CBOR string

[arnaudbrejeon]

I received some CBOR files that contain strings (major type 2) and not UTF-8 strings(major type 3). It looks like parsing fails. Is it on purpose?

[nlohmann]

This is currently not supported, as we only support UTF-8 as encoding. I we would accept byte strings, we would also need to support additional encodings.

While this would just be a programming exercise (it seems that only base64url, base64, base16, and CBOR are mentioned in RFC 7049), it is currently not a priority for this library. However, PRs are always welcome :-)

[azadkuh]

ATM, it's possible to misuse the json::to_cbor() / json::to_msgpack() as:

const std::string binary = load_some_binary_like_a_jpeg_file();
json src = {"data", binary};

const auto bin_serialization = json::to_cbor(src); // json::to_msgpack()
auto dest = json::from_cbor(bin_serialization);    // json::from_msgpack()
// now: dest["data"].get<std::string>() == binary
// if: serialization/deserialization by this library

it may sound weird, but as long as all modern std:string implementations can handle binary data as well as utf8 text, what if to_cbor()/from_cbor() consider all strings as byte strings (raw/binary)? then both binary/text are covered in msgpack/cbor serialization.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[nlohmann]

Binary strings are now there, see #1662.