以下代码

builder = SparkSession.builder\
   .appName("PythonTest11")
spark = builder.getOrCreate()

# spark.conf.set("spark.sql.debug.maxToStringFields", 10000)

# Subscribe to 1 topic

df = spark \
  .readStream \
  .format("kafka") \
  .option("kafka.bootstrap.servers", config["kafka"]["bootstrap.servers"]) \
  .option("subscribe", dataFlowTopic) \
  .load()

df = df \
    .selectExpr("LENGTH(value)")

# .selectExpr("CAST(value as string)") \

df.printSchema()

# Start running the query that prints the running counts to the console

query = df \
    .writeStream \
    .outputMode('append') \
    .format('console') \
    .start()

query.awaitTermination()

印刷品

+-------------+
|length(value)|
+-------------+
|         4095|
+-------------+

对于任何大消息，也就是说，它会截断传入的字符串。
如何解决这个问题？