That's only one part of the fix. But we also want to test the Fancy date format: see D21442.

This patch will only work if kdepim-addons run make install before running the tests. This issue is fixed by patch D21444 especially the part that diff on messageviewer/src/antispam/antispamconfig.cpp.

This is not the way it should be fixed. I already added a patch that fixes that issue (but havn't commited it yet).

I also like your suggestion to use only "More >>" and "Fewer <<".

In D21448#485183, @dvratil wrote:

So why do you want to remove it in the first place, if you know there's a code in a header style plugin that uses those methods? And what is your proposed fix for kdepim-addons?

In D21448#471111, @dvratil wrote:

I assume you checked that everything still builds? I see some reference to this method in kdepim-addons?

In D21447#471109, @dvratil wrote:

What does resentFrom actually represent? The sender of the vcard, if the vcard is not of the sender?

It looks fine now. The two open issues will be addressed separately.

overall I like this patch - that is great an brings postgresql more to a usable db interface.

@security-team: should I split the commits code/test so they are easier to backport? On the other side adding the tests also to old code makes it more clear that this is fixed.

The review is getting a little bit messy, because of all those new test data.

• Add test cases for a Decryption Oracle based on PGP inline.
• TemplateParser: test htmlReply.
• Rename convertedHtmlContent-> convertedHtmlContent for reply_Plain
• Make unexpected data leak harder via reply.

• Add test for a normal encrypted message, to make sure decrpytion works.
• templateparserjobtest: HTML mime parts are handled differently
• WIP: Decryption Oracle is also possible with PGP inline.
• TemplateParser: test htmlReply.
• Rename convertedHtmlContent-> convertedHtmlContent for reply_Plain
• Make unexpected data leak harder via reply.

Please add a simple test case to make sure we catch those changes by tests.

seems good.

• MimeTreeParser returns wrong content for inline mesasges with multiple encrypted blocks.
• TemplateParser: test htmlReply.

• Add test for a normal encrypted message, to make sure decryption works.
• templateparserjobtest: HTML mime parts are handled differently
• WIP: Decryption Oracle is also possible with PGP inline.

In D20757#454735, @vkrause wrote:

This makes sense I think (and should go in IMHO), but I'm not sure if it is complete to cover all of bug 404698.

• Do we need to be concerned about forwarding (which does include attachments AFAIK)?

make sense, yes.

In D20757#454566, @aacid wrote:

Sorry if this is a stupid question :D

But does this test make sense?

I mean how is the test supposed to decrypt the encrypted part of it? How does it know the keys/password/whatever to decrypt it?

This is git logic, that it defines this as a copy. And I think git it more or less right, as we can't move every tests because some tests need MessageViewer::BufferedHtmlWriter. Only

update description.

In D20757#454549, @dfaure wrote:

So these pass? There's no bug?

• pidgin
  • use the proxy settings from the preferences and make use every account is taking the general proxy setting.
  • nice thing about this, that you can enable/disable tor on an account base.

@vkrause: I havn't found the tool you mentioned "tcpconnect"?

and for a kde wide search:
https://lxr.kde.org/search?_filestring=&_string=setStrictTransportSecurityEnabled