json/docs/mkdocs/docs/features/binary_formats/ubjson.md

UBJSON

Universal Binary JSON (UBJSON) is a binary form directly imitating JSON, but requiring fewer bytes of data. It aims to achieve the generality of JSON, combined with being much easier to process than JSON.

!!! abstract "References"

- [UBJSON Website](https://ubjson.org)

Byte layout

UBJSON is built from just a one-byte, human-readable ASCII type marker (for example, i for int8, S for string, [ for array) followed by the payload of that type. Strings and the like are prefixed with their length, which is itself a UBJSON integer value including its marker. Numbers are stored in big-endian byte order. Containers can be written in two ways: as an unoptimized container that is opened with [ or { and closed with ] or }, or as an optimized container that declares a count (#) after the opening marker, optionally preceded by an element type ($); an element type is only allowed together with a count. An optimized container saves the closing marker and, with a type, the per-element markers.

Consider the following JSON value (shown formatted for readability):

{
  "id": 300,
  "name": "json",
  "pi": 3.14,
  "tags": [
    "a",
    "b"
  ]
}

to_ubjson encodes it to 51 bytes (shown in hexadecimal; the encoding contains no whitespace). The colors mark the role of each byte: type/marker length/count/size data.

JSON	Bytes	Meaning
{	7B	object start ({)
"id": 300,	69 02 69 64 49 01 2C	key: length 2 as int8 (i), then UTF-8 bytes; value: int16 (I), big-endian
"name": "json",	69 04 6E 61 6D 65 53 69 04 6A 73 6F 6E	key: length 4; value: string (S), length 4 as int8, UTF-8 bytes
"pi": 3.14,	69 02 70 69 44 40 09 1E B8 51 EB 85 1F	key: length 2; value: float64 (D), big-endian
"tags": [	69 04 74 61 67 73 5B	key: length 4; value: array start ([)
"a",	53 69 01 61	string, length 1
"b"	53 69 01 62	string, length 1
]	5D	array end (])
}	7D	object end (})

Serialization

The library uses the following mapping from JSON values types to UBJSON types according to the UBJSON specification:

JSON value type	value/range	UBJSON type	marker
null	null	null	Z
boolean	true	true	T
boolean	false	false	F
number_integer	-9223372036854775808..-2147483649	int64	L
number_integer	-2147483648..-32769	int32	l
number_integer	-32768..-129	int16	I
number_integer	-128..127	int8	i
number_integer	128..255	uint8	U
number_integer	256..32767	int16	I
number_integer	32768..2147483647	int32	l
number_integer	2147483648..9223372036854775807	int64	L
number_unsigned	0..127	int8	i
number_unsigned	128..255	uint8	U
number_unsigned	256..32767	int16	I
number_unsigned	32768..2147483647	int32	l
number_unsigned	2147483648..9223372036854775807	int64	L
number_unsigned	9223372036854775808..18446744073709551615	high-precision	H
number_float	any value	float64 (float32 if number_float_t is float)	D (d)
string	with shortest length indicator	string	S
array	see notes on optimized format	array	[
object	see notes on optimized format	object	{

!!! success "Complete mapping"

The mapping is **complete** in the sense that any JSON value type can be converted to a UBJSON value.

Any UBJSON output created by `to_ubjson` can be successfully parsed by `from_ubjson`.

!!! warning "Size constraints"

The following values can **not** be converted to a UBJSON value:

  - strings with more than 9223372036854775807 bytes (theoretical)

!!! warning "UTF-8 validation of string values and object keys"

UBJSON's required string encoding is UTF-8. By default (the [`error_handler`](../../api/basic_json/to_ubjson.md)
parameter left at `keep`), `to_ubjson()` writes the bytes of string values and object keys unchanged, even if they
are not valid UTF-8. With `error_handler_t::strict`, it throws
[`type_error.316`](../../home/exceptions.md#jsonexceptiontype_error316) for ill-formed UTF-8 instead;
`replace`/`ignore` sanitize the string. [`JSON_STRICT_BINARY_UTF8`](../../api/macros/json_strict_binary_utf8.md)
makes `strict` the default.

!!! info "Unused UBJSON markers"

The following markers are not used in the conversion:

- `N`: no-op values are never written (they are skipped when reading).
- `C`: single-byte strings are serialized with `S` markers.

!!! info "NaN/infinity handling"

NaN and infinity are written as IEEE 754 float64 values (`D`) and read back unchanged. This differs from
[`dump()`](../../api/basic_json/dump.md), which serializes them to `null`, and from the UBJSON specification, which
encodes infinity as `null` (it does not mention NaN); other UBJSON readers may therefore not accept these values.

!!! info "Optimized formats"

The optimized formats for containers are supported: Parameter `use_size` adds size information to the beginning of a
container and removes the closing marker. Parameter `use_type` further checks whether all elements of a container
have the same type and adds the type marker to the beginning of the container. The `use_type` parameter must only be
used together with `use_size = true`: combining `use_type` with `use_size = false` throws
[`other_error.502`](../../home/exceptions.md#jsonexceptionother_error502) if the value contains a non-empty array,
object, or binary value.

Two elements have the same type if they are written with the same type marker. For example, `[1, 2]` is written with
the type `i`, but `[1, 200]` (`i`, `U`), `[true, false]` (`T`, `F`), and `[1, 2.5]` (`i`, `D`) are not; an empty
container is never given a `$` annotation of its own (an array of empty arrays is still typed `[`).

Note that `use_size = true` alone may result in larger representations - the benefit of this parameter is that the
receiving side is immediately informed on the number of elements of the container.

An array whose type marker is `Z` (null), `T` (true) or `F` (false) stores no payload at all, because the marker
already is the value. Its declared count is therefore the only thing that decides how much memory the receiving side
allocates, and a handful of bytes can describe billions of elements. `from_ubjson` rejects such an array with
[`out_of_range.408`](../../home/exceptions.md#jsonexceptionout_of_range408) when the count exceeds 1,048,576
(`1 << 20`), or when the counts of all such arrays in the input add up to more than 1,048,576 plus 4 for every byte
read so far. `to_ubjson` therefore writes an array of these types in the counted but untyped form when it is longer
than that, or when it would bring the total of the typed arrays written so far above 1,048,576: the `$` type marker
is dropped, the `#` count is kept, and every element is written with its own marker (`[#l 1048577 Z Z ...`), so
everything it produces can be read back. Only input from another encoder, or written by version 3.12.0 and earlier of
this library, can exceed the limit: for example, 1,100 rows of 1,000 `null` values or 210,000 rows of 25, each with
a `$Z` annotation. Version 3.12.0 writes such a row in 5 bytes (6 above 255 elements), so rows of 20 or fewer such
values are never rejected. Objects and untyped counted arrays are not subject to this limit.

An optimized array of type `N` (no-op) is empty whatever its count is, because a no-op is not a value. Its count is
not passed on: `sax_parse` reports `start_array(0)` (versions before 3.13.0 reported the declared count), and the DOM
parser reserves nothing for it.

!!! info "Binary values"

If the JSON data contains the binary type, the value stored is a list of integers, as suggested by the UBJSON
documentation.  In particular, this means that serialization and the deserialization of a JSON containing binary
values into UBJSON and back will result in a different JSON object.

??? example "Example: serialize JSON values to UBJSON, with and without size/type optimization"

```cpp
--8<-- "examples/to_ubjson.cpp"
```

Output:

```c
--8<-- "examples/to_ubjson.output"
```

Deserialization

The library maps UBJSON types to JSON value types as follows:

UBJSON type	JSON value type	marker
no-op	no value, next value is read	N
null	null	Z
false	false	F
true	true	T
float32	number_float	d
float64	number_float	D
uint8	number_unsigned	U
int8	number_integer	i
int16	number_integer	I
int32	number_integer	l
int64	number_integer	L
string	string	S
char	string	C
high-precision	number_unsigned, number_integer, or number_float	H
array	array (optimized values are supported)	[
object	object (optimized values are supported)	{

!!! info "High-precision numbers"

The payload of `H` is a length-prefixed string that is parsed as a JSON number (surrounding whitespace is ignored).
Integer text in the range of `number_unsigned_t` (non-negative) or `number_integer_t` (negative) is stored as such;
any other valid number text (larger or smaller integers, fractions, and exponents such as `1e2`) is stored as
`number_float`, so precision may be lost. Text that is not a valid JSON number raises
[`parse_error.115`](../../home/exceptions.md#jsonexceptionparse_error115), a number outside the range of
`number_float_t` (`1e999`) raises [`out_of_range.406`](../../home/exceptions.md#jsonexceptionout_of_range406).

!!! success "Complete mapping"

Every well-formed UBJSON value can be converted to a JSON value, with these exceptions: a high-precision number
outside the range of `number_float_t` ([`out_of_range.406`](../../home/exceptions.md#jsonexceptionout_of_range406)),
optimized arrays of `Z`, `T`, or `F` with more elements than the limit allows (1,048,576 plus 4 per byte read),
and a count that exceeds the capacity of the container
([`out_of_range.408`](../../home/exceptions.md#jsonexceptionout_of_range408)). High-precision numbers that do not
fit an integer type are stored as floating-point numbers and may lose precision. Ill-formed input is rejected with a
parse error.

!!! info "Duplicate object keys"

If an object contains the same key more than once, the last value is kept (the UBJSON specification does not
address duplicate keys).

!!! info "Type marker of optimized containers"

The type marker after `$` must be one of `Z`, `N`, `T`, `F`, `i`, `U`, `I`, `l`, `L`, `d`, `D`, `H`, `C`, `S`, `[`, or
`{`. It is checked even if the count is 0; any other marker raises
[`parse_error.112`](../../home/exceptions.md#jsonexceptionparse_error112). A counted object typed `N` (`{$N#i1 ...`)
is not supported.

!!! warning "Ill-formed UTF-8 in string values and object keys"

UBJSON's required string encoding is UTF-8, but checking it on read is opt-in: with the
[`error_handler`](../../api/basic_json/from_ubjson.md) parameter left at `keep` (the default), `from_ubjson()`
accepts a string value or object key whose bytes are not valid UTF-8 and hands them back unchanged. Passing
`error_handler_t::strict` makes `from_ubjson()` check and throw
[`parse_error.113`](../../home/exceptions.md#jsonexceptionparse_error113) for ill-formed UTF-8, and
`replace`/`ignore` sanitize the string instead of keeping it. However,
[`dump()`](../../api/basic_json/dump.md) still requires valid UTF-8 and throws
[`type_error.316`](../../home/exceptions.md#jsonexceptiontype_error316) for a value read with the default
`keep` handler, unless an error handler is passed that replaces or ignores the ill-formed bytes. `to_ubjson()`'s
own `error_handler` parameter defaults to `keep` (see above), so such a value is written back unchanged.

??? example "Example: deserialize a JSON value from UBJSON"

```cpp
--8<-- "examples/from_ubjson.cpp"
```

Output:

```json
--8<-- "examples/from_ubjson.output"
```

See also

• Binary formats - overview and comparison of the supported binary formats
• Binary values - how binary values and subtypes map to UBJSON
• to_ubjson - serialize a JSON value to UBJSON
• from_ubjson - parse a JSON value from UBJSON
• JSON_STRICT_BINARY_UTF8 - reject ill-formed UTF-8 in strings
• SAX interface - decode binary formats without building a value