使用適用於 Python 的 SDK (Boto3) 的 Kinesis 範例 - AWS SDK 程式碼範例
動作無伺服器範例

文件 AWS 開發套件範例 GitHub 儲存庫中有更多可用的 AWS SDK 範例。

本文為英文版的機器翻譯版本，如內容有任何歧義或不一致之處，概以英文版為準。

使用適用於 Python 的 SDK (Boto3) 的 Kinesis 範例

下列程式碼範例示範如何使用 AWS SDK for Python (Boto3) 搭配 Kinesis 來執行動作和實作常見案例。

Actions 是大型程式的程式碼摘錄，必須在內容中執行。雖然動作會告訴您如何呼叫個別服務函數，但您可以在其相關情境中查看內容中的動作。

每個範例都包含完整原始程式碼的連結，您可以在其中找到如何在內容中設定和執行程式碼的指示。

動作

下列程式碼範例示範如何使用 CreateStream。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在 AWS 程式碼範例儲存庫中設定和執行。

class KinesisStream:
    """Encapsulates a Kinesis stream."""

    def __init__(self, kinesis_client):
        """
        :param kinesis_client: A Boto3 Kinesis client.
        """
        self.kinesis_client = kinesis_client
        self.name = None
        self.details = None
        self.stream_exists_waiter = kinesis_client.get_waiter("stream_exists")


    def create(self, name, wait_until_exists=True):
        """
        Creates a stream.

        :param name: The name of the stream.
        :param wait_until_exists: When True, waits until the service reports that
                                  the stream exists, then queries for its metadata.
        """
        try:
            self.kinesis_client.create_stream(StreamName=name, ShardCount=1)
            self.name = name
            logger.info("Created stream %s.", name)
            if wait_until_exists:
                logger.info("Waiting until exists.")
                self.stream_exists_waiter.wait(StreamName=name)
                self.describe(name)
        except ClientError:
            logger.exception("Couldn't create stream %s.", name)
            raise

  • 如需 API 詳細資訊，請參閱 SDK AWS for Python (Boto3) API 參考中的 CreateStream。

下列程式碼範例示範如何使用 DeleteStream。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在 AWS 程式碼範例儲存庫中設定和執行。

class KinesisStream:
    """Encapsulates a Kinesis stream."""

    def __init__(self, kinesis_client):
        """
        :param kinesis_client: A Boto3 Kinesis client.
        """
        self.kinesis_client = kinesis_client
        self.name = None
        self.details = None
        self.stream_exists_waiter = kinesis_client.get_waiter("stream_exists")


    def delete(self):
        """
        Deletes a stream.
        """
        try:
            self.kinesis_client.delete_stream(StreamName=self.name)
            self._clear()
            logger.info("Deleted stream %s.", self.name)
        except ClientError:
            logger.exception("Couldn't delete stream %s.", self.name)
            raise

  • 如需 API 詳細資訊，請參閱 SDK AWS for Python (Boto3) API 參考中的 DeleteStream。

下列程式碼範例示範如何使用 DescribeStream。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在 AWS 程式碼範例儲存庫中設定和執行。

class KinesisStream:
    """Encapsulates a Kinesis stream."""

    def __init__(self, kinesis_client):
        """
        :param kinesis_client: A Boto3 Kinesis client.
        """
        self.kinesis_client = kinesis_client
        self.name = None
        self.details = None
        self.stream_exists_waiter = kinesis_client.get_waiter("stream_exists")


    def describe(self, name):
        """
        Gets metadata about a stream.

        :param name: The name of the stream.
        :return: Metadata about the stream.
        """
        try:
            response = self.kinesis_client.describe_stream(StreamName=name)
            self.name = name
            self.details = response["StreamDescription"]
            logger.info("Got stream %s.", name)
        except ClientError:
            logger.exception("Couldn't get %s.", name)
            raise
        else:
            return self.details

  • 如需 API 詳細資訊，請參閱 SDK AWS for Python (Boto3) API 參考中的 DescribeStream。

下列程式碼範例示範如何使用 GetRecords。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在 AWS 程式碼範例儲存庫中設定和執行。

class KinesisStream:
    """Encapsulates a Kinesis stream."""

    def __init__(self, kinesis_client):
        """
        :param kinesis_client: A Boto3 Kinesis client.
        """
        self.kinesis_client = kinesis_client
        self.name = None
        self.details = None
        self.stream_exists_waiter = kinesis_client.get_waiter("stream_exists")


    def get_records(self, max_records):
        """
        Gets records from the stream. This function is a generator that first gets
        a shard iterator for the stream, then uses the shard iterator to get records
        in batches from the stream. The shard iterator can be accessed through the
        'details' property, which is populated using the 'describe' function of this class.
        Each batch of records is yielded back to the caller until the specified
        maximum number of records has been retrieved.

        :param max_records: The maximum number of records to retrieve.
        :return: Yields the current batch of retrieved records.
        """
        try:
            response = self.kinesis_client.get_shard_iterator(
                StreamName=self.name,
                ShardId=self.details["Shards"][0]["ShardId"],
                ShardIteratorType="LATEST",
            )
            shard_iter = response["ShardIterator"]
            record_count = 0
            while record_count < max_records:
                response = self.kinesis_client.get_records(
                    ShardIterator=shard_iter, Limit=10
                )
                shard_iter = response["NextShardIterator"]
                records = response["Records"]
                logger.info("Got %s records.", len(records))
                record_count += len(records)
                yield records
        except ClientError:
            logger.exception("Couldn't get records from stream %s.", self.name)
            raise



    def describe(self, name):
        """
        Gets metadata about a stream.

        :param name: The name of the stream.
        :return: Metadata about the stream.
        """
        try:
            response = self.kinesis_client.describe_stream(StreamName=name)
            self.name = name
            self.details = response["StreamDescription"]
            logger.info("Got stream %s.", name)
        except ClientError:
            logger.exception("Couldn't get %s.", name)
            raise
        else:
            return self.details

  • 如需 API 詳細資訊，請參閱 SDK AWS for Python (Boto3) API 參考中的 GetRecords。

下列程式碼範例示範如何使用 PutRecord。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在 AWS 程式碼範例儲存庫中設定和執行。

class KinesisStream:
    """Encapsulates a Kinesis stream."""

    def __init__(self, kinesis_client):
        """
        :param kinesis_client: A Boto3 Kinesis client.
        """
        self.kinesis_client = kinesis_client
        self.name = None
        self.details = None
        self.stream_exists_waiter = kinesis_client.get_waiter("stream_exists")


    def put_record(self, data, partition_key):
        """
        Puts data into the stream. The data is formatted as JSON before it is passed
        to the stream.

        :param data: The data to put in the stream.
        :param partition_key: The partition key to use for the data.
        :return: Metadata about the record, including its shard ID and sequence number.
        """
        try:
            response = self.kinesis_client.put_record(
                StreamName=self.name, Data=json.dumps(data), PartitionKey=partition_key
            )
            logger.info("Put record in stream %s.", self.name)
        except ClientError:
            logger.exception("Couldn't put record in stream %s.", self.name)
            raise
        else:
            return response

  • 如需 API 詳細資訊，請參閱 SDK AWS for Python (Boto3) API 參考中的 PutRecord。

無伺服器範例

下列程式碼範例示範如何實作 Lambda 函數，該函數接收從 Kinesis 串流接收記錄所觸發的事件。此函數會擷取 Kinesis 承載、從 Base64 解碼，並記錄記錄內容。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在無伺服器範例儲存庫中設定和執行。

使用 Python 搭配 Lambda 來使用 Kinesis 事件。

# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
# SPDX-License-Identifier: Apache-2.0
import base64
def lambda_handler(event, context):

    for record in event['Records']:
        try:
            print(f"Processed Kinesis Event - EventID: {record['eventID']}")
            record_data = base64.b64decode(record['kinesis']['data']).decode('utf-8')
            print(f"Record Data: {record_data}")
            # TODO: Do interesting work based on the new data
        except Exception as e:
            print(f"An error occurred {e}")
            raise e
    print(f"Successfully processed {len(event['Records'])} records.")

下列程式碼範例示範如何為從 Kinesis 串流接收事件的 Lambda 函數實作部分批次回應。此函數會在回應中報告批次項目失敗，指示 Lambda 稍後重試這些訊息。

SDK for Python (Boto3)
注意

GitHub 上提供更多範例。尋找完整範例，並了解如何在無伺服器範例儲存庫中設定和執行。

透過使用 Python 的 Lambda 報告 Kinesis 批次項目失敗。

# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
# SPDX-License-Identifier: Apache-2.0
def handler(event, context):
    records = event.get("Records")
    curRecordSequenceNumber = ""
    
    for record in records:
        try:
            # Process your record
            curRecordSequenceNumber = record["kinesis"]["sequenceNumber"]
        except Exception as e:
            # Return failed record's sequence number
            return {"batchItemFailures":[{"itemIdentifier": curRecordSequenceNumber}]}

    return {"batchItemFailures":[]}