Upload a DynamoDB array from Json string in boto3-CodePudding

To upload a JSON file to an AWS DynamoDB table in Python I am happily using the script found on this page, but I can't understand if it is possible to tell Python to split a single string of the JSON file on a specific character in order to create an array of elements on DynamoDB.

For example, let's use this data.json file

[
  {
    "artist": "Romero Allen",
    "song": "Atomic Dim",
    "id": "b4b0da3f-36e3-4569-b196-3ad982f72bbd",
    "priceUsdCents": 392,
    "publisher": "QUAREX|IME|RUME"
  },
  {
    "artist": "Hilda Barnes",
    "song": "Almond Dutch",
    "id": "eeb58c73-603f-4d6b-9e3b-cf587488f488",
    "priceUsdCents": 161,
    "publisher": "LETPRO|SOUNDSCARE"
  }
]

and this script.py file

import boto3
import json

dynamodb = boto3.client('dynamodb')

def upload():
    with open('data.json', 'r') as datafile:
        records = json.load(datafile)
    for song in records:
        print(song)
        item = {
                'artist':{'S':song['artist']},
                'song':{'S':song['song']},
                'id':{'S': song['id']},
                'priceUsdCents':{'S': str(song['priceUsdCents'])},
                'publisher':{'S': song['publisher']}
        }
        print(item)
        response = dynamodb.put_item(
            TableName='basicSongsTable', 
            Item=item
        )
        print("UPLOADING ITEM")
        print(response)

upload()

My target is to edit the script so the publisher column won't include the string

publisher: "QUAREX|IME|RUME"

but a nested array of elements

publisher:["QUAREX","IME","RUME"]

For me, an extra edit of the JSON file with Python before running the upload script is an option.

CodePudding user response：

You can just use .split('|')

item = {
                'artist':{'S':song['artist']},
                'song':{'S':song['song']},
                'id':{'S': song['id']},
                'priceUsdCents':{'S': str(song['priceUsdCents'])},
                'publisher':{'L': song['publisher'].split('|')}
        }