Bug 170803

Summary: Add content-based classification for binary data, too.
Product: [Applications] kleopatra Reporter: Marc Mutz <mutz>
Component: generalAssignee: Andre Heinecke <aheinecke>
Status: RESOLVED FIXED    
Severity: normal CC: emanuel, kdepim-bugs
Priority: NOR    
Version: 2.0.0-rc5   
Target Milestone: ---   
Platform: Compiled Sources   
OS: Unspecified   
Latest Commit: http://commits.kde.org/kleopatra/0499cc6822fab0f3c2036954ecebeb788fae5962 Version Fixed In:
Sentry Crash Report:

Description Marc Mutz 2008-09-10 15:32:49 UTC 
Version:           2.0.0-rc5 (using Devel)
Installed from:    Compiled sources

The lack of it is a problem for .p7s files in particular: If both A and A.p7s are present, we cannot determine whether A.p7s is a detached signature over A, or an opaque signature made over A, with A left-over.

There's some code in GpgOL, and on-going work in gpg, to classify binary content. Once done, we should make use of that to get rid of the heuristics we now have.
Comment 1 Emanuel Schütze 2013-05-08 10:12:10 UTC 
Andre, can you check this issue with current Kleopatra?
Comment 2 Andre Heinecke 2016-07-14 09:45:25 UTC 
Git commit 0499cc6822fab0f3c2036954ecebeb788fae5962 by Andre Heinecke.
Committed on 14/07/2016 at 09:17.
Pushed by aheinecke into branch 'master'.

Handle working gpgme_data_identify in classify

Now that we might have a working identify version we
want to prefer that over our own detection if it is available.

M  +54   -25   src/utils/classify.cpp

http://commits.kde.org/kleopatra/0499cc6822fab0f3c2036954ecebeb788fae5962