More seahorse refactoring: Using GObjects instead of native GPGME structures
- From: Stef <stef-list memberwebs com>
- To: Seahorse mailing list <seahorse-list gnome org>
- Subject: More seahorse refactoring: Using GObjects instead of native GPGME structures
- Date: Sat, 13 Dec 2008 00:57:19 +0000 (UTC)
I've done some more refactoring work in the PGP code in seahorse:
44 files changed, 3129 insertions(+), 2113 deletions(-)
In the PGP code I've added GObject based classes for subkeys, photos,
and refined the ones for the UIDs.
One interesting aspect of the code is that those classes keep track of
their own GPG indexes, and are then usable directly with the various key
operations. For example you can call seahorse_pgp_key_op_sign_uid()
directly with a SeahorsePgpUID object.
The direction this is going, is becoming less dependant on GPGME and
it's structures. It was a big kludge to have the remote PGP keys (search
results) be represented by the gpgme_key_t structures. But this hasn't
changed yet. :(
An interesting part of the code is bit that calculates the 'actual' GPG
UID indexes from the dumb GPGME UID indexes. Adam, do you have time to
check and make sure I got this right?
renumber_actual_uids() in seahorse-pgp-key.c
Cheers,
Stef Walter
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]