logicalType =
(Schema.LogicalType<Object, Object>) type.getLogicalType();
String identifier = logicalType.getIdentifier();
if (SqlTypes.DATE.getIdentifier().equals(identifier)) {
gen.writeString(((LocalDate) value).toString()); // ISO 8601 format
} else if (SqlTypes.TIME.getIdentifier().equals(identifier)) {
gen.writeString(((LocalTime) value).toString()); // ISO 8601 format
} else if (SqlTypes.DATETIME.getIdentifier().equals(identifier)) {
gen.writeString(((LocalDateTime) value).toString()); // ISO 8601 format
} else {
writeValue(gen, type.getLogicalType().getBaseType(), value);
writeValue(gen, logicalType.getBaseType(), logicalType.toBaseType(value));
}
break;
default:
Expand Down
5 changes: 5 additions & 0 deletions sdks/java/core/src/test/java/org/apache/beam/sdk/transforms/JsonToRowTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,9 @@
*/
package org.apache.beam.sdk.transforms;

import static org.hamcrest.MatcherAssert.assertThat;
import static org.hamcrest.Matchers.equalTo;

import java.io.Serializable;
import java.util.Collection;
import java.util.stream.Collectors;
Expand Down Expand Up @@ -317,6 +320,8 @@ public void testParsesErrorWithErrorMsgRowsDeadLetterWithCustomFieldNames() thro
PCollection<Row> personRows = results.getResults();
PCollection<Row> errorsWithMsg = results.getFailedToParseLines();

assertThat(errorsWithMsg.getSchema(), equalTo(customSchema));

PAssert.that(personRows).containsInAnyOrder(PERSON_ROWS);

PAssert.that(errorsWithMsg)
Expand Down
8 changes: 6 additions & 2 deletions sdks/java/core/src/test/java/org/apache/beam/sdk/util/RowJsonTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,7 @@
import java.util.Map;
import org.apache.beam.sdk.schemas.Schema;
import org.apache.beam.sdk.schemas.Schema.FieldType;
import org.apache.beam.sdk.schemas.logicaltypes.EnumerationType;
import org.apache.beam.sdk.schemas.logicaltypes.VariableString;
import org.apache.beam.sdk.util.RowJson.RowJsonDeserializer;
import org.apache.beam.sdk.util.RowJson.RowJsonDeserializer.NullBehavior;
Expand Down Expand Up @@ -134,15 +135,18 @@ private static Object[] makeFlatRowTestCase() {
}

private static Object[] makeLogicalTypeTestCase() {
EnumerationType enumType = EnumerationType.create("ZERO", "ONE", "TWO");
Schema schema =
Schema.builder()
.addLogicalTypeField("f_passThroughString", VariableString.of(10))
.addLogicalTypeField("f_enum", enumType)
.build();

// fixed string will do padding
String rowString = "{\n" + "\"f_passThroughString\" : \"hello\"\n" + "}";
String rowString =
"{\n" + "\"f_passThroughString\" : \"hello\",\n" + "\"f_enum\" : 1\n" + "}";

Row expectedRow = Row.withSchema(schema).addValues("hello").build();
Row expectedRow = Row.withSchema(schema).addValues("hello", enumType.valueOf("ONE")).build();

return new Object[] {"Logical Types", schema, rowString, expectedRow};
}
Expand Down
Loading