Specifically make sure that we don't actually allocate a buffer for a read that
extends beyond the end of the file.
BUG:101401
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@438035 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
that takes a maximum number of values. This should probably be used more widely in
places where the max is known ahead of time for this to be a more useful fix than nailing
just this special case. Anyway, fixes the bug.
BUG:103622
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@438030 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
update the configure.in.in in the root...
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@438007 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
This is generally much better thought out than the current method of creating
FileRefs using the native mime system and also allows for extension to additional
file formats.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@437382 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
Make it possible to save files using the save() with a tag mask
without stripping the tags that do not fit the mask. I still need
to do a little testing on this to make sure that the offsets are
still computed correctly.
BUG:96460
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@437380 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
should be functional enough to hold over to the next binary
incompatible change and has a minimum of differences to the previous
API.
I did a slightly nasty hack so that the API docs will see just methods
with an optional argument, but those are actually two separate methods
(for BC).
Brian, please feel free to take a look at this prior to 1.4 going out.
BUG:107025
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@437115 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
to working. The more complicated fix for #107025 will
follow shortly.
BUG:95545
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@437091 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
UnknownFrame and set the flag to discard it on write.
As a special bonus this meant implementing the discard-on-tag-alter
flag.
BUG:100515
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@415248 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
Almost the same as a patch from Clemens Fuchslocher
CCMAIL:clemens@allesdurcheinander.de
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@415211 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
at the end of the list. Thanks to Gunnar Roth for catching this one.
BUG:96926
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@378404 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
there are actually more items that there actually are.
BUG:92028
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@359382 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
find it online and it's poorly written once found.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@359360 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
the appropriate seek location. I'm pretty sure that this is what was causing
Ogg length information to not work on x86_64. Can you confirm Hamish?
BUG:86806
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@358653 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
to an integer before trying to convert it and (b) make sure that there's data
in an APE::Item before trying to parse it.
BUG:92028
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@358637 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
loop through lists rather than while, fix bracket style.
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@358627 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
is if a bogus frame is hit it may try to parse it as an UnknownFrame rather
than just discarding it.
CCMAIL:Espen Tveit <espen.tveit@gmail.com>
git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@355686 283d02a7-25f6-0310-bc7c-ecb5cbfe19da
