Python parser for Outlook OLK binary caches

I discovered that a long-deleted email account when I used Outlook had left behind a cache of emails and calendar invites on my Mac - I was curious, so I started chipping away at the binary files to see if I could fully decipher them (enough of the data is stored in unicode or plain text that I could tell what the files were).

I was mostly successful, and think I ended up able to decode and map more than half the fields that Outlook included in the cache, and definitely all of the important ones.

By default, on macOS Outlook will have put its cache in ~/Users/harry/Library/Group Containers/UBF8T346G9.Office/Outlook/Outlook 15 Profiles/Main Profile - if you just import pyolk and initialize the main class, it'll look in that directory and parse the whole cache. Then calling export will create a folder structure mirroring your cached Outlook inbox, and write all of the emails, events, notes, etc. into those folders.

By default export writes everything to a new Recovered Outlook Data folder in the current working directory.

from pyolk import PyOLKReader
p = PyOLKReader()
p.export()

pyolk.py includes the caller and the interface to Outlook.sqlite, the cache's database / index.

mailobjects.py are @dataclass interfaces for the various different objects that are cached (emails, calendar invites, tasks, mailboxes, etc.)

datafiles.py is the main parser class for the olk15* binary files. All of these use basically the same binary encoding patterns, so a single parser is able to read olk15Message, olk15Category, olk15Event, etc.

utils.py includes helper functions for parsing specific binary data types that were short and used multiple places.