Hi,
this patch adds testing vcardparser to cmake, so that the code compiles and
is usable.
1.
I did not add check target as in the original automake, but rather added a
test.sh script that can be executed from within vcardparser directory.
2.
The patch also includes the patch by Fat-Zer, which works for the testing
tools, but as I reported before, it results in utf8 converted chars each
time I load kontact (each time vcf file is read (and perhaps written)).
To me it looks like parseVCards is good with that patch but there is a
problem elsewhere in the chain - write??(see 4).
3.
I added two tests in utf8
tests/vcard8.vcf (german)
tests/vcard9.vcf (cyrillic)
and have no idea why vcard9 fails on testread2 and why the Name is not read
properly, when in vcard8 it passes the test
4.
I then added a testwrite2 file that exposes the problem with the utf8 to
utf8 conversion and will look into it next.
all together we are 1 step further
[to be continued]
regards
PS:
Unfortunately when building debian packages it adds
X-Ubuntu-Gettext-Domain=desktop_tdelibs-trinity
to each desktop source file, which is really bad.
I quick workaround it like following
find . -name '*.desktop' -exec grep X-Ubuntu-Gettext-Domain {} \; -print |
grep '^\./' | while read line; do echo ">>>" $line;
perl -pi -e 's:X-Ubuntu-Gettext-Domain=desktop_tdelibs-trinity::g' $line;
done