though much of this could be moved into the base class as there's a lot
of copy-paste action going on.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@768992 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
preferring a comment without an empty description.
FEATURE:137553
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@768865 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
ISO-8859-1 are in fact ISO-8859-1 values (assuming they were passed into TagLib
properly) and if not automatically switches those frames to UTF8.
FEATURE:90635
CCBUG:90635
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@768857 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
happen when some of the characters are null. This behavior can still be customized
via the StringHandler.
CCBUG:99149
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@768597 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
- '\xFF\xFE' and '\xFE\xFF' are valid, but empty, UTF-16 strings -- don't print a warning
- print a warning on strings that don't start with a valid BOM
BUG:156076
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@764385 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
Patch by Vincent Castellano
CCMAIL:surye80@gmail.com
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@747837 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
The code is inside a '#ifndef NO_ITUNES_HACKS' block, so I hope it's ok to add it.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@743534 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
out the configure check is fixed. Once the configure check is fixed this can be
commented back in.
CCMAIL:lalinsky@gmail.com
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@735161 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
This needs to be fixed to fail gracefully for users without CppUnit available.
CCMAIL:lalinsky@gmail.com
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@735160 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
This is last of the three formats implemented for libtunepimp, now used also by various players.
BUG:114982
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@735116 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
According to the spec, the identifier contains arbitrary binary data (which
includes \0). That means splitting it on \0 and checking if we have only
two parts is wrong, because it rejects valid frames.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@735109 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
In case where the pattern has the same length as the bytevector, it should
degrade to A == B, not just return "not found" (e.g. A.find(A) should
return 0, not -1).
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@735107 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
caused by using uninitialized pointer. (BUG:151078)
- Make Frame::readStringField to actually read the string field.
- Fixed parsing of APIC frames -- there is one-byte type between mime type
and description. The code worked only thanks to the previous bug.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@735035 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
