使用以下 Beam 代码来从 Kafka 读取数据并将其写入到文件中，每 10 秒钟创建一个新文件：

import apache_beam as beam
from apache_beam import window
from apache_beam.transforms.trigger import AfterProcessingTime

# Kafka configuration
KAFKA_BOOTSTRAP_SERVERS = ''
KAFKA_TOPIC = ''

# Beam pipeline configuration
OUTPUT_DIR = ''

# Pipeline
def run():
    with beam.Pipeline() as p:
        records = (
            p
            | 'ReadFromKafka' >> beam.io.ReadFromKafka(
                consumer_config={'bootstrap.servers': KAFKA_BOOTSTRAP_SERVERS},
                topics=[KAFKA_TOPIC],
            )
            | 'Format' >> beam.Map(lambda r: str(r.value, 'utf-8'))
            | 'Window' >> beam.WindowInto(window.FixedWindows(10))
            | 'WriteToFiles' >> beam.io.WriteToText(
                OUTPUT_DIR,
                file_name_suffix='.txt'
            )
        )

if __name__ == '__main__':
    run()

这段代码定义了一个 Beam 管道，它包含从 Kafka 读取记录、转换记录、应用固定窗口和将数据写入文本文件的步骤。这里的窗口大小为10秒钟，并且每10秒钟都会创建一个新文件。