added more packet types #15

Closed
wants to merge 13 commits into
from

Projects

None yet

2 participants

@alios
alios commented Jul 13, 2012

added type constructor and parsers for:

  • SymetricallyEncryptedDataPacket
  • SymmetricKeyEncryptedSessionKeyPacket
  • TrustPacket
  • UserAttributePacket
@singpolyma
Owner

Trust packets are really just a type tag that happens to be registered. Nothing about their structure is known. What is the advantage of using a TrustPacket constructor over using UnsupportedPacket ?

@singpolyma
Owner

These patches seem to use spaces for both indenting as well as alignment. I can clean that up, but in future please use tabs for indentation.

@singpolyma
Owner

This UserAttribute Packet implementation causes the tests to fail

@singpolyma singpolyma closed this Aug 6, 2012
@singpolyma singpolyma reopened this Aug 6, 2012
@singpolyma
Owner

Oops. Didn't mean to close this.

I've put a cleaned up version of these patches at https://github.com/singpolyma/OpenPGP-Haskell/tree/pull_request_15

I'm still unsure about TrustPacket, and the tests will need to accept UserAttribute. Also, I'd really prefer that UserAttribute actually parse the subpackets usefully: the partial implementation is a bit strange.

@alios
alios commented Aug 7, 2012

The TrustPacket might be useful when parsing f.e. a gpg keyring.

@singpolyma
Owner

Hmm... I'm not against the TrustPacket, I'm just not sure what it gets you over the UnsupportedPacket. I guess it means you don't have to pattern-match the actual tag number yourself. That might be enough.

@singpolyma
Owner

Similar functionality mainlined at singpolyma@3bd3777

@singpolyma

I can't mainline this :)

If Data and Typeable are really needed, they can be derived using the derive tool in a Haskell2010-compatible way. What's your use case for this?

@singpolyma
Owner

Has this been tested at all? http://tools.ietf.org/html/rfc4880#section-5.3 and http://tools.ietf.org/html/rfc4880#section-3.7 say to me that a full s2k specifier is supposed to be in this packet, not just a one-octet useage.

Owner

Equivalent, but correct, functionality added in singpolyma@e2f9358

@singpolyma
Owner

Similar functionality mainlined at singpolyma@86d6407

@singpolyma
Owner

Cleaned up and mainlined at singpolyma@60b8c7b

@singpolyma
Owner

Just parsing to [(Word8, B.ByteString)] makes me a bit uncomfortable. Would prefer to parse the subpackets as with SignatureSubpacket.

@singpolyma
Owner

alios@80e22d2
alios@0224e70
alios@8af7590

are the only feature commits that haven't been in some way merged.

@singpolyma singpolyma closed this Sep 27, 2014
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment