As I understand it, the following should result in an OrderedDict object with the items sorted by the key, which is a datetime object.
ledger = OrderedDict(
sorted(ledger.items(), key=lambda x: datetime.strftime(x[0], '%d-%m-%Y'))
)
However, in my case, the objects are being sorted out of order:
OrderedDict([
(datetime.date(2021, 10, 1), {'net': -33924, 'entries': [{'text': 'Applied to 102101', 'amount': -16962, 'mode': 'debit'}, {'text': 'Applied to 102104', 'amount': -16962, 'mode': 'debit'}]}),
(datetime.date(2021, 11, 3), {'net': -96118, 'entries': [{'text': 'Applied to 102104', 'amount': -96118, 'mode': 'debit'}]}),
(datetime.date(2021, 9, 24), {'net': 18000.0, 'entries': [{'text': 'Payment Received', 'amount': 18000.0, 'mode': 'credit'}]})
])
What are the caveats that I am missing here?
If you want to sort in some way that is not chronological, you can access the attributes you want to sort by and use tuple sorting:
sorted(datetimes, key=lambda d: (d.month, d.seconds)) # or some other weird sorting
If it's just a chronological sort however, then you don't need to bother with any of this:
In [3]: OrderedDict(sorted(ledger.items()))
Out[3]:
OrderedDict([(datetime.date(2021, 9, 24),
{'net': 18000.0,
'entries': [{'text': 'Payment Received',
'amount': 18000.0,
'mode': 'credit'}]}),
(datetime.date(2021, 10, 1),
{'net': -33924,
'entries': [{'text': 'Applied to 102101',
'amount': -16962,
'mode': 'debit'},
{'text': 'Applied to 102104',
'amount': -16962,
'mode': 'debit'}]}),
(datetime.date(2021, 11, 3),
{'net': -96118,
'entries': [{'text': 'Applied to 102104',
'amount': -96118,
'mode': 'debit'}]})])