Introduction#

The data types in AES70 are:

Basic data types (e.g. OcaFloat32, OcaString, OcaList<T>)
Enums (e.g. OcaStatus) and bitsets
Structure types that are compositions of other types

For each data type defined in the AES70 standard, this library defines a corresponding C++ type using standard library types. For example:

OcaFloat32 is a typedef for float
OcaString is a typedef for std::string
OcaList<T> is a typedef for std::vector<T>

The full list of data types can be found in the files included from include/aes70/types.hpp.

OCP.1 Encoding#

AES70 uses a binary encoding called OCP.1. OCP.1 defines how data types in return values and arguments are encoded for transfer over the network.

The implementation of OCP.1 is defined in the aes70::OCP1 namespace. For instance, the implementation of the encoding of OcaString is defined as aes70::OCP1::OcaString.

One important detail about these encoder classes is that they can decode to and encode more than just the standard types. This means, for instance, that the encoder for OcaString supports not only std::string, but also several other string-like types such as std::string_view.

Encoding happens in the following situations:

Return value of control methods in devices (see here)
Arguments of controller classes (see here)
When sending property changed events

Decoding happens in the following situations:

Argument of control methods in devices (see here)
Return values of controller classes (see here)
When receiving property changed events in controllers

As a consequence, return values and arguments of AES70 methods are flexible.

The following sections describe the encoding and decoding possibilities in detail.

To understand in detail which data types are supported, look at the encoder implementations in include/aes70/OCP1/encoders.hpp and especially the signatures of the methods

    uint8_t *encode_to(uint8_t *dst, const T &x);
    const uint8_t *decode_from(const uint8_t *src, T &x);

There are some special types which any encoder supports:

aes70::OCP1::ignore. When encoding, this results in the trivial encoding of the corresponding type (e.g. 0 for integers, the empty list for OcaList<T>, the empty string, etc.). When used during decoding, the encoder decodes and discards the result. This is useful for unused arguments in OCA method implementations.
aes70::buffer_view. This class is similar to std::string_view; it represents a pointer and a length. When used as an argument during decoding, it will contain the part of the network buffer that contains the encoding of the corresponding parameter. This can be useful in situations where custom decoding logic needs to be implemented.

Integer types#

All integer types (OcaInt8, OcaUint16, etc.) can decode into and encode any integer type.

Note

Keep in mind that when decoding into a smaller integer type, the value may not fit. For example, when an AES70 method is defined to receive a value of type OcaInt64, using an int16_t could result in a different value. It is therefore recommended to use the integer of the correct size, e.g. int64_t.

OcaFloat32, OcaFloat64#

These encode and decode into the corresponding floating point types float and double. When decoding values, they are validated using the macro AES70_IS_VALID_FLOAT(x), which by default will not allow NaN.