Helper class to convert Apache Avro schema to BigQuery Table schema

AvroToBigQuery.java

package com.mallo64.dataflow

import com.google.api.client.json.GenericJson;
import com.google.api.services.bigquery.model.TableCell;
import com.google.api.services.bigquery.model.TableFieldSchema;
import com.google.api.services.bigquery.model.TableRow;
import com.google.api.services.bigquery.model.TableSchema;
import com.google.cloud.dataflow.sdk.transforms.DoFn;
import org.apache.avro.Schema;
import org.apache.avro.specific.SpecificRecord;

import java.util.List;
import java.util.stream.Collectors;

import static org.apache.avro.Schema.*;

public class AvroToBigQuery<TRecord extends SpecificRecord>  extends DoFn<TRecord, TableRow> {

    @Override
    public void processElement(ProcessContext processContext) throws Exception {
        processContext.output(getTableRow(processContext.element()));
    }

    static TableRow getTableRow(SpecificRecord record) {
        TableRow row = new TableRow();
        encode(record, row);
        return row;
    }


    static TableCell getTableCell(SpecificRecord record) {
        TableCell cell = new TableCell();
        encode(record, cell);
        return cell;
    }

    private static void encode(SpecificRecord record, GenericJson row) {
        Schema schema = record.getSchema();
        schema.getFields().forEach(field -> {
            Type type = field.schema().getType();
            switch (type) {
                case RECORD:
                    row.set(field.name(), getTableCell((SpecificRecord) record.get(field.pos())));
                    break;
                case INT:
                case LONG:
                    row.set(field.name(), ((Number)record.get(field.pos())).longValue());
                    break;
                case BOOLEAN:
                    row.set(field.name(), record.get(field.pos()));
                    break;
                case FLOAT:
                case DOUBLE:
                    row.set(field.name(), ((Number)record.get(field.pos())).doubleValue());
                    break;
                default:
                    row.set(field.name(), String.valueOf(record.get(field.pos())));
            }
        });
    }

    public static TableSchema getTableSchemaRecord(Schema schema) {
        return new TableSchema().setFields(getFieldsSchema(schema.getFields()));
    }

    static List<TableFieldSchema> getFieldsSchema(List<Schema.Field> fields) {
        return fields.stream().map(field -> {
            TableFieldSchema column = new TableFieldSchema().setName(field.name());
            Type type = field.schema().getType();
            switch (type) {
                case RECORD:
                    column.setType("RECORD");
                    column.setFields(getFieldsSchema(fields));
                    break;
                case INT:
                case LONG:
                    column.setType("INTEGER");
                    break;
                case BOOLEAN:
                    column.setType("BOOLEAN");
                    break;
                case FLOAT:
                case DOUBLE:
                    column.setType("FLOAT");
                    break;
                default:
                    column.setType("STRING");
            }
            return column;
        }).collect(Collectors.toList());
    }
}

That's a cool helper, thanks. Wonder if there is an helper to do the opposite.

Did you ever get the reverse?

Hello guys. Can you help me to get info about how I can use the TableSchema for creating an empty BQ table? New BQ java sdk version push me to use the new class that represents the BQ schema: com.google.cloud.bigquery.Schema. It not so flexible to create the new bq schema dynamically like in the example above.