Implementation:Apache Flink FileSource ForRecordStreamFormat

Knowledge Sources	Apache Flink Apache Flink FileSource
Domains	Stream_Processing, File_IO
Last Updated	2026-02-09 00:00 GMT

Overview

Concrete tool for constructing file source connectors with stream-wise or bulk format readers provided by the Apache Flink connector-files module.

Description

The FileSource class provides two static factory methods. forRecordStreamFormat accepts a StreamFormat<T> and input paths, internally adapting it to a BulkFormat via StreamFormatAdapter. forBulkFileFormat accepts a BulkFormat<T, FileSourceSplit> directly. Both return a FileSourceBuilder that auto-selects the file enumerator based on format splittability and defaults to LocalityAwareSplitAssigner.

Usage

Import these factory methods to create file sources for Flink pipelines. Use forRecordStreamFormat for text files and forBulkFileFormat for Parquet/ORC.

Code Reference

Source Location

• Repository: Apache Flink
• File: flink-connectors/flink-connector-files/src/main/java/org/apache/flink/connector/file/src/FileSource.java
• Lines: L181-199

Signature

@PublicEvolving
public final class FileSource<T> extends AbstractFileSource<T, FileSourceSplit>
        implements DynamicParallelismInference {

    public static <T> FileSourceBuilder<T> forRecordStreamFormat(
            final StreamFormat<T> streamFormat, final Path... paths) {
        return forBulkFileFormat(new StreamFormatAdapter<>(streamFormat), paths);
    }

    public static <T> FileSourceBuilder<T> forBulkFileFormat(
            final BulkFormat<T, FileSourceSplit> bulkFormat, final Path... paths) {
        checkNotNull(bulkFormat, "reader");
        checkNotNull(paths, "paths");
        checkArgument(paths.length > 0, "paths must not be empty");
        return new FileSourceBuilder<>(paths, bulkFormat);
    }
}

Import

import org.apache.flink.connector.file.src.FileSource;
import org.apache.flink.connector.file.src.reader.StreamFormat;
import org.apache.flink.connector.file.src.reader.BulkFormat;
import org.apache.flink.core.fs.Path;

I/O Contract

Inputs

Name	Type	Required	Description
streamFormat	StreamFormat<T>	Yes (stream)	Record-by-record reader
bulkFormat	BulkFormat<T, FileSourceSplit>	Yes (bulk)	Batch reader (e.g., Parquet)
paths	Path...	Yes	Input file or directory paths

Outputs

Name	Type	Description
builder	FileSourceBuilder<T>	Configurable builder supporting .monitorContinuously(), .processStaticFileSet(), .build()
FileSource<T>	FileSource<T>	Final source instance (after .build())

Usage Examples

Reading Text Files

import org.apache.flink.connector.file.src.FileSource;
import org.apache.flink.formats.text.TextLineInputFormat;
import org.apache.flink.core.fs.Path;

FileSource<String> source = FileSource
    .forRecordStreamFormat(new TextLineInputFormat(), new Path("/input/logs/"))
    .monitorContinuously(Duration.ofSeconds(30))
    .build();

DataStream<String> stream = env.fromSource(source, WatermarkStrategy.noWatermarks(), "file-source");

Reading Parquet Files

import org.apache.flink.connector.file.src.FileSource;
import org.apache.flink.formats.parquet.avro.AvroParquetReaders;

FileSource<GenericRecord> source = FileSource
    .forBulkFileFormat(AvroParquetReaders.forGenericRecord(schema), new Path("/input/parquet/"))
    .processStaticFileSet()
    .build();

Related Pages

Implements Principle

• Principle:Apache_Flink_File_Source_Builder_Configuration

Requires Environment

• Environment:Apache_Flink_Java_Build_Environment