Python AVRO阅读器在解码kafka消息时返回AssertionError

Posted 2023-02-23

【中文标题】Python AVRO阅读器在解码kafka消息时返回AssertionError

【英文标题】：Python AVRO reader returns AssertionError when decoding kafka messages

【发布时间】：2020-02-16 18:21:09

【问题描述】：

新手在玩 Kafka 和 AVRO。 我正在尝试使用kafka-python、avro-python3 包和关注this answer 在 Python 3.7.3 中反序列化 AVRO 消息。

负责解码Kafka消息的函数是

def decode_message(msg_value, reader):
    from io import BytesIO
    from avro.io import BinaryDecoder

    message_bytes = BytesIO(msg_value)
    decoder = BinaryDecoder(message_bytes)
    event_dict = reader.read(decoder)

return event_dict

其中reader 定义为avro.io.DatumReader 实例：

def create_reader(filename_path):
    from avro.io import DatumReader
    import avro.schema

    schema = avro.schema.Parse(open(filename_path).read())
    reader = DatumReader(schema)

    return reader

不幸的是，它失败了。这是回溯：

<_io.BytesIO object at 0x7fab73fe5530>
<avro.io.BinaryDecoder object at 0x7fab74300090>
Traceback (most recent call last):
File "app.py", line 19, in <module>
  kfk.read_messages(kafka_consumer, avro_reader)
File "/app/modules/consume_kafka.py", line 17, in read_messages
  decoded_message = decode_message(msg_value, reader)
File "/app/modules/consume_kafka.py", line 50, in decode_message
  event_dict = reader.read(decoder)
File "/usr/local/lib/python3.7/site-packages/avro/io.py", line 489, in read
  return self.read_data(self.writer_schema, self.reader_schema, decoder)
File "/usr/local/lib/python3.7/site-packages/avro/io.py", line 534, in read_data
  return self.read_record(writer_schema, reader_schema, decoder)
File "/usr/local/lib/python3.7/site-packages/avro/io.py", line 734, in read_record
  field_val = self.read_data(field.type, readers_field.type, decoder)
File "/usr/local/lib/python3.7/site-packages/avro/io.py", line 512, in read_data
  return decoder.read_utf8()
File "/usr/local/lib/python3.7/site-packages/avro/io.py", line 257, in read_utf8
  input_bytes = self.read_bytes()
File "/usr/local/lib/python3.7/site-packages/avro/io.py", line 249, in read_bytes
  assert (nbytes >= 0), nbytes
AssertionError: -40

我能够阅读该消息，它看起来像

b'Obj\x01\x04\x14avro.codec\x08null\x16avro.schema\xbe\t"type":"record","name":"tracks","namespace":"integration","fields":["name":"name","type":"string","name":"data","type":["type":"record","name":"track_upload_verified","namespace":"integration.tracks","fields":["name":"track_id","type":"string","name":"audio_filename","type":"string","name":"track_type","type":"string"],"type":"record","name":"audio_processed","namespace":"integration.tracks","fields":["name":"track_id","type":"string","name":"audio_mp3_filename","type":"string","name":"waveform_samples","type":"type":"array","items":"int","name":"duration","type":"string"]]]\x00\xc4\x8ad\xceF\x9c\xef\x99\n#y7\x96\xba\xb4\x02\xe2\x01*track_upload_verified\x00H341aa6a3-5ecb-4ac0-8f27-bc2fe5abc9d4^tracks-audio/-1khgyI4kYfSf8hq2XiXZjg-1569510465\x08main\xc4\x8ad\xceF\x9c\xef\x99\n#y7\x96\xba\xb4'

这是我所期望的，即生咬。

当我使用 this tool 对其进行验证时，我对架构非常确定。

有人遇到过类似的问题吗？

【问题讨论】：

我建议使用 Confluent Schema Registry 和他们的 python 客户端来处理 Avro

【参考方案1】：

看起来像 avro 文件中的损坏，但根据您的问题和输入很难理解。

尝试以下方法：

安装 aws_schema_registry 从 aws_schema_registry.avro 导入 AvroSchema 初始化 - 架构 = AvroSchema(my_schema) 现在使用来自 kafka 的字节事件： 结果 = schema.read(message)