以下是一个使用AWS数据管道将CSV文件从S3上传到DynamoDB的解决方案，包含代码示例：

1. 创建一个Lambda函数来处理数据管道的转换和加载任务：

import boto3

def lambda_handler(event, context):
    input_bucket = event['inputBucket']
    input_key = event['inputKey']
    output_table = event['outputTable']

    # 从S3下载CSV文件
    s3 = boto3.client('s3')
    s3.download_file(input_bucket, input_key, '/tmp/input.csv')
    
    # 处理CSV文件，转换为DynamoDB条目
    items = []
    with open('/tmp/input.csv', 'r') as file:
        lines = file.readlines()
        headers = lines[0].strip().split(',')
        for line in lines[1:]:
            values = line.strip().split(',')
            item = {}
            for i in range(len(headers)):
                item[headers[i]] = values[i]
            items.append(item)
    
    # 将转换后的条目加载到DynamoDB表中
    dynamodb = boto3.resource('dynamodb')
    table = dynamodb.Table(output_table)
    with table.batch_writer() as batch:
        for item in items:
            batch.put_item(Item=item)
    
    return 'Data pipeline completed successfully.'

2. 创建一个数据管道定义文件（pipeline-definition.json），指定输入和输出：

{
  "objects": [
    {
      "id": "S3Input",
      "name": "S3Input",
      "schedule": {
        "ref": "DefaultSchedule"
      },
      "type": "S3DataNode",
      "dataFormat": {
        "ref": "CSVDataFormat"
      },
      "scheduleType": "ONDEMAND",
      "directoryPath": "#{myInputS3Path}",
      "errorThreshold": 0,
      "errorHandlingConfig": {
        "failOnError": true
      }
    },
    {
      "id": "DynamoDBOutput",
      "name": "DynamoDBOutput",
      "schedule": {
        "ref": "DefaultSchedule"
      },
      "type": "DynamoDBDataNode",
      "scheduleType": "ONDEMAND",
      "tableName": "#{myOutputDynamoDBTable}",
      "errorThreshold": 0,
      "errorHandlingConfig": {
        "failOnError": true
      }
    }
  ],
  "parameters": [
    {
      "id": "myInputS3Path",
      "description": "S3 input path",
      "type": "String"
    },
    {
      "id": "myOutputDynamoDBTable",
      "description": "DynamoDB output table",
      "type": "String"
    }
  ]
}

3. 创建一个数据管道任务，指定Lambda函数和数据管道定义文件：

import boto3

def create_data_pipeline(input_bucket, input_key, output_table):
    client = boto3.client('datapipeline')
    
    # 创建数据管道定义
    with open('pipeline-definition.json', 'r') as file:
        pipeline_definition = file.read()
    
    # 创建数据管道
    response = client.create_pipeline(
        name='csv-to-dynamodb-pipeline',
        uniqueId='csv-to-dynamodb-pipeline',
        description='Pipeline to upload CSV file to DynamoDB',
        pipelineTags=[
            {
                'key': 'environment',
                'value': 'development'
            },
        ],
        pipelineObjects=[
            {
                'id': 'S3Input',
                'name': 'S3Input',
                'fields': [
                    {
                        'key': 'myInputS3Path',
                        'stringValue': 's3://' + input_bucket + '/' + input_key
                    }
                ]
            },
            {
                'id': 'DynamoDBOutput',
                'name': 'DynamoDBOutput',
                'fields': [
                    {
                        'key': 'myOutputDynamoDBTable',
                        'stringValue': output_table
                    }
                ]
            }
        ],
        parameterObjects=[
            {
                'id': 'myInputS3Path',
                'attributes': []
            },
            {
                'id': 'myOutputDynamoDBTable',
                'attributes': []
            }
        ],
        parameterValues=[
            {
                'id': 'myInputS3Path',
                'stringValue': 's3://' + input_bucket + '/' + input_key
            },
            {
                'id': 'myOutputDynamoDBTable',
                'stringValue': output_table
            }
        ],
        pipelineDefinition=pipeline_definition
    )