roker
|
c81740b0e6
|
add class nfc_string
|
4 years ago |
|
roker
|
4cd987c24f
|
add German design document for libpEpDatatypes.
|
4 years ago |
|
roker
|
01941f70ad
|
C++17 is minimum requirement now. So replace pEp::string_view with std::string_view etc.
|
4 years ago |
|
Volker Birk
|
2330093cff
|
remove support for C++ < 17 and add more convenience functions
|
4 years ago |
|
roker
|
f5f12e2105
|
add UTF<char16_t>::utf_length(). Be more pedantic about surrogates where they're not allowed.
|
4 years ago |
|
roker
|
4c9ff8de3a
|
Merge remote-tracking branch 'refs/remotes/origin/master'
|
4 years ago |
|
heck
|
f677af8732
|
doc.
|
4 years ago |
|
heck
|
3c8069fa14
|
concept.
|
4 years ago |
|
roker
|
8205b7b9de
|
bundle UTF handling functions as static member functions of a UTF class template to circumvent prohibition of partial function specialization. *sigh*
|
4 years ago |
|
roker
|
c0c278aad9
|
add u32string_view typedef.
|
4 years ago |
|
roker
|
f5b13062c3
|
start to encapsulate the functions in a class, to avoid the prohibition of partial function specialization.
|
4 years ago |
|
roker
|
f1179d91da
|
move operations are 'noexcept'. implement copy_out() for stringpair and message.
|
4 years ago |
|
roker
|
bd7a9b4e93
|
use LANG_VERSION to switch from C++14 (default) to C++17, if necessary
|
4 years ago |
|
heck
|
5349d21158
|
Missing include
|
4 years ago |
|
roker
|
c48088d788
|
add char* strdup_NFC(string_view s). FIXME: The more efficient implementation needs some more internal re-work, that's why this correct but stupid quick'n'dirty implementation is there for now.
|
4 years ago |
|
roker
|
065e730d87
|
unrelated, but still not commited.
|
4 years ago |
|
roker
|
5cdb6e3bdc
|
add additional explicit template instantiations. :-/
|
4 years ago |
|
roker
|
f914202788
|
add identity_list* Wrapper<::identity_list*>::copy_out()
|
4 years ago |
|
roker
|
7973a9638b
|
okay, we allow copying for all Wrapper where copy_out() is implemented.
|
4 years ago |
|
roker
|
d965b3115b
|
allow non-const access to wrapped type. Can be dangerous! USE WITH CAUTION!
|
4 years ago |
|
roker
|
af5720bed0
|
add copy_out() to get a copy of the wrapped value.
|
4 years ago |
|
roker
|
8f7f181899
|
oh, the missing comma confuses libetpan, letting it reject the whole Cc: line. pEpMIME parses 4 entries here. Who is rignt?
|
4 years ago |
|
roker
|
5c399ab820
|
fix Wrapper<::message*>::_new() to use mime_decode_message() instead of a dummy.
|
4 years ago |
|
roker
|
8306edaf09
|
resort the test to see what is working and what not, yet.
|
4 years ago |
|
roker
|
0c82dcc5dc
|
add an EXPECT_NE() to check for nullptr before derefencing it. Avoids segfault. But the test is still red.
|
4 years ago |
|
roker
|
1da02a1eb2
|
add more comprehensive test for message parser. And... it segfaults. :-(
|
4 years ago |
|
heck
|
b536dac898
|
Build: Add local.conf.example
|
4 years ago |
|
roker
|
5ffd66f169
|
fix the unittests for UTF-16 NFC tests
|
4 years ago |
|
roker
|
1d1575e971
|
copy unittest_nfc.cc to unittest_nfc16.cc
|
4 years ago |
|
roker
|
6e1f680ec5
|
use string_view in unittest_nfc. That avoids copying and makes compiler happy, too!
|
4 years ago |
|
roker
|
d3747a8b1f
|
add UTF-16 stuff, and try to re-use as much code as possible.
|
4 years ago |
|
roker
|
1a9c3fc1a8
|
implement parseUtf16().
|
4 years ago |
|
roker
|
b9a0dde181
|
add u16string_view.
|
4 years ago |
|
roker
|
57ee30017d
|
test the wrapper for message.
|
4 years ago |
|
roker
|
2348e3a080
|
inherit from std::iterator<> to add some typedefs that make GNU G++ happy. Unfortunately, std::iterator<> is deprecated in C++17, for whatever reason. :-/
|
4 years ago |
|
roker
|
fea6876528
|
remove warning. :-D
|
4 years ago |
|
roker
|
d36141f3b1
|
forgot Wrapper<pEp_identity*>::_new() and ~::free().
|
4 years ago |
|
roker
|
9278dd9e9b
|
add unittest for identity & identity_list
|
4 years ago |
|
roker
|
8dd301c251
|
fix cut & paste errors, so the compiler is happy again. :)
|
4 years ago |
|
roker
|
9f2d01a2bf
|
add identity & identity_list.
|
4 years ago |
|
roker
|
2e1241bb5f
|
okay, ListWrapper<bloblist_t*, void> is a full specialization, so its members don't need "template<>". Clang++ did not tell me that, g++ did. But still bizarre somehow...
|
4 years ago |
|
roker
|
23d25ef428
|
G++ does not like "non-template functions as friends of class templates. So I change the operator== into member function. :-|
|
4 years ago |
|
roker
|
176d649476
|
add bloblist specialization. DOES NOT COMPILE, YET.
|
4 years ago |
|
roker
|
76ead4b179
|
add identity.cc I had fogotten.
|
4 years ago |
|
roker
|
801f23a746
|
uuuh... a const_cast would silent the compiler. hmm.
|
4 years ago |
|
roker
|
392662b981
|
unittest to show how the intended usage will be.
|
4 years ago |
|
roker
|
cbbd03ae01
|
okay, if I change the value_type from 'char*' to 'const char*' I got other troubles there... *sigh*
|
4 years ago |
|
roker
|
16cf6ce9c3
|
add stringlist wrapper.
|
4 years ago |
|
roker
|
b8529ff268
|
change return type.
|
4 years ago |
|
roker
|
39e8dd2d10
|
Yay, all tests are green now! \o/
|
4 years ago |
