Handling response from DynamoDB - Python

Question

I'm receiving following data from DynamoDB as a response to client scan API call.

Data = [{'id': {'S': '5'},
         'FirstName': {'S': 'Prashanth'},
         'LastName': {'S': 'Wadeyar'},
         'ClientName': {'S': 'Test'}}]

I want to handle this response and get the output as

{'FirstName':'Prashanth', 'LastName': 'Wadeyar', 'ClientName': 'Test'}

I can handle it by separating it like

for field_obj in data:
    obj = (field_obj['FirstName'])

but to get value of Firstname, Key 'S' may differ for each object. like boolean string, list etc.

is there a easy way to get the key and value pairs.

Andrey · Accepted Answer · 2020-04-16 08:24:02Z

2

If you don't want to bring external dependencies you can use Table class as described here.

import boto3

dynamodb = boto3.resource('dynamodb')
table = dynamodb.Table('users')

response = table.get_item(
    Key={
        'username': 'janedoe',
        'last_name': 'Doe'
    }
)
item = response['Item']
print(item)

Expected Output:

{u'username': u'janedoe',
 u'first_name': u'Jane',
 u'last_name': u'Doe',
 u'account_type': u'standard_user',
 u'age': Decimal('25')}

But my personal preference everytime I hear about Python and DynamoDB is to use PynamoDB: https://pynamodb.readthedocs.io/en/latest/ which is sort of an ORM for DynamoDB.

answered Apr 16, 2020 at 8:24

Andrey

60.3k12 gold badges127 silver badges169 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Prashanth · Accepted Answer · 2020-04-16 12:21:35Z

0

Above answer from Mayank seems to work well, but I also wanted to map fields I want to return in a list. Hence I mapped field and then was able to get the result I was looking for

FIELD_MAP = ['firstName', 'lastName',
               'clientName']

    for field_obj in response['Items']:
        for key in FIELD_MAP:
            if field_obj.get(key):
                obj = (field_obj.get(key))
                for i in obj:
                    value = obj[i]
                db_result.append({key: value})
    print(db_result)

answered Apr 16, 2020 at 12:21

Prashanth

1414 silver badges14 bronze badges

4 Comments

Andrey Over a year ago

What about numbers and complex types like map/list/set?

Prashanth Over a year ago

We would need to loop through again if its a list. Although my query result does not have any other types. it would be the same

Andrey Over a year ago

you will need recursive walk, not a loop, do you really want it? There is much simpler 1 line solution, I am genuinely wondering why would you want to reimplement something that is already implemented in boto for you.

Prashanth Over a year ago

yeah, I got it @Andrey, I was restricted to use only client and not resource method. Hence had to deal with it. Using resource object makes job lot more simpler. Thanks for your comment

Mayank Porwal · Accepted Answer · 2020-04-16 08:32:06Z

-1

Something like this could work, where you don't need to care about the internal keys(S in your case):

In [1018]: d = {}

In [1016]: for i in Data: 
      ...:     for k,v in i.items():
      ...:         if k != 'id':
      ...:             if isinstance(v, dict): 
      ...:                 for j, val in v.items(): 
      ...:                     d[k] = val 
      ...:                  

In [1024]: d                                                                                                                                                                                                
Out[1024]: {'FirstName': 'Prashanth', 'LastName': 'Wadeyar', 'ClientName': 'Test'}

answered Apr 16, 2020 at 8:32

Mayank Porwal

34.2k9 gold badges45 silver badges65 bronze badges

2 Comments

Prashanth Over a year ago

Thank you, I was able to get through with something like below as I wanted to iterate each item and append through the list for field_obj in data: for key in FIELD_MAP: if field_obj.get(key): obj = (field_obj.get(key)) for i in obj: value = obj[i] db_result.append({key: value})

Andrey Over a year ago

This is a bad approach for the following reasons: 1) it ignores types, all primitive values will come out as strings 2) it doesn't support nested structures at all (set/map/list). For those reasons I am downvoting it (and not because I have my answer).

Collectives™ on Stack Overflow

Handling response from DynamoDB - Python

3 Answers 3

Comments

4 Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

4 Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related