search

dovecot / imaptest

IMAP server tester
Other
51 stars 21 forks source link

Urlauth #3

Closed brong closed 2 years ago

brong commented 7 years ago

This is a series of 4 patches.

The first one just changes the urlauth-binary tests to match my reading of what the spec actually intends.

In particular the existing tests are clearly bogus because they look for a BODYPARTSTRUCTURE which still says that the encoding is BASE64, while the BODYPARTSTRUCTURE is supposed to represent the converted data and the encoding should contain the identity domain of the BINARY content if BINARY is requested.

The second patch just adds a boring additional test.

The third adds a test for the case of an unparseable part, to make sure that the server returns the original bodystructure and a NIL BINARY part.

The fourth adds a part with a quoted-printable encoded version of the MIME boundary, which would cause broken results from a server which blindly decodes each individual part without checking for boundaries in the resulting MIME structure.

cmouse commented 7 years ago

Hi! Thanks we'll take a look at this!

sirainen commented 3 years ago

Huh, took forever to look into this. Doesn't seem like there are many/any people using URLAUTH.. I think:

brong commented 3 years ago

That sure is forever! I've just been running a locally modified set of patches so our tests pass. Would be very happy to be able to run vanilla upstream though :)

On Tue, Mar 30, 2021, at 23:18, sirainen wrote:

Huh, took forever to look into this. Doesn't seem like there are many/any people using URLAUTH.. I think:

  • Returning (BODYPARTSTRUCTURE ... BINARY ...) - Right, this is definitely a bug
  • BODYPARTSTRUCTURE returning converted content-transfer-encoding: Yes, but I think it's valid for it to return either 7bit, 8bit or binary, right?
  • As for returning multiple MIME parts .. Didn't we discuss this at some point related to BINARY extension alone, and decide that it's just not allowed to be used for non-leaf parts? Or that it's server-defined behavior what happens when it's non-leaf. So I think I should just remove it from the test.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/dovecot/imaptest/pull/3#issuecomment-810174114, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAEJ3PNT2YYFGYB2M26LYTTGG6TBANCNFSM4CVPMVSQ.

-- Bron Gondwana, CEO, Fastmail Pty Ltd @.***

sirainen commented 2 years ago

DOP-279 - will eventually be merged once we get URLAUTH fixed in Dovecot..

cmouse commented 1 year ago

please reopen against main