summaryrefslogtreecommitdiffstats
path: root/test/tests/spec/2_6_5-uid.vcf
diff options
context:
space:
mode:
authorGravatar Nikias Bassen2013-02-20 18:03:42 +0100
committerGravatar Nikias Bassen2013-02-20 18:03:42 +0100
commitd86c2a34bdaa37dfd07c8ded8a19afbfc1066949 (patch)
treece99c4eb7efd55459f4655356b9b4b3bca720e37 /test/tests/spec/2_6_5-uid.vcf
parentf70f1876b9287dc21302ebce4d5374291087fc2b (diff)
downloadlibvformat-d86c2a34bdaa37dfd07c8ded8a19afbfc1066949.tar.gz
libvformat-d86c2a34bdaa37dfd07c8ded8a19afbfc1066949.tar.bz2
Fix base64 decoding for multi-line data
The base64 decoder assumed that all data passed to it is a complete base64 blob. However due to the nature of different encoding schemes vformat data can also be spanned over multiple lines, usually prefixed by one or more spaces. The problem arises when a base64 line is broken in a non-by-4-dividable manner. We fix this by storing a remainder if the base64 data is not dividable by 4 and prepend it to the next decode operation.
Diffstat (limited to 'test/tests/spec/2_6_5-uid.vcf')
0 files changed, 0 insertions, 0 deletions