Convert long to ISODate using pymongo

Question

I have a collection with date field as long. A sample json format of the collection is as follows:

{
"processID" : "1410449146441-3-8863e29e0055",
"department" : "ABC",
"price" : 9.99,
"unitSold" : 19,
"date" : 1410449146442
},
{
"processID" : "14104491tf-alhb-8863e29e0055",
"department" : "XYZ",
"price" : 19.99,
"unitSold" : 21,
"date" : 1410985985985  
}

I need to convert this date field as ISOdate format so that I can extract year, month and day from it. I need to do it in pymongo. I tried doing it as datetime.datetime.fromtimestamp({"$divide": ["$date", 1e3]}), but got error "TypeError: a float is required"

Can you please let me know a good way of doing it in pymongo

Answer 1

You got the error because fromtimestamp is a Python function not Mongo DB function. .fromtimestamp expects float instead you provided a dict. Use datetime module to convert while iterating through your cursor:

import datetime
for x in cursor:
    timestamp = datetime.datetime.fromtimestamp(x.get("date")/1000)
    # timestamp = object of type datetime.

Note: If you don't divide the long value by 1000, you'll get the error of "year not in range".