Skip to content

[HUDI-5058]Fix flink catalog read spark table error : primary key col can not be nullable #7009

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

Merged
danny0405 merged 4 commits into from
Oct 26, 2022
Merged

[HUDI-5058]Fix flink catalog read spark table error : primary key col can not be nullable #7009

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
da6ea1d
1.fix read spark table error : primary key col cat not nullable
waywtdcc Oct 20, 2022
96c1c1d
1.patch
waywtdcc Oct 21, 2022
fac1600
1.patch2
waywtdcc Oct 25, 2022
77dd5d5
1.commit 3
waywtdcc Oct 26, 2022
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
15 changes: 11 additions & 4 deletions ...-datasource/hudi-flink/src/main/java/org/apache/hudi/table/catalog/HoodieHiveCatalog.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@
import org.apache.hudi.sync.common.util.ConfigUtils;
import org.apache.hudi.table.format.FilePathUtils;
import org.apache.hudi.util.AvroSchemaConverter;
import org.apache.hudi.util.DataTypeUtils;
import org.apache.hudi.util.StreamerUtil;

import org.apache.avro.Schema;
Expand Down Expand Up @@ -66,6 +67,7 @@
import org.apache.flink.table.catalog.stats.CatalogColumnStatistics;
import org.apache.flink.table.catalog.stats.CatalogTableStatistics;
import org.apache.flink.table.expressions.Expression;
import org.apache.flink.table.types.DataType;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hive.conf.HiveConf;
import org.apache.hadoop.hive.metastore.IMetaStoreClient;
Expand Down Expand Up @@ -397,17 +399,22 @@ public CatalogBaseTable getTable(ObjectPath tablePath) throws TableNotExistExcep
String path = hiveTable.getSd().getLocation();
Map<String, String> parameters = hiveTable.getParameters();
Schema latestTableSchema = StreamerUtil.getLatestTableSchema(path, hiveConf);
String pkColumnsStr = parameters.get(FlinkOptions.RECORD_KEY_FIELD.key());
List<String> pkColumns = StringUtils.isNullOrEmpty(pkColumnsStr)
? null : StringUtils.split(pkColumnsStr, ",");
org.apache.flink.table.api.Schema schema;
if (latestTableSchema != null) {
// if the table is initialized from spark, the write schema is nullable for pk columns.
DataType tableDataType = DataTypeUtils.ensureColumnsAsNonNullable(
AvroSchemaConverter.convertToDataType(latestTableSchema), pkColumns);
org.apache.flink.table.api.Schema.Builder builder = org.apache.flink.table.api.Schema.newBuilder()
.fromRowDataType(AvroSchemaConverter.convertToDataType(latestTableSchema));
.fromRowDataType(tableDataType);
String pkConstraintName = parameters.get(PK_CONSTRAINT_NAME);
String pkColumns = parameters.get(FlinkOptions.RECORD_KEY_FIELD.key());
if (!StringUtils.isNullOrEmpty(pkConstraintName)) {
// pkColumns expect not to be null
builder.primaryKeyNamed(pkConstraintName, StringUtils.split(pkColumns, ","));
builder.primaryKeyNamed(pkConstraintName, pkColumns);
} else if (pkColumns != null) {
builder.primaryKey(StringUtils.split(pkColumns, ","));
builder.primaryKey(pkColumns);
}
schema = builder.build();
} else {
Expand Down
40 changes: 40 additions & 0 deletions hudi-flink-datasource/hudi-flink/src/main/java/org/apache/hudi/util/DataTypeUtils.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@

package org.apache.hudi.util;

import org.apache.flink.table.api.DataTypes;
import org.apache.flink.table.types.DataType;
import org.apache.flink.table.types.logical.LocalZonedTimestampType;
import org.apache.flink.table.types.logical.LogicalType;
Expand All @@ -26,10 +27,14 @@
import org.apache.flink.table.types.logical.RowType;
import org.apache.flink.table.types.logical.TimestampType;

import javax.annotation.Nullable;

import java.math.BigDecimal;
import java.time.LocalDate;
import java.time.LocalDateTime;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;

/**
* Utilities for {@link org.apache.flink.table.types.DataType}.
Expand Down Expand Up @@ -123,4 +128,39 @@ public static Object resolvePartition(String partition, DataType type) {
"Can not convert %s to type %s for partition value", partition, type));
}
}

/**
* Ensures the give columns of the row data type are not nullable(for example, the primary keys).
*
* @param dataType The row data type
* @param pkColumns The primary keys
* @return a new row data type if any column nullability is tweaked or the original data type
*/
public static DataType ensureColumnsAsNonNullable(DataType dataType, @Nullable List<String> pkColumns) {
Copy link
Contributor

@alexeykudinkin alexeykudinkin Oct 24, 2022

Choose a reason for hiding this comment

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

nit: this method should be generic no reason to couple it to primary-keys (we can call it for non-PK use-cases as well)

Copy link
Contributor Author

@waywtdcc waywtdcc Oct 25, 2022

Choose a reason for hiding this comment

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

Hello, thank you for your reviews. This method is used to convert the primary key field into not null type.

Copy link
Contributor

@alexeykudinkin alexeykudinkin Oct 26, 2022

Choose a reason for hiding this comment

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

@waywtdcc i understand that. What i'm saying though is that there's no reason for us to couple it to primary keys (it should be generic to be able to handle any columns)

if (pkColumns == null || pkColumns.isEmpty()) {
return dataType;
}
RowType rowType = (RowType) dataType.getLogicalType();
List<DataType> oriFieldTypes = dataType.getChildren();
Copy link
Contributor

@alexeykudinkin alexeykudinkin Oct 24, 2022

Choose a reason for hiding this comment

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

nit: originalFieldTypes

List<String> fieldNames = rowType.getFieldNames();
List<DataType> fieldTypes = new ArrayList<>();
boolean tweaked = false;
for (int i = 0; i < fieldNames.size(); i++) {
if (pkColumns.contains(fieldNames.get(i)) && rowType.getTypeAt(i).isNullable()) {
fieldTypes.add(oriFieldTypes.get(i).notNull());
tweaked = true;
} else {
fieldTypes.add(oriFieldTypes.get(i));
}
}
if (!tweaked) {
return dataType;
}
List<DataTypes.Field> fields = new ArrayList<>();
for (int i = 0; i < fieldNames.size(); i++) {
fields.add(DataTypes.FIELD(fieldNames.get(i), fieldTypes.get(i)));
}
return DataTypes.ROW(fields.toArray(new DataTypes.Field[fields.size()])).notNull();
Copy link
Contributor

@alexeykudinkin alexeykudinkin Oct 24, 2022

Choose a reason for hiding this comment

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

Better to do: fields.toArray(DataTypes:Field[]::new)

}

}