The changes are best reviewed by a developer to check that I didn't make a mistake when I replaced usage of the canSign, canEncrypt, etc. methods with the new hasSign, hasEncrypt, etc. methods or the corresponding compatibility helpers keyHasSign, etc..

It isn't a duplicate. See T6325#176719.

This task is about the creation of OpenPGP certificates or S/MIME CSR from existing card keys. In particular, it isn't about the creation of card keys.

With the above changes a group with an expired key is now expanded. But, as in T6742, the key is shown with the blue info icon instead of with a red X icon and the OK button is enabled.

C++ bindings also done.

The tag of the last displayed user ID that has a tag is chosen. And that's tag tagC1 in the above scenario.

In T6683#176424, @ebo wrote:

See also https://dev.gnupg.org/T6401#176099 which explains why OpenPGP seems to behave differently than S/MIME.

I have created a merge request for upstream to remove the useless "No date" option:
https://invent.kde.org/frameworks/kwidgetsaddons/-/merge_requests/213
I have added a patch with the changes to gpg4win, so that we do not have to wait for the fixed upstream version.

You forgot to add src/inst-pcre2.nsi and src/uninst-pcre2.nsi

The certification of certificate groups (first two points of T6469#174361) is mostly done.

In T5903#176175, @uwi wrote:

I'm late on the train, but (talking about Kleopatra) there is a menu entry to refresh all keys, but I miss a RMB (right mouse button) context menu for the list of certificates to refresh just the current one (much similar like the feature requested, but "one UI level higher").

I know Microsoft is probably not the best example, but copying an already copied file with the Windows Explorer you get "Copy of Copy of Copy of originalname.txt". Moreover, I think foo_signed_signed_signed.pdf makes it pretty clear that this is a PDF that has been signed multiple times. I would leave it as-is. People who don't like the name can easily change it.

I think there is a timing issue between the termination of a job and the retrieval of gpg's output, so that gpg's output is sometimes truncated or even completely empty. This is a general problem and not specific for this ticket.

Somebody else needs to take over.

Encrypting folders with umlauts in the folder name and decrypting archives with umlauts in the file name now works for me on Windows. Encrypting folders and decrypting archives with kanji in the names still fails.

Re-open to add auto-detection.

Relevant lines from Arch's qconfig.pri:

QT.global.enabled_features = shared pkg-config rpath reduce_relocations signaling_nan zstd thread future concurrent dbus openssl-linked opensslv30 shared shared rpath reduce_exports no_direct_extern_access reduce_relocations openssl
QT.global.disabled_features = static cross_compile debug_and_release separate_debug_info appstore-compliant simulator_and_device force_asserts framework c++20 c++2a c++2b c++2b wasm-simd128 wasm-exceptions opensslv11
QT.global.disabled_features += release build_all
QT_CONFIG += shared rpath reduce_exports no_direct_extern_access reduce_relocations openssl release
CONFIG +=  shared plugin_manifest

I have added the additional suffixes. And I have added a patch to gpg4win which makes the fix work on Windows.

The Qt 6 bindings will now be compiled with the -mno-direct-extern-access flag if --enable-no-direct-extern-access is passed to configure.

In T6696#175866, @antonio-rojas wrote:

This is a Qt issue. Reported at https://bugreports.qt.io/browse/QTBUG-117398

Sure. That's why I wrote that my fix needs to be checked on Windows. And if the regression T6726 wouldn't have blocked the test of my local fix for Windows, ....