Storing DateTime fields in Azure Table Storage

3
votes

I'm trying to store an ISO formatted DateTime in my Azure Table Storage. But every time I try, it results in this String format: https://i.stack.imgur.com/kcrlN.png

This is a summary of the effective code I'm running:

import jsons
from azure.functions import Out

def main(parameterTableOutputJSON: Out[str]):
    row = {'ReadOn': '2021-04-28T09:35:26.123456Z'}
    parameterTableOutputJSON.set(jsons.dumps(row))

Of course my real code also has the PartitionKey and RowKey columns, but I can't show those here.

How can I insert the DateTime in the Azure Table Storage like this: 2021-04-28T09:35:26.123456Z
Instead of this: 04/28/2021 09:35:26

The Azure Table Storage docs tell me it supports ISO formatted date times...

2
pythonazureazure-table-storageazure-storage-account
Please share the complete code where you're writing the entity into Table Storage. - Gaurav Mantri
@GauravMantri I'm writing the entity into Table Storage by calling the set function. The JSON String that goes into this function is the payload that gets written to Table Storage. - Stefan Rang

2 Answers

1
votes

Using the newest Azure Data Tables library you can use the Python datetime object directly within your entity and submit it for creation. Here is an example snippet:

from azure.data.tables import TableClient
from datetime import datetime

client = TableClient.from_connection_string(<my_conn_str>)
my_entity = {
    "PartitionKey": <my_pk>,
    "RowKey": <my_rk>,
    "ReadOn": datetime(2021, 4, 28, 9, 36, 26, 123456)
}
client.upsert_entity(my_entity)

Here is what the entity looks like in the Storage Explorer in the portal

Disclaimer: I work on the Azure SDK for Python team.

0
votes

I found the solution to my own problem.

The jsons.dumps method strips the milliseconds by default. This results in a datetime that's formatted like this: 2021-04-28T09:35:26Z
This format is not supported as a DateTime in Azure Table Storage.

The problem is solved by writing the last line of my code like this:

parameterTableOutputJSON.set(jsons.dumps(row, strip_microseconds=False))