Skip to content

Releases: nylas/exchangelib

v1.11.4-0.12.9: Revert "Increase session pool size to 500"

24 Nov 20:34
Compare
Choose a tag to compare
This reverts commit 94164fae

This change affected all EWS accounts and caused memory issues

Increase session pool size to 500

17 Nov 16:56
1d1e9c4
Compare
Choose a tag to compare
Pre-release

This release is unstable - causes memory issues

Typing only for <Py35

27 Oct 21:27
c91dfc6
Compare
Choose a tag to compare
Merge pull request #55 from nylas/typing-lt-py35

typing is a Py35 backport; don't depend on it if you're above py35

Catch another UnicodeDecodeError

20 Aug 19:44
16ccf8f
Compare
Choose a tag to compare
Merge pull request #53 from nylas/yusra-ahmed/catch-unicode-decode-error

catch another unicodedecodeerror

Decode and encode with utf-8 before chardet detected encoding

04 Aug 21:30
0a3f29b
Compare
Choose a tag to compare
v1.11.4-0.12.4

Merge pull request #52 from nylas/yusra-ahmed/ch26621/schema-validati…

Try different encoding approach in post_ratelimited

13 Jul 21:51
Compare
Choose a tag to compare
We were getting some SSL bad write retry errors, and mostly the internet
just seems to think that passing in unicode is a bad idea. This tries to
encode the data instead. Try utf-8 first, and if that fails we see if
chardet can get the encoding for us. Also had to encode the
authorization header because it was causing the data body to be coerced
(unsuccessfully) to unicode.

Make Content-Length header value a string

01 Jul 19:49
Compare
Choose a tag to compare
To hopefully fix this error
https://sentry.nylas.com/sentry/syncback-prod/issues/54469/

Update post_ratelimited to better handle unicode data

01 Jul 13:10
Compare
Choose a tag to compare
We have this pretty hairy UnicodeDecodeError bug going on:
https://app.clubhouse.io/nylas/story/18092/event-creation-fails-in-syncback-with-unicodedecodeerror

It looks like we have a string with some utf-8 characters in it, but the
string isn't specified as unicode.

My first attempt to fix this was just to call `.decode('utf-8')` on the
data before we passed it down into requests, but that came back to us
with a schema validation error from the provider.

After lots of digging, turns out the Content-Length header was incorrect
:face-palm: Requests doesn't have good enough unicode support to realize
it should calculate the byte length instead of the number of characters.
So, I've rejiggered our request building so that we can calculate the
content length ourselves.

Sort AdditionalProperties FieldPath

29 Jun 20:49
6998918
Compare
Choose a tag to compare

Implement sorting on the AdditionalProperties FieldPath attributes from getitem requests.

Implement MoveFolder

24 Jun 15:46
Compare
Choose a tag to compare
v1.11.4-0.11.0

Implement MoveFolder and add folder.move()