Export DynamoDB to CSV on S3 with Lambda function (python)

Question

Hello im trying to generate a CSV from dynamoDB to S3 using lambda function. the thing is I just get an empty file on s3. Please your help!

import csv
import boto3
import json
dynamodb = boto3.resource('dynamodb')
db = dynamodb.Table('ReporteTelefonica')
def lambda_handler(event, context):
    AWS_BUCKET_NAME = 'reportetelefonica'
    s3 = boto3.resource('s3')
    bucket = s3.Bucket(AWS_BUCKET_NAME)
    path = 'test.csv'
    try:
        response = db.scan()
        myFile = open(path, 'w')  

        for i in response['Items']:
            csv.register_dialect('myDialect', delimiter=',', quoting=csv.QUOTE_NONE)
            with myFile:
                writer = csv.writer(myFile, dialect='myDialect')
                writer.writerows(i)
            print(i)
    except :
        print("error")

    bucket.put_object(
        ACL='public-read-write',
        ContentType='application/csv',
        Key=path,
        # Body=json.dumps(i),
    )
    # print("here")
    body = {
        "uploaded": "true",
        "bucket": AWS_BUCKET_NAME,
        "path": path,
    }
    # print("then here")
    return {
        "statusCode": 200,
        "body": json.dumps(body)
    }

I'm kind of noob on this, so I was wondering what should I modify to successfully make a complete scan of the table and write the values on the CSV on S3???

Answer 1

You either have to close the file after you finished writing the cvs records and then reopen for reading and pass to the put_obkect method. Alternatively you open the file for reading and writing and after writing you seek to position 0 so that the put_object method reads from the start.

Answer 2

Here's a working lambda that will do the job.

import boto3
import json
import os
import pandas as pd

TABLE_NAME = os.environ.get("DDB_TABLE_NAME")
OUTPUT_BUCKET = os.environ.get("BUCKET_NAME")
TEMP_FILENAME = '/tmp/export.csv'
OUTPUT_KEY = 'export.csv'

s3_resource = boto3.resource('s3')
dynamodb_resource = boto3.resource('dynamodb')
table = dynamodb_resource.Table(TABLE_NAME)


def lambda_handler(event, context):
    response = table.scan()
    df = pd.DataFrame(response['Items'])
    df.to_csv(TEMP_FILENAME, index=False, header=True)

    # Upload temp file to S3
    s3_resource.Bucket(OUTPUT_BUCKET).upload_file(TEMP_FILENAME, OUTPUT_KEY)

    return {
        'statusCode': 200,
        'headers': {
            "Access-Control-Allow-Origin": "*",
            "Access-Control-Allow-Credentials": True,
            "content-type": "application/json"
        },
        'body': json.dumps('OK')
    }