Releases: dampcake/bencode
v1.4.2
What's Changed
- Bump GHA actions and target supported java versions by @dampcake in #19
- Add test for debian bttracker by @dampcake in #18
- Bump junit to 4.13.2 and plugins to latest by @dampcake in #20
- Fix resource leaks in stream usage with try-with-resources by @dampcake in #21
- Cleanup test code by @dampcake in #22
- Add missing test case from coverage report by @dampcake in #23
- Fix README.md build badge by @dampcake in #25
- Remove redundant check from readObject condition by @dampcake in #24
- Remove encode conditions by using a consumer pattern by @dampcake in #26
- Update to use new Maven Central Portal by @dampcake in #27
Full Changelog: bencode-1.4.1...bencode-1.4.2
v1.4.1
v1.4
Fix #12 - Properly encode multibyte codepoints. Thanks @manuel-sugawara for the fix!
v1.3.2
v1.3.1
Fix for #7, thanks @lfcnassif for filing.
Allow many more types of numbers to be parsed, but will still only write/return integers (Long
s) as per the bencode spec.
v1.3
Fix for #6, thanks @zzyandzzy for filing.
New constructor argument for Bencode
and BencodeInputStream
a boolean
called useBytes
. This is especially useful when working with torrent files as not all of the byte strings can be converted to String
s without data loss.
If useBytes
is true
:
List
andMap
values (not keys inMaps
) will not be coerced into aString
value but instead be kept as binary data in aByteBuffer
.ByteBuffer
is now understood by theBencodeOutputStream
and will be added as byte string data.- New methods added for writing/reading ByteBuffers added.
If useBytes
is false
:
List
andMap
values will continue to be coerced intoStrings
.
v1.2.2
v1.2.1
Speedup encoding by using StringBuilder instead of String.format.
See: #4
Thanks to @EonTechnology for the fix.
v1.2
Fix encoding on locales with non-arabic digits.
See: #3
Thanks to @EonTechnology for the fix.