This package provides an efficient implementation of the Concise
Binary Object Representation (CBOR), as specified by
RFC 7049.
If you are looking for a library for serialisation of Haskell values,
have a look at the serialise package, which is
built upon this library.
An implementation of the standard bijection between CBOR and JSON is
provided by the cborg-json package. Also see
cbor-tool for a convenient command-line utility
for working with CBOR data.
This package was formerly known as binary-serialise-cbor.
Changes
Revision history for cborg
0.2.10.0
Return TypeUInt64 and TypeNInt64 from tokenType when appropriate, fixing #324
Don’t rely on MonadFail Gen
Support for GHC 9.8
0.2.9.0
Fix build with base >= 4.17 on platforms without unaligned memory operations.
Fix Eq, Ord, Show and IsList instances for SlicedByteArray when offset is present
Fix toBuilder and encodeByteArray for SlicedByteArray when offset is present
0.2.8.0 – 2022-09-24
Support GHC 9.4
Fix compatibility with primitive 0.7.4.0
Drop GHC 8.0 and 8.2 support
Support aeson 2.1
0.2.3.1 – 2020-05-10
Bounds updates for GHC 8.10
Decoder is now a newtype.
0.2.2.1 – 2019-12-29
Testsuite updates for GHC 8.8
0.2.2.0 – 2019-07-31
Add peekByteOffset for the current ByteOffset in the input byte sequence.
By keeping track of the byte offsets before and after decoding asubterm
(a pattern captured by decodeWithByteSpan) and if the overall input data
is retained then this is enables later retrieving the span of bytes for
the subterm.
Add encodePreEncoded function. This allows pre-encoded CBOR data to be
included into an Encoding. This is useful in cases where one has known-valid
encoded CBOR data, e.g. on disk, that you want to include into a larger CBOR
data stream. This makes it possible in such cases to avoid decoding and
re-encoding.
Improved test suite property coverage. We now have property coverering most
parts of a commuting diagram, which gives more confidence about what are
the right properties to test and what is enough.
Improved test coverage for decoding non-canonical terms
Fix a bug in the checks in the canonical decoding of float16 NaNs. There
are multiple bit representations of NaNs and while we were checking this
correctly for float32 and float64 we were not checking this correctly for
the float16 encoding.
Improved test coverage for special float values. We now have pretty
comprehensive coverage of round-tripping properties for special float
values, +/- infinity and non-canonical NaNs.
Improved the structure of the test suite
Use new GHC primitives castWord{32ToFloat,64ToDouble} rather than home grown
Support GHC built with integer-simple
Support GHC 8.8
0.2.1.0 – 2018-10-11
Bounds bumps and GHC 8.6 compatibility
0.2.0.0 – 2017-11-30
Improved robustness of non-UTF-8 strings
Add encoders and decoders for ByteArray
Add decoding variants that check for canonical encodings
