6

I tried to use exchangelib to extract group mailbox for analysis, and i want to extract within a date range.
tried to use Filter function but seems only work for calender, may I have your advise is there any sample for email ? thanks all.

user2672033
  • 63
  • 1
  • 5

2 Answers2

8

You need to filter on a datetime field that is available on Message items. Message.FIELDS contains all available fields on the Message class. You can list all datetime fields with something like:

>>> [f.name for f in Message.FIELDS if f.value_cls == EWSDateTime]
['datetime_received', 'datetime_sent', 'datetime_created', 'reminder_due_by', 'last_modified_time']

The README shows examples using .filter(start__range(x, y)), but the start field is only available on CalendarItem objects. Instead, use e.g. datetime_received to filter Message objects:

tz = EWSTimeZone.localzone()
emails_from_2017 = account.inbox.filter(datetime_received__range=(
    tz.localize(EWSDateTime(2017, 1, 1)),
    tz.localize(EWSDateTime(2018, 1, 1))
))
Erik Cederstrand
  • 9,643
  • 8
  • 39
  • 63
3
pytz_tz = pytz.timezone('Europe/Copenhagen') #setting the timezone

py_dt = pytz_tz.localize(datetime(year,month,day)) #building the custom date filter with a timezone object

ews_bfr = EWSDateTime.from_datetime(py_dt) #converting the custom date timezone object to a EWS (Exchange Web Service) date object


for item in account.inbox.all().order_by('-datetime_received')[:10000]: #look into the inbox the first 10K emails order desc by date received
    if item.datetime_received < ews_bfr: #if the mail if older than the custom date in the EWS format then apply rule
        item.delete() #delete all filtered emails
        print("Mail deleted Successfully")
guignol
  • 404
  • 4
  • 10
amrutha
  • 31
  • 1
  • Please do explain the answer you have provided rather than posting code only. Also, posting an example with a delete operation is risky. Please update your example to make it more copy-paste friendly. – xenodevil Feb 18 '20 at 08:22
  • I wouldn't recommend this solution as it doesn't scale, if you have a lots of email (like > 500) it will crash. also you don't have any guaranty that it will process all emails due to the [:10000] list limit. it's better to do the filtering on server side. meaning using EWS filters. – guignol Sep 02 '21 at 08:00