Java SchemaBuilder.FieldAssembler方法代码示例

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
/**
 * Creates a flat Avro Schema for testing.
 */
public static Schema createFlatAvroSchema(String[] fieldNames, TypeInformation[] fieldTypes) {
	final SchemaBuilder.FieldAssembler<Schema> fieldAssembler = SchemaBuilder
		.record("BasicAvroRecord")
		.namespace(NAMESPACE)
		.fields();

	final Schema nullSchema = Schema.create(Schema.Type.NULL);

	for (int i = 0; i < fieldNames.length; i++) {
		Schema schema = ReflectData.get().getSchema(fieldTypes[i].getTypeClass());
		Schema unionSchema = Schema.createUnion(Arrays.asList(nullSchema, schema));
		fieldAssembler.name(fieldNames[i]).type(unionSchema).noDefault();
	}

	return fieldAssembler.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
private static Schema schemaFor(StructType structType, String record, String namespace, int recordCount) {

    // Increment after using the recordCount
    record = (null == record) ? "record" + recordCount++ : record;

    LOG.debug("Converting {} to Avro Record schema [{}:{}]", structType, record, namespace);
    SchemaBuilder.RecordBuilder<Schema> schema = SchemaBuilder.record(record);

    if (null != namespace) {
      schema.namespace(namespace);
    }

    schema.doc("Auto-generated from Spark DataFrame");

    SchemaBuilder.FieldAssembler<Schema> assembler = schema.fields();
    StructField[] structFields = structType.fields();

    for (StructField f : structFields) {
      assembler.name(f.name()).type(typeFor(f.dataType(), f.nullable(), recordCount)).noDefault();
    }

    return assembler.endRecord();
  }
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
public Schema getMakeRowSchema() {
    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.builder().record("MakeRowRecord").fields() //
            .name("ID").type(AvroUtils._decimal()).noDefault() //
            .name("C1").type().nullable().stringType().noDefault() //
            .name("C2").type().nullable().booleanType().noDefault() //
            .name("C3").type().nullable().doubleType().noDefault() //
            // date
            .name("C4").type(AvroUtils._logicalDate()).noDefault() //
            // time
            .name("C5").type(AvroUtils._logicalTime()).noDefault() //
            // timestamp
            .name("C6").type(AvroUtils._logicalTimestamp()).noDefault() //
            // variant
            .name("C7").type().nullable().stringType().noDefault();
    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
@Override
public Schema getMakeRowSchema(boolean isDynamic) {
    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.builder().record("MakeRowRecord").fields() //
            .name("Id").type(AvroUtils._string()).noDefault() //
            .name("Name").type(AvroUtils._string()).noDefault() //
            .name("ShippingStreet").type(AvroUtils._string()).noDefault() //
            .name("ShippingPostalCode").type(AvroUtils._int()).noDefault() //
            .name("BillingStreet").type(AvroUtils._string()).noDefault() //
            .name("BillingState").type(AvroUtils._string()).noDefault() //
            .name("BillingPostalCode").type(AvroUtils._string()).noDefault();
    if (isDynamic) {
        fa = fa.name("ShippingState").type(AvroUtils._string()).noDefault();
    }

    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
@Override
public Schema getMakeRowSchema(boolean isDynamic) {
    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.builder().record("MakeRowRecord").fields() //
            .name("Id").type().nullable().stringType().noDefault() //
            .name("Name").type().nullable().stringType().noDefault() //
            .name("ShippingStreet").type().nullable().stringType().noDefault() //
            .name("ShippingPostalCode").type().nullable().intType().noDefault() //
            .name("BillingStreet").type().nullable().stringType().noDefault() //
            .name("BillingState").type().nullable().stringType().noDefault() //
            .name("BillingPostalCode").type().nullable().stringType().noDefault() //
            .name("BillingCity").type().nullable().stringType().noDefault();
    if (isDynamic) {
        fa = fa.name("ShippingState").type().nullable().stringType().noDefault();
    }

    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
private static Schema assembleNestedSchema(MapEventType mapEventType, ConfigurationEngineDefaults.EventMeta.AvroSettings avroSettings, Annotation[] annotations, EventAdapterService eventAdapterService, String statementName, String engineURI, TypeRepresentationMapper optionalMapper) {
    SchemaBuilder.FieldAssembler<Schema> assembler = record(mapEventType.getName()).fields();

    for (Map.Entry<String, Object> prop : mapEventType.getTypes().entrySet()) {
        AvroSchemaUtil.assembleField(prop.getKey(), prop.getValue(), assembler, annotations, avroSettings, eventAdapterService, statementName, engineURI, optionalMapper);
    }
    return assembler.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
private Schema getMakeRowSchema() {
    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.builder().record("MakeRowRecord").fields() //
            .name("col_1").type().nullable().stringType().noDefault() //
            .name("col_2").type().nullable().stringType().noDefault() //
            .name("col_3").type().nullable().intType().noDefault() //
            .name("col_4").type().nullable().doubleType().noDefault() //
            .name("col_5").prop(SchemaConstants.TALEND_COLUMN_PATTERN, "yyyy-MM-dd'T'HH:mm:ss'.000Z'").type(AvroUtils._date()).noDefault() //
            .name("col_6").type().nullable().stringType().noDefault()
            //the UI model may pass a empty string as the pattern, please see MetadataToolAvroHelper
            .name("col_7").prop(SchemaConstants.TALEND_COLUMN_PATTERN, "").type().nullable().longType().noDefault();
    Schema schema = fa.endRecord();
    return schema;
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
/**
 * Infers an Avro schema for the given DescribeSObjectResult. This can be an expensive operation so the schema
 * should be cached where possible. This is always an {@link Schema.Type#RECORD}.
 *
 * @param in the DescribeSObjectResult to analyse.
 * @return the schema for data given from the object.
 */
private Schema inferStringArray(String[] in) {
    List<Schema.Field> fields = new ArrayList<>();

    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.record(RECORD_NAME).fields();
    for (int i = 0; i < in.length; i++) {
        fa = fa.name(FIELD_PREFIX + i).type(Schema.create(Schema.Type.STRING)).noDefault();
    }
    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
/**
 * Infers an Avro schema for the given String array. This can be an expensive operation so the schema should be
 * cached where possible. This is always an {@link Schema.Type#RECORD}.
 *
 * @param in the String array to analyse.
 * @return the schema for data given from the object.
 */
private Schema inferStringArray(String[] in) {
    List<Schema.Field> fields = new ArrayList<>();

    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.record(RECORD_NAME).fields();
    for (int i = 0; i < in.length; i++) {
        fa = fa.name(FIELD_PREFIX + i).type(Schema.create(Schema.Type.STRING)).noDefault();
    }
    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
/**
 * Infers an Avro schema for the given String array. This can be an expensive operation so the schema should be
 * cached where possible. This is always an {@link Schema.Type#RECORD}.
 *
 * @param in the DescribeSObjectResult to analyse.
 * @return the schema for data given from the object.
 */
private Schema inferCsvRecord(CSVRecord in) {
    List<Schema.Field> fields = new ArrayList<>();

    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.record(RECORD_NAME).fields();
    for (int i = 0; i < in.size(); i++) {
        fa = fa.name(FIELD_PREFIX + i).type(Schema.create(Schema.Type.STRING)).noDefault();
    }
    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
private org.apache.avro.Schema inferBigQuerySchema(Schema schema) {
    List<Field> bqFields = schema.getFields();
    if (bqFields.size() == 0) {
        return SchemaBuilder.builder().record("EmptyRecord").fields().endRecord();
    }

    SchemaBuilder.FieldAssembler<org.apache.avro.Schema> fieldAssembler = SchemaBuilder.record("BigQuerySchema").fields();
    for (Field bqField : bqFields) {
        String name = bqField.getName();
        org.apache.avro.Schema fieldSchema = inferSchemaField(bqField);
        fieldAssembler = fieldAssembler.name(name).type(fieldSchema).noDefault();
    }
    return fieldAssembler.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
public static Schema getSchema(boolean isDynamic) {
    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.builder().record("MakeRowRecord").fields() //
            .name("Id").type().nullable().stringType().noDefault() //
            .name("Name").type().nullable().stringType().noDefault() //
            .name("ShippingStreet").type().nullable().stringType().noDefault() //
            .name("ShippingPostalCode").type().nullable().intType().noDefault() //
            .name("BillingStreet").type().nullable().stringType().noDefault() //
            .name("BillingState").type().nullable().stringType().noDefault() //
            .name("BillingPostalCode").type().nullable().stringType().noDefault();
    if (isDynamic) {
        fa = fa.name("ShippingState").type().nullable().stringType().noDefault();
    }

    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
@Override
public Schema guessSchema(String soqlQuery) throws IOException {
    SoqlQuery query = SoqlQuery.getInstance();
    query.init(soqlQuery);

    SchemaBuilder.FieldAssembler fieldAssembler = SchemaBuilder.record("GuessedSchema").fields();
    DescribeSObjectResult describeSObjectResult = null;

    try {
        describeSObjectResult = connect(null).connection.describeSObject(query.getDrivingEntityName());
    } catch (ConnectionException e) {
        throw new RuntimeException(e.getMessage());
    }

    Schema entitySchema = SalesforceAvroRegistry.get().inferSchema(describeSObjectResult);

    for (FieldDescription fieldDescription : query.getFieldDescriptions()) {
        Schema.Field schemaField = entitySchema.getField(fieldDescription.getSimpleName());

        SchemaBuilder.FieldBuilder builder = fieldAssembler.name(fieldDescription.getFullName());

        Schema fieldType = null;
        if (schemaField != null) {
            Map<String, Object> props = schemaField.getObjectProps();
            for (Map.Entry<String, Object> entry : props.entrySet()) {
                builder.prop(entry.getKey(), String.valueOf(entry.getValue()));
            }
            fieldType = schemaField.schema();
        } else {
            fieldType = DEFAULT_GUESS_SCHEMA_TYPE;
        }
        builder.type(fieldType).noDefault();
    }

    return (Schema) fieldAssembler.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
public Schema inferStringArray(String[] in) {
    List<Schema.Field> fields = new ArrayList<>();

    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.record(RECORD_NAME).fields();
    for (int i = 0; i < in.length; i++) {
        fa = fa.name(FIELD_PREFIX + i).type(Schema.create(Schema.Type.STRING)).noDefault();
    }
    return fa.endRecord();
}
 

import org.apache.avro.SchemaBuilder; //导入方法依赖的package包/类
/**
 * Infers an Avro schema for the given String array. This can be an expensive operation so the schema should be
 * cached where possible. This is always an {@link Schema.Type#RECORD}.
 *
 * @param in the DescribeSObjectResult to analyse.
 * @return the schema for data given from the object.
 */
static Schema inferSchema(CSVRecord in) {
    List<Schema.Field> fields = new ArrayList<>();
    SchemaBuilder.FieldAssembler<Schema> fa = SchemaBuilder.record(RECORD_NAME).fields();
    for (int i = 0; i < in.size(); i++) {
        fa = fa.name(in.get(i)).type(Schema.create(Schema.Type.STRING)).noDefault();
    }
    return fa.endRecord();
}