UNOFFICIAL - This is a draft for a not-yet-released version 2.3 of Jansson for Big Number Extensions

API Reference

Preliminaries

All declarations are in jansson.h, so it’s enough to

#include <jansson.h>

in each source file.

All constants are prefixed with JSON_ (except for those describing the library version, prefixed with JANSSON_). Other identifiers are prefixed with json_. Type names are suffixed with _t and typedef‘d so that the struct keyword need not be used.

Library Version

The Jansson version is of the form A.B.C, where A is the major version, B is the minor version and C is the micro version. If the micro version is zero, it’s omitted from the version string, i.e. the version string is just A.B.

When a new release only fixes bugs and doesn’t add new features or functionality, the micro version is incremented. When new features are added in a backwards compatible way, the minor version is incremented and the micro version is set to zero. When there are backwards incompatible changes, the major version is incremented and others are set to zero.

The following preprocessor constants specify the current version of the library:

JANSSON_VERSION_MAJOR, JANSSON_VERSION_MINOR, JANSSON_VERSION_MICRO
Integers specifying the major, minor and micro versions, respectively.
JANSSON_VERSION
A string representation of the current version, e.g. "1.2.1" or "1.3".
JANSSON_VERSION_HEX

A 3-byte hexadecimal representation of the version, e.g. 0x010201 for version 1.2.1 and 0x010300 for version 1.3. This is useful in numeric comparisions, e.g.:

#if JANSSON_VERSION_HEX >= 0x010300
/* Code specific to version 1.3 and above */
#endif

Value Representation

The JSON specification (RFC 4627) defines the following data types: object, array, string, number, boolean, and null. JSON types are used dynamically; arrays and objects can hold any other data type, including themselves. For this reason, Jansson’s type system is also dynamic in nature. There’s one C type to represent all JSON values, and this structure knows the type of the JSON value it holds.

json_t

This data structure is used throughout the library to represent all JSON values. It always contains the type of the JSON value it holds and the value’s reference count. The rest depends on the type of the value.

Objects of json_t are always used through a pointer. There are APIs for querying the type, manipulating the reference count, and for constructing and manipulating values of different types.

Unless noted otherwise, all API functions return an error value if an error occurs. Depending on the function’s signature, the error value is either NULL or -1. Invalid arguments or invalid input are apparent sources for errors. Memory allocation and I/O operations may also cause errors.

Type

The type of a JSON value is queried and tested using the following functions:

enum json_type

The type of a JSON value. The following members are defined:

JSON_OBJECT
JSON_ARRAY
JSON_STRING
JSON_INTEGER
JSON_BIGINTEGER
JSON_REAL
JSON_BIGREAL
JSON_TRUE
JSON_FALSE
JSON_NULL

These correspond to JSON object, array, string, number, boolean and null. A number is represented by one of the values of type JSON_INTEGER, JSON_BIGINTEGER, JSON_REAL, or JSON_BIGREAL. A true boolean value is represented by a value of the type JSON_TRUE and false by a value of the type JSON_FALSE.

The two big-number types, JSON_BIGINTEGER and JSON_BIGREAL, are used with add-on extensions to allow arbitrarily large or high-precision numbers to be represented. Unless an extension is provided and enabled using the big number extension API, these types will not be used.

int json_typeof(const json_t *json)

Return the type of the JSON value (a json_type cast to int). json MUST NOT be NULL. This function is actually implemented as a macro for speed.

json_is_object(const json_t *json)
json_is_array(const json_t *json)
json_is_string(const json_t *json)
json_is_integer(const json_t *json)
json_is_biginteger(const json_t *json)
json_is_real(const json_t *json)
json_is_bigreal(const json_t *json)
json_is_true(const json_t *json)
json_is_false(const json_t *json)
json_is_null(const json_t *json)

These functions (actually macros) return true (non-zero) for values of the given type, and false (zero) for values of other types and for NULL.

json_is_number(const json_t *json)

Returns true for values of types JSON_INTEGER and JSON_REAL; and false for other types and for NULL.

json_is_bignumber(const json_t *json)

Returns true for values of types JSON_BIGINTEGER and JSON_BIGREAL, and false for other types and for NULL.

json_is_anynumber(const json_t *json)

Returns true for values of types JSON_INTEGER, JSON_BIGINTEGER, JSON_REAL and JSON_BIGREAL, and false for other types and for NULL.

json_is_anyinteger(const json_t *json)

Returns true for values of types JSON_INTEGER and JSON_BIGINTEGER, and false for other types and for NULL.

json_is_anyreal(const json_t *json)

Returns true for values of types JSON_REAL and JSON_BIGREAL, and false for other types and for NULL.

json_is_boolean(const json_t *json)

Returns true for types JSON_TRUE and JSON_FALSE, and false for values of other types and for NULL.

Reference Count

The reference count is used to track whether a value is still in use or not. When a value is created, it’s reference count is set to 1. If a reference to a value is kept (e.g. a value is stored somewhere for later use), its reference count is incremented, and when the value is no longer needed, the reference count is decremented. When the reference count drops to zero, there are no references left, and the value can be destroyed.

The following functions are used to manipulate the reference count.

json_t *json_incref(json_t *json)

Increment the reference count of json if it’s not non-NULL. Returns json.

void json_decref(json_t *json)

Decrement the reference count of json. As soon as a call to json_decref() drops the reference count to zero, the value is destroyed and it can no longer be used.

Functions creating new JSON values set the reference count to 1. These functions are said to return a new reference. Other functions returning (existing) JSON values do not normally increase the reference count. These functions are said to return a borrowed reference. So, if the user will hold a reference to a value returned as a borrowed reference, he must call json_incref(). As soon as the value is no longer needed, json_decref() should be called to release the reference.

Normally, all functions accepting a JSON value as an argument will manage the reference, i.e. increase and decrease the reference count as needed. However, some functions steal the reference, i.e. they have the same result as if the user called json_decref() on the argument right after calling the function. These functions are suffixed with _new or have _new_ somewhere in their name.

For example, the following code creates a new JSON array and appends an integer to it:

json_t *array, *integer;

array = json_array();
integer = json_integer(42);

json_array_append(array, integer);
json_decref(integer);

Note how the caller has to release the reference to the integer value by calling json_decref(). By using a reference stealing function json_array_append_new() instead of json_array_append(), the code becomes much simpler:

json_t *array = json_array();
json_array_append_new(array, json_integer(42));

In this case, the user doesn’t have to explicitly release the reference to the integer value, as json_array_append_new() steals the reference when appending the value to the array.

In the following sections it is clearly documented whether a function will return a new or borrowed reference or steal a reference to its argument.

Circular References

A circular reference is created when an object or an array is, directly or indirectly, inserted inside itself. The direct case is simple:

json_t *obj = json_object();
json_object_set(obj, "foo", obj);

Jansson will refuse to do this, and json_object_set() (and all the other such functions for objects and arrays) will return with an error status. The indirect case is the dangerous one:

json_t *arr1 = json_array(), *arr2 = json_array();
json_array_append(arr1, arr2);
json_array_append(arr2, arr1);

In this example, the array arr2 is contained in the array arr1, and vice versa. Jansson cannot check for this kind of indirect circular references without a performance hit, so it’s up to the user to avoid them.

If a circular reference is created, the memory consumed by the values cannot be freed by json_decref(). The reference counts never drops to zero because the values are keeping the references to each other. Moreover, trying to encode the values with any of the encoding functions will fail. The encoder detects circular references and returns an error status.

True, False and Null

These values are implemented as singletons, so each of these functions returns the same value each time.

json_t *json_true(void)
Return value: New reference.

Returns the JSON true value.

json_t *json_false(void)
Return value: New reference.

Returns the JSON false value.

json_t *json_null(void)
Return value: New reference.

Returns the JSON null value.

String

Jansson uses UTF-8 as the character encoding. All JSON strings must be valid UTF-8 (or ASCII, as it’s a subset of UTF-8). Normal null terminated C strings are used, so JSON strings may not contain embedded null characters. All other Unicode codepoints U+0001 through U+10FFFF are allowed.

json_t *json_string(const char *value)
Return value: New reference.

Returns a new JSON string, or NULL on error. value must be a valid UTF-8 encoded Unicode string.

json_t *json_string_nocheck(const char *value)
Return value: New reference.

Like json_string(), but doesn’t check that value is valid UTF-8. Use this function only if you are certain that this really is the case (e.g. you have already checked it by other means).

const char *json_string_value(const json_t *string)

Returns the associated value of string as a null terminated UTF-8 encoded string, or NULL if string is not a JSON string.

The retuned value is read-only and must not be modified or freed by the user. It is valid as long as string exists, i.e. as long as its reference count has not dropped to zero.

int json_string_set(const json_t *string, const char *value)

Sets the associated value of string to value. value must be a valid UTF-8 encoded Unicode string. Returns 0 on success and -1 on error.

int json_string_set_nocheck(const json_t *string, const char *value)

Like json_string_set(), but doesn’t check that value is valid UTF-8. Use this function only if you are certain that this really is the case (e.g. you have already checked it by other means).

Number

The JSON specification only contains one numeric type, “number”. The C programming language has distinct types for integer and floating-point numbers, so for practical reasons Jansson also has distinct types for the two. They are called “integer” and “real”, respectively.

Additionally, Jansson provides an extension API to allow external packages to be used to represent arbitrarily large integers or arbitrarily high-precision real numbers. So a JSON number may be represented by any of four different C types.

For more information, see RFC Conformance.

json_int_t

This is the C type that is used to store JSON integer values in the absence of a big number extension. It represents the widest integer type available on your system. In practice it’s just a typedef of long long if your compiler supports it, otherwise long.

Usually, you can safely use plain int in place of json_int_t, and the implicit C integer conversion handles the rest. Only when you know that you need the full 64-bit range, you should use json_int_t explicitly.

double

The C type double is used to store JSON real values in the absence of a big number extension.

Note that the C type long double is only supported by using the big number extension API.

json_bigz_t
json_bigz_const_t

This is an unspecified C pointer type used to reference JSON integer values that are stored using an external big number package. The _const_t type is the same only it is used to point to a constant value.

json_bigr_t
json_bigr_const_t

This is an unspecified C pointer type used to reference JSON real values that are stored using an external big number package. The _const_t type is the same only it is used to point to a constant value.

All of the functions and macros for dealing with the big number types json_bigz_t and json_bigr_t are documented separately in the section Big Number Extensions.

JSON_INTEGER_IS_LONG_LONG

This is a preprocessor variable that holds the value 1 if json_int_t is long long, and 0 if it’s long. It can be used as follows:

#if JSON_INTEGER_IS_LONG_LONG
/* Code specific for long long */
#else
/* Code specific for long */
#endif

JSON_INTEGER_FORMAT

This is a macro that expands to a printf() conversion specifier that corresponds to json_int_t, without the leading % sign, i.e. either "lld" or "ld". This macro is required because the actual type of json_int_t can be either long or long long, and printf() requires different length modifiers for the two.

Example:

json_int_t x = 123123123;
printf("x is %" JSON_INTEGER_FORMAT "\n", x);
json_t *json_integer(json_int_t value)
Return value: New reference.

Returns a new JSON integer, or NULL on error.

json_int_t json_integer_value(const json_t *integer)

Returns the associated value of integer, or 0 if json is not a JSON integer.

int json_integer_set(const json_t *integer, json_int_t value)

Sets the associated value of integer to value. Returns 0 on success and -1 if integer is not a JSON integer.

json_t *json_real(double value)
Return value: New reference.

Returns a new JSON real, or NULL on error.

double json_real_value(const json_t *real)

Returns the associated value of real, or 0.0 if real is not a JSON real.

int json_real_set(const json_t *real, double value)

Sets the associated value of real to value. Returns 0 on success and -1 if real is not a JSON real.

In addition to the functions above, there’s a common query function for integers and reals:

double json_number_value(const json_t *json)

Returns the associated value of the JSON integer or JSON real json, cast to double regardless of the actual type. If json is neither JSON real nor JSON integer, 0.0 is returned.

Array

A JSON array is an ordered collection of other JSON values.

json_t *json_array(void)
Return value: New reference.

Returns a new JSON array, or NULL on error. Initially, the array is empty.

size_t json_array_size(const json_t *array)

Returns the number of elements in array, or 0 if array is NULL or not a JSON array.

json_t *json_array_get(const json_t *array, size_t index)
Return value: Borrowed reference.

Returns the element in array at position index. The valid range for index is from 0 to the return value of json_array_size() minus 1. If array is not a JSON array, if array is NULL, or if index is out of range, NULL is returned.

int json_array_set(json_t *array, size_t index, json_t *value)

Replaces the element in array at position index with value. The valid range for index is from 0 to the return value of json_array_size() minus 1. Returns 0 on success and -1 on error.

int json_array_set_new(json_t *array, size_t index, json_t *value)

Like json_array_set() but steals the reference to value. This is useful when value is newly created and not used after the call.

int json_array_append(json_t *array, json_t *value)

Appends value to the end of array, growing the size of array by 1. Returns 0 on success and -1 on error.

int json_array_append_new(json_t *array, json_t *value)

Like json_array_append() but steals the reference to value. This is useful when value is newly created and not used after the call.

int json_array_insert(json_t *array, size_t index, json_t *value)

Inserts value to array at position index, shifting the elements at index and after it one position towards the end of the array. Returns 0 on success and -1 on error.

int json_array_insert_new(json_t *array, size_t index, json_t *value)

Like json_array_insert() but steals the reference to value. This is useful when value is newly created and not used after the call.

int json_array_remove(json_t *array, size_t index)

Removes the element in array at position index, shifting the elements after index one position towards the start of the array. Returns 0 on success and -1 on error. The reference count of the removed value is decremented.

int json_array_clear(json_t *array)

Removes all elements from array. Returns 0 on sucess and -1 on error. The reference count of all removed values are decremented.

int json_array_extend(json_t *array, json_t *other_array)

Appends all elements in other_array to the end of array. Returns 0 on success and -1 on error.

Object

A JSON object is a dictionary of key-value pairs, where the key is a Unicode string and the value is any JSON value.

json_t *json_object(void)
Return value: New reference.

Returns a new JSON object, or NULL on error. Initially, the object is empty.

size_t json_object_size(const json_t *object)

Returns the number of elements in object, or 0 if object is not a JSON object.

json_t *json_object_get(const json_t *object, const char *key)
Return value: Borrowed reference.

Get a value corresponding to key from object. Returns NULL if key is not found and on error.

int json_object_set(json_t *object, const char *key, json_t *value)

Set the value of key to value in object. key must be a valid null terminated UTF-8 encoded Unicode string. If there already is a value for key, it is replaced by the new value. Returns 0 on success and -1 on error.

int json_object_set_nocheck(json_t *object, const char *key, json_t *value)

Like json_object_set(), but doesn’t check that key is valid UTF-8. Use this function only if you are certain that this really is the case (e.g. you have already checked it by other means).

int json_object_set_new(json_t *object, const char *key, json_t *value)

Like json_object_set() but steals the reference to value. This is useful when value is newly created and not used after the call.

int json_object_set_new_nocheck(json_t *object, const char *key, json_t *value)

Like json_object_set_new(), but doesn’t check that key is valid UTF-8. Use this function only if you are certain that this really is the case (e.g. you have already checked it by other means).

int json_object_del(json_t *object, const char *key)

Delete key from object if it exists. Returns 0 on success, or -1 if key was not found. The reference count of the removed value is decremented.

int json_object_clear(json_t *object)

Remove all elements from object. Returns 0 on success and -1 if object is not a JSON object. The reference count of all removed values are decremented.

int json_object_update(json_t *object, json_t *other)

Update object with the key-value pairs from other, overwriting existing keys. Returns 0 on success or -1 on error.

The following functions implement an iteration protocol for objects, allowing to iterate through all key-value pairs in an object. The items are not returned in any particular order, as this would require sorting due to the internal hashtable implementation.

void *json_object_iter(json_t *object)

Returns an opaque iterator which can be used to iterate over all key-value pairs in object, or NULL if object is empty.

void *json_object_iter_at(json_t *object, const char *key)

Like json_object_iter(), but returns an iterator to the key-value pair in object whose key is equal to key, or NULL if key is not found in object. Iterating forward to the end of object only yields all key-value pairs of the object if key happens to be the first key in the underlying hash table.

void *json_object_iter_next(json_t *object, void *iter)

Returns an iterator pointing to the next key-value pair in object after iter, or NULL if the whole object has been iterated through.

const char *json_object_iter_key(void *iter)

Extract the associated key from iter.

json_t *json_object_iter_value(void *iter)
Return value: Borrowed reference.

Extract the associated value from iter.

int json_object_iter_set(json_t *object, void *iter, json_t *value)

Set the value of the key-value pair in object, that is pointed to by iter, to value.

int json_object_iter_set_new(json_t *object, void *iter, json_t *value)

Like json_object_iter_set(), but steals the reference to value. This is useful when value is newly created and not used after the call.

The iteration protocol can be used for example as follows:

/* obj is a JSON object */
const char *key;
json_t *value;
void *iter = json_object_iter(obj);
while(iter)
{
    key = json_object_iter_key(iter);
    value = json_object_iter_value(iter);
    /* use key and value ... */
    iter = json_object_iter_next(obj, iter);
}

Error reporting

Jansson uses a single struct type to pass error information to the user. See sections Decoding, Building Values and Parsing and Validating Values for functions that pass error information using this struct.

json_error_t
char text[]

The error message (in UTF-8), or an empty string if a message is not available.

char source[]

Source of the error. This can be (a part of) the file name or a special identifier in angle brackers (e.g. <string>).

int line

The line number on which the error occurred.

int column

The column on which the error occurred. Note that this is the character column, not the byte column, i.e. a multibyte UTF-8 character counts as one column.

size_t position

The position in bytes from the start of the input. This is useful for debugging Unicode encoding problems.

The normal use of json_error_t is to allocate it on the stack, and pass a pointer to a function. Example:

int main() {
    json_t *json;
    json_error_t error;

    json = json_load_file("/path/to/file.json", 0, &error);
    if(!json) {
        /* the error variable contains error information */
    }
    ...
}

Also note that if the call succeeded (json != NULL in the above example), the contents of error are unspecified.

All functions also accept NULL as the json_error_t pointer, in which case no error information is returned to the caller.

Encoding

This sections describes the functions that can be used to encode values to JSON. By default, only objects and arrays can be encoded directly, since they are the only valid root values of a JSON text. To encode any JSON value, use the JSON_ENCODE_ANY flag (see below).

By default, the output has no newlines, and spaces are used between array and object elements for a readable output. This behavior can be altered by using the JSON_INDENT and JSON_COMPACT flags described below. A newline is never appended to the end of the encoded JSON data.

Each function takes a flags parameter that controls some aspects of how the data is encoded. Its default value is 0. The following macros can be ORed together to obtain flags.

JSON_INDENT(n)
Pretty-print the result, using newlines between array and object items, and indenting with n spaces. The valid range for n is between 0 and 31 (inclusive), other values result in an undefined output. If JSON_INDENT is not used or n is 0, no newlines are inserted between array and object items.
JSON_COMPACT
This flag enables a compact representation, i.e. sets the separator between array and object items to "," and between object keys and values to ":". Without this flag, the corresponding separators are ", " and ": " for more readable output.
JSON_ENSURE_ASCII
If this flag is used, the output is guaranteed to consist only of ASCII characters. This is achived by escaping all Unicode characters outside the ASCII range.
JSON_SORT_KEYS
If this flag is used, all the objects in output are sorted by key. This is useful e.g. if two JSON texts are diffed or visually compared.
JSON_PRESERVE_ORDER
If this flag is used, object keys in the output are sorted into the same order in which they were first inserted to the object. For example, decoding a JSON text and then encoding with this flag preserves the order of object keys.
JSON_ENCODE_ANY

Specifying this flag makes it possible to encode any JSON value on its own. Without it, only objects and arrays can be passed as the root value to the encoding functions.

Note: Encoding any value may be useful in some scenarios, but it’s generally discouraged as it violates strict compatiblity with RFC 4627. If you use this flag, don’t expect interoperatibility with other JSON systems. Even Jansson itself doesn’t have any means to decode JSON texts whose root value is not object or array.

New in version 2.1.

The following functions perform the actual JSON encoding. The result is in UTF-8.

char *json_dumps(const json_t *root, size_t flags)

Returns the JSON representation of root as a string, or NULL on error. flags is described above. The return value must be freed by the caller using free().

int json_dumpf(const json_t *root, FILE *output, size_t flags)

Write the JSON representation of root to the stream output. flags is described above. Returns 0 on success and -1 on error. If an error occurs, something may have already been written to output. In this case, the output is undefined and most likely not valid JSON.

int json_dump_file(const json_t *json, const char *path, size_t flags)

Write the JSON representation of root to the file path. If path already exists, it is overwritten. flags is described above. Returns 0 on success and -1 on error.

json_dump_callback_t

A typedef for a function that’s called by json_dump_callback():

typedef int (*json_dump_callback_t)(const char *buffer, size_t size, void *data);

buffer points to a buffer containing a chunk of output, size is the length of the buffer, and data is the corresponding json_dump_callback() argument passed through.

On error, the function should return -1 to stop the encoding process. On success, it should return 0.

New in version 2.2.

int json_dump_callback(const json_t *json, json_dump_callback_t callback, void *data, size_t flags)

Call callback repeatedly, passing a chunk of the JSON representation of root each time. flags is described above. Returns 0 on success and -1 on error.

New in version 2.2.

Decoding

This sections describes the functions that can be used to decode JSON text to the Jansson representation of JSON data. The JSON specification requires that a JSON text is either a serialized array or object, and this requirement is also enforced with the following functions. In other words, the top level value in the JSON text being decoded must be either array or object.

See RFC Conformance for a discussion on Jansson’s conformance to the JSON specification. It explains many design decisions that affect especially the behavior of the decoder.

Each function takes a flags parameter that can be used to control the behavior of the decoder. Its default value is 0. The following macros can be ORed together to obtain flags.

JSON_REJECT_DUPLICATES

Issue a decoding error if any JSON object in the input text contains duplicate keys. Without this flag, the value of the last occurence of each key ends up in the result. Key equivalence is checked byte-by-byte, without special Unicode comparison algorithms.

New in version 2.1.

JSON_DISABLE_EOF_CHECK

By default, the decoder expects that its whole input constitutes a valid JSON text, and issues an error if there’s extra data after the otherwise valid JSON input. With this flag enabled, the decoder stops after decoding a valid JSON array or object, and thus allows extra data after the JSON text.

New in version 2.3.

JSON_USE_BIGINT
This will enable the use of a big number package to be used to store large integer values, assuming a suitable big number package has been registered. Only those JSON integers whose values can not be stored in a json_int_t will use the big number extension.
JSON_USE_BIGINT_ALWAYS
This flag implies JSON_USE_BIGINT and differs by using the big number extension to store all JSON integers, even those that could have otherwise been stored in a json_int_t.
JSON_USE_BIGREAL
This will enable the use of a big number package to be used to store large real values, assuming a suitable big number package has been registered. Only those JSON reals whose values can not be accurately stored in a json_real_t, either because their values or exponents are out of range or there would be a loss of precision (dropped significant digits), will use the big number extension.
JSON_USE_BIGREAL_ALWAYS
This flag implies JSON_USE_BIGREAL and differs by using the big number extension to store all JSON reals, even those that could have otherwise been accurately stored in a json_real_t.

The following functions perform the actual JSON decoding.

json_t *json_loads(const char *input, size_t flags, json_error_t *error)
Return value: New reference.

Decodes the JSON string input and returns the array or object it contains, or NULL on error, in which case error is filled with information about the error. flags is described above.

json_t *json_loadb(const char *buffer, size_t buflen, size_t flags, json_error_t *error)
Return value: New reference.

Decodes the JSON string buffer, whose length is buflen, and returns the array or object it contains, or NULL on error, in which case error is filled with information about the error. This is similar to json_loads() except that the string doesn’t need to be null-terminated. flags is described above.

New in version 2.1.

json_t *json_loadf(FILE *input, size_t flags, json_error_t *error)
Return value: New reference.

Decodes the JSON text in stream input and returns the array or object it contains, or NULL on error, in which case error is filled with information about the error. flags is described above.

This function will start reading the input from whatever position the input file was, without attempting to seek first. If an error occurs, the file position will be left indeterminate. On success, the file position will be at EOF, unless JSON_DISABLE_EOF_CHECK flag was used. In this case, the file position will be at the first character after the last ] or } in the JSON input. This allows calling json_loadf() on the same FILE object multiple times, if the input consists of consecutive JSON texts, possibly separated by whitespace.

json_t *json_load_file(const char *path, size_t flags, json_error_t *error)
Return value: New reference.

Decodes the JSON text in file path and returns the array or object it contains, or NULL on error, in which case error is filled with information about the error. flags is described above.

Building Values

This section describes functions that help to create, or pack, complex JSON values, especially nested objects and arrays. Value building is based on a format string that is used to tell the functions about the expected arguments.

For example, the format string "i" specifies a single integer value, while the format string "[ssb]" or the equivalent "[s, s, b]" specifies an array value with two strings and a boolean as its items:

/* Create the JSON integer 42 */
json_pack("i", 42);

/* Create the JSON array ["foo", "bar", true] */
json_pack("[ssb]", "foo", "bar", 1);

Here’s the full list of format characters. The type in parentheses denotes the resulting JSON type, and the type in brackets (if any) denotes the C type that is expected as the corresponding argument.

s (string) [const char *]
Convert a NULL terminated UTF-8 string to a JSON string.
n (null)
Output a JSON null value. No argument is consumed.
b (boolean) [int]
Convert a C int to JSON boolean value. Zero is converted to false and non-zero to true.
i (integer) [int]
Convert a C int to JSON integer.
I (integer) [json_int_t]
Convert a C json_int_t to JSON integer.
z (big integer) [json_bigz_t]
Convert a C json_bigz_const_t (a constant-pointer version of a json_bigz_t) to a JSON integer. The provided value will be copied, no reference to it will be maintained. You must have already registered a suitable big number package extension.
f (real) [double]
Convert a C double to JSON real.
r (big real) [json_bigr_t]
Convert a C json_bigr_const_t (a constant-pointer version of a json_bigr_t) to a JSON real. The provided value will be copied, no reference to it will be maintained. You must have already registered a suitable big number package extension.
o (any value) [json_t *]
Output any given JSON value as-is. If the value is added to an array or object, the reference to the value passed to o is stolen by the container.
O (any value) [json_t *]
Like o, but the argument’s reference count is incremented. This is useful if you pack into an array or object and want to keep the reference for the JSON value consumed by O to yourself.
[fmt] (array)
Build an array with contents from the inner format string. fmt may contain objects and arrays, i.e. recursive value building is supported.
{fmt} (object)
Build an object with contents from the inner format string fmt. The first, third, etc. format character represent a key, and must be s (as object keys are always strings). The second, fourth, etc. format character represent a value. Any value may be an object or array, i.e. recursive value building is supported.

The following functions compose the value building API:

json_t *json_pack(const char *fmt, ...)
Return value: New reference.

Build a new JSON value according to the format string fmt. For each format character (except for {}[]n), one argument is consumed and used to build the corresponding value. Returns NULL on error.

json_t *json_pack_ex(json_error_t *error, size_t flags, const char *fmt, ...)
json_t *json_vpack_ex(json_error_t *error, size_t flags, const char *fmt, va_list ap)
Return value: New reference.

Like json_pack(), but an in the case of an error, an error message is written to error, if it’s not NULL. The flags parameter is currently unused and should be set to 0.

As only the errors in format string (and out-of-memory errors) can be caught by the packer, these two functions are most likely only useful for debugging format strings.

More examples:

/* Build an empty JSON object */
json_pack("{}");

/* Build the JSON object {"foo": 42, "bar": 7} */
json_pack("{sisi}", "foo", 42, "bar", 7);

/* Like above, ':', ',' and whitespace are ignored */
json_pack("{s:i, s:i}", "foo", 42, "bar", 7);

/* Build the JSON array [[1, 2], {"cool": true}] */
json_pack("[[i,i],{s:b]]", 1, 2, "cool", 1);

Parsing and Validating Values

This sectinon describes functions that help to validate complex values and extract, or unpack, data from them. Like building values, this is also based on format strings.

While a JSON value is unpacked, the type specified in the format string is checked to match that of the JSON value. This is the validation part of the process. In addition to this, the unpacking functions can also check that all items of arrays and objects are unpacked. This check be enabled with the format character ! or by using the flag JSON_STRICT. See below for details.

Here’s the full list of format characters. The type in parentheses denotes the JSON type, and the type in brackets (if any) denotes the C type whose address should be passed.

s (string) [const char *]
Convert a JSON string to a pointer to a NULL terminated UTF-8 string. The resulting string is extracted by using json_string_value() internally, so it exists as long as there are still references to the corresponding JSON string.
n (null)
Expect a JSON null value. Nothing is extracted.
b (boolean) [int]
Convert a JSON boolean value to a C int, so that true is converted to 1 and false to 0.
i (integer) [int]
Convert a JSON integer to C int.
I (integer) [json_int_t]
Convert a JSON integer to C json_int_t.
z (big integer) [json_bigz_t]
Convert a JSON big integer to a C json_bigz_t. The returned pointer will reference a newly allocated big number; you are responsible for eventually freeing it. You must have already registered a suitable big number package extension.
Z (any integer) [json_bigz_t]
Like z, except that both plain integers and big integers will be accepted. When extracting values, a big integer will always be returned.
f (real) [double]
Convert a JSON real to C double.
F (integer or real) [double]
Convert a JSON number (integer or real) to C double.
r (big real) [json_bigr_t]
Convert a JSON big real to a C json_bigr_t. The returned pointer will reference newly allocated big number; you are responsible for eventually freeing it. You must have already registered a suitable big number package extension.
R (any integer) [json_bigr_t]
Like``r``, except that both plain reals and big reals will be accepted. When extracting values, a big real will always be returned.
o (any value) [json_t *]
Store a JSON value with no conversion to a json_t pointer.
O (any value) [json_t *]
Like O, but the JSON value’s reference count is incremented.
v (any scalar) [json_t *]
Store any JSON scalar value (any type except lists or objects) with no conversion to a json_t pointer.
V (any scalar) [json_t *]
Like v, but the JSON value’s reference count is incremented.
[fmt] (array)
Convert each item in the JSON array according to the inner format string. fmt may contain objects and arrays, i.e. recursive value extraction is supporetd.
{fmt} (object)

Convert each item in the JSON object according to the inner format string fmt. The first, third, etc. format character represent a key, and must be s. The corresponding argument to unpack functions is read as the object key. The second fourth, etc. format character represent a value and is written to the address given as the corresponding argument. Note that every other argument is read from and every other is written to.

fmt may contain objects and arrays as values, i.e. recursive value extraction is supporetd.

!
This special format character is used to enable the check that all object and array items are accessed, on a per-value basis. It must appear inside an array or object as the last format character before the closing bracket or brace. To enable the check globally, use the JSON_STRICT unpacking flag.
*
This special format character is the opposite of !. If the JSON_STRICT flag is used, * can be used to disable the strict check on a per-value basis. It must appear inside an array or object as the last format character before the closing bracket or brace.

The following functions compose the parsing and validation API:

int json_unpack(json_t *root, const char *fmt, ...)

Validate and unpack the JSON value root according to the format string fmt. Returns 0 on success and -1 on failure.

int json_unpack_ex(json_t *root, json_error_t *error, size_t flags, const char *fmt, ...)
int json_vunpack_ex(json_t *root, json_error_t *error, size_t flags, const char *fmt, va_list ap)

Validate and unpack the JSON value root according to the format string fmt. If an error occurs and error is not NULL, write error information to error. flags can be used to control the behaviour of the unpacker, see below for the flags. Returns 0 on success and -1 on failure.

Note

The unpacking will halt at the first error, which may leave some of the variables that you designated to hold unpacked values in an uninitialized state. In partciular, as the big number formats (z, Z, r, and R) return pointers to newly allocated memory, it is good practice to always initialize those corresponding variables to NULL prior to unpacking, and to remember to free the memory for those (if not null) afterwards, regardless if the unpacking succeeded or resulted in an error.

Note

The first argument of all unpack functions is json_t *root instead of const json_t *root, because the use of O format character causes the reference count of root, or some value reachable from root, to be increased. Furthermore, the o format character may be used to extract a value as-is, which allows modifying the structure or contents of a value reachable from root.

If the O and o format character are not used, it’s perfectly safe to cast a const json_t * variable to plain json_t * when used with these functions.

The following unpacking flags are available:

JSON_STRICT
Enable the extra validation step checking that all object and array items are unpacked. This is equivalent to appending the format character ! to the end of every array and object in the format string.
JSON_VALIDATE_ONLY
Don’t extract any data, just validate the JSON value against the given format string. Note that object keys must still be specified after the format string.

Examples:

/* root is the JSON integer 42 */
int myint;
json_unpack(root, "i", &myint);
assert(myint == 42);

/* root is the JSON object {"foo": "bar", "quux": true} */
const char *str;
int boolean;
json_unpack(root, "{s:s, s:b}", "foo", &str, "quux", &boolean);
assert(strcmp(str, "bar") == 0 && boolean == 1);

/* root is the JSON array [[1, 2], {"baz": null} */
json_error_t error;
json_unpack_ex(root, &error, JSON_VALIDATE_ONLY, "[[i,i], {s:n}]", "baz");
/* returns 0 for validation success, nothing is extracted */

/* root is the JSON array [1, 2, 3, 4, 5] */
int myint1, myint2;
json_unpack(root, "[ii!]", &myint1, &myint2);
/* returns -1 for failed validation */

Equality

Testing for equality of two JSON values cannot, in general, be achieved using the == operator. Equality in the terms of the == operator states that the two json_t pointers point to exactly the same JSON value. However, two JSON values can be equal not only if they are exactly the same value, but also if they have equal “contents”:

  • Two integer or two real values are equal if their contained numeric values are equal. Comparisons between big number and corresponding regular number types are allowable, and are likewise determined based upon their equivalent numeric values. However, an integer value is never equal to a real value.
  • Two strings are equal if their contained UTF-8 strings are equal, byte by byte. Unicode comparison algorithms are not implemented.
  • Two arrays are equal if they have the same number of elements and each element in the first array is equal to the corresponding element in the second array.
  • Two objects are equal if they have exactly the same keys and the value for each key in the first object is equal to the value of the corresponding key in the second object.
  • Two true, false or null values have no “contents”, so they are equal if their types are equal. (Because these values are singletons, their equality can actually be tested with ==.)

The following function can be used to test whether two JSON values are equal.

int json_equal(json_t *value1, json_t *value2)

Returns 1 if value1 and value2 are equal, as defined above. Returns 0 if they are inequal or one or both of the pointers are NULL.

Copying

Because of reference counting, passing JSON values around doesn’t require copying them. But sometimes a fresh copy of a JSON value is needed. For example, if you need to modify an array, but still want to use the original afterwards, you should take a copy of it first.

Jansson supports two kinds of copying: shallow and deep. There is a difference between these methods only for arrays and objects. Shallow copying only copies the first level value (array or object) and uses the same child values in the copied value. Deep copying makes a fresh copy of the child values, too. Moreover, all the child values are deep copied in a recursive fashion.

json_t *json_copy(json_t *value)
Return value: New reference.

Returns a shallow copy of value, or NULL on error.

json_t *json_deep_copy(json_t *value)
Return value: New reference.

Returns a deep copy of value, or NULL on error.

Custom Memory Allocation

By default, Jansson uses malloc() and free() for memory allocation. These and other memory-related functions can be overridden if custom behavior is needed.

json_malloc_t

A typedef for a function pointer with malloc()‘s signature:

typedef void *(*json_malloc_t)(size_t);
json_free_t

A typedef for a function pointer with free()‘s signature:

typedef void (*json_free_t)(void *);
json_realloc_t

A typedef for a function pointer with realloc()‘s signature:

typedef void (*json_realloc_t)(void *, size_t);

Jansson does not itself use a realloc function, though it will pass it on to big number extensions that may require a realloc.

json_overwrite_t

A typedef for a function pointer that will securely overwrite a region of memory that has a signature:

typedef void (*json_overwrite_t)(void *, size_t);
void json_set_alloc_funcs(json_malloc_t malloc_fn, json_free_t free_fn)

Use malloc_fn instead of malloc() and free_fn instead of free(). This function has to be called before any other Jansson’s API functions to ensure that all memory operations use the same functions.

Supplying NULL as either function pointer will restore the default of using malloc() or free().

void json_set_realloc_func(json_realloc_t realloc_fn)

Uses realloc_fn instead of realloc(). This function will get passed to any big number extension for its possible use. Note that Jansson does not itself use realloc. This function should be called before any other of Jansson’s APIs are called which may deal with big number types.

Supplying NULL as the function pointer will restore the default of using realloc().

void json_set_overwrite_func(json_overwrite_t overwrite_fn)

Uses overwrite_fn to overwrite a region of memory. The default is to use an internal function that wraps memset() to write zero-bytes. However as some systems may optimize away the memset technique, a caller may wish to provide a more secure memory overwriting function.

Supplying NULL as the function pointer will restore the default of using an internal memset() based wrapper.

Examples:

Use the Boehm’s conservative garbage collector for memory operations:

json_set_alloc_funcs(GC_malloc, GC_free);
json_set_realloc_func(GC_realloc)

Allow storing sensitive data (e.g. passwords or encryption keys) in JSON structures by zeroing all memory when freed. You do not need to provide the secure_realloc function unless you use a big number extension which uses realloc:

static void secure_overwrite(void *ptr, size_t size)
{
    guaranteed_memset(ptr, 0, size);
}

static void *secure_malloc(size_t size)
{
    /* Store the memory area size in the beginning of the block */
    void *ptr = malloc(size + 8);
    *((size_t *)ptr) = size;
    return ptr + 8;
}

static void secure_free(void *ptr)
{
    size_t size;

    ptr -= 8;
    size = *((size_t *)ptr);

    secure_overwrite(ptr, size);
    free(ptr);
}

static void *secure_realloc(void *ptr, size_t size)
{
    size_t oldsize;

    if(ptr == NULL)
        return secure_malloc(size);
    if(size == 0) {
        secure_free(ptr);
        return NULL;
    }

    oldsize  = *((size_t *)(ptr - 8));

    if( oldsize > size ) {
        secure_overwrite(ptr+size, oldsize-size);
    }
    else if( oldsize < size ) {
        void *newptr = secure_malloc(size);
        memcpy(newptr, ptr, size);
        secure_free(ptr);
        ptr = newptr;
    }
    return ptr;
}

int main()
{
    json_set_alloc_funcs(secure_malloc, secure_free);
    json_set_realloc_func(secure_realloc);
    json_set_overwrite_func(secure_overwrite);
    /* ... */
}

For more information about the issues of storing sensitive data in memory, see http://www.dwheeler.com/secure-programs/Secure-Programs-HOWTO/protect-secrets.html. The page also explains the guaranteed_memset() function used in the example and gives a sample implementation for it.

Big Number Extensions

It is possible to extend Jansson so that it may support numeric values that are larger or more precise than may be represented by the native C types. Big number extensions may be independently provided for integer values and real values.

The extension API is designed to be generic so that many different big number packages may be used. Each extension is enabled by registering a set of user-supplied callback functions that perform basic operations on a big number. These callbacks will often be thin wrapper functions around the routines provided by the big number package being used.

For more information on big numbers and a list of some software packages which support them see http://en.wikipedia.org/wiki/Arbitrary-precision_arithmetic

Big number types and functions

Within Jansson, a big number type is represented as an opaque pointer type of json_bigz_t and json_bigr_t for big integers and big reals respectively. There are corresponding constant-pointer types as well, json_bigz_const_t and json_bigr_const_t.

By default these pointer types are declared as void * or void const *. To allow for better type safety, the user may provide a more specific type name for these pointers by defining a macro prior to including the Jansson header file; for example if using GMP:

#define JSON_BIGZ_TYPE mpz_t
#define JSON_BIGR_TYPE mpf_t
#include <jansson.h>

Then the json_bigz_t type will be equivalent to mpz_t *, and similar for the real types.

Note

Jansson adopts the convention of using the letter Z to mean integer and R to mean a real number. Do not let this cause you confusion, as some big number packages may use F for reals instead.

json_t *json_biginteger(json_bigz_const_t value);
Return value: New reference.

Returns a new JSON big integer, or NULL on error. The passed-in value is copied.

json_bigz_const_t json_biginteger_value(const json_t *biginteger)

Returns the pointer to the value of biginteger, or NULL if it is not a JSON big integer. Note that the returned pointer is a reference to the existing value and not a copy; use caution if retaining this reference.

int json_biginteger_set(json_t *integer, json_bigz_const_t value)

Sets the associated value of biginteger to value. Returns 0 on success and -1 if biginteger is not a JSON big integer. A copy is made of value.

json_t *json_bigreal(json_bigr_const_t value);
Return value: New reference.

Returns a new JSON big real, or NULL on error. The passed-in value is copied.

json_bigr_const_t json_bigreal_value(const json_t *bigreal)

Returns the pointer to the value of bigreal, or NULL if it is not a JSON big real. Note that the returned pointer is a reference to the existing value and not a copy; use caution if retaining this reference.

int json_bigreal_set(json_t *bigreal, json_bigr_const_t value)

Sets the associated value of bigreal to value. Returns 0 on success and -1 if bigreal is not a JSON big real. A copy is made of value.

Sample bignum packages

Included with the Jansson distribution are a sampling of extensions for supporting several popular big number formats. These samples are C code files which may be included and compiled into your own application.

C long double: The C standard provides a native type long double which may have greater range and precision than a plain double for real numbers. To use it add the sample file json_bignum_ldbl.c into your project sources. Then arrange your own source similar to:

#define JSON_BIGR_TYPE long double
#include <jansson.h>

int main()
{
    json_use_ldbl_for_bigreals();
    ...
}

GNU Quadmath: The GNU libquadmath library provides support for quad-precision floating-point numbers, which may have even greater range and precision than a long double. Quadmath is built into newer versions of the GCC compiler on some platforms, so that it acts like a native type. To use it add the sample file json_bignum_quad.c into your project sources. Then arrange your own source similar to:

#include <quadmath.h>
#define JSON_BIGR_TYPE __float128
#include <jansson.h>

int main()
{
    json_use_quad_for_bigreals();
    ...
}

Note that you may need to link your project with an additional system library typically named libquadmath (use the -lquadmath linker option in Unix-like environments). For more information see http://gcc.gnu.org/onlinedocs/libquadmath/

GNU Multiprecision Library (GMP): The GMP package supports arbitrary sized integers and arbitrary precision real numbers. To use it add the sample file json_bignum_gmp.c into your project sources. Then arrange your own source similar to:

#include <gmp.h>
#define JSON_BIGZ_TYPE mpz_t
#define JSON_BIGR_TYPE mpf_t
#include <jansson.h>

int main()
{
    json_use_gmp_for_bigintegers();
    json_use_gmp_for_bigreals();
    mpf_set_default_prec( 1024 ); /* precision in bits */
    ...
}

Notice that you may wish to call the GMP function mpf_set_default_prec() to set up the default number of bits of precision that are kept. Jansson will always attempt to preserve all of the digits, but this GMP default if not set appropriately may still result in loss of significant digits.

You will need to link your project with the GMP library (use the -lgmp linker option in Unix-like environments). For more information see http://gmplib.org/

OpenSSL Big Numbers (BN): The OpenSSL cryptographic library provides generic support for arbitrary sized integers. To use it add the sample file json_bignum_openssl.c into your project sources. Then arrange your own source similar to:

#include <openssl/bn.h>
#define JSON_BIGZ_TYPE BIGNUM
#include <jansson.h>

int main()
{
    json_use_openssl_for_bigintegers();
    ...
}

You will need to link your project with the OpenSSL crypto library (use the -lcrypto linker option in Unix-like environments). For more information see http://www.openssl.org/docs/crypto/bn.html

The callback functions

To use a big number package you must provide a set of callback functions that manipulate big number values. There are six callbacks needed: copy, delete, compare, convert to string, convert from string, and up-convert from a native number type. The callbacks are generally the same for both big integers and big reals except for the specific type signatures.

When invoked by Jansson, all of the callback function will be provided with a json_memory_funcs_t argument that contains the set of function pointers for common memory-related routines. The callback functions should make use of these functions when possible so that both Jansson and the big number package use compatible memory management routines. The members of the json_memory_funcs_t include:

malloc_fn(size_t size)
free_fn(void *ptr)
realloc_fn(void *ptr, size_t size)
overwrite_fn(void *ptr, size_t size)
strdup_fn(const char *str)

Each big number package must provide the following callback functions, shown here for big integers:

json_bigint_copy_t

A typedef for a function pointer that will make a copy of a big integer value. It has a signature:

typedef json_bigz_t (*json_bigint_copy_t)(json_bigz_const_t bignum, const json_memory_funcs_t *memfuncs)
json_bigint_del_t

A typedef for a function pointer that will delete a big integer value. It has a signature:

typedef void (*json_bigint_del_t)(json_bigz_t bignum, const json_memory_funcs_t *memfuncs)
json_bigint_cmp_t

A typedef for a function pointer that will numerically compare two big integer values, returning -1 (less-than), 0 (equal), or +1 (greater-than). It has a signature:

typedef int (*json_bigint_cmp_t)(json_bigz_const_t bignum1, json_bigz_const_t bignum2, const json_memory_funcs_t *memfuncs)
json_bigint_to_str_t

A typedef for a function pointer that will convert a big number value into a decimal string format. The resulting format must adhere strictly to the JSON standard syntax, e.g., “+0” or ”.3” are invalid. It has a signature:

typedef int (*json_bigint_to_str_t)(json_bigz_const_t bignum, char *buffer, size_t size, const json_memory_funcs_t *memfuncs)

The function will be provided a buffer, identified with buffer and size, into which it should write the string. The string must be null terminated. The length of the string excluding the null byte must be returned.

If the resulting string is too large to fit into the provided buffer, then the function should return the number of bytes needed (excluding the null terminator), after which the function will be called again with a larger buffer. In this case it is not necessary to write into the buffer or insure it is null terminated.

json_bigint_from_str_t

A typedef for a function pointer that will convert a decimal string repreentation of a number (in standard JSON syntax) into a big number value. It has the signature:

typedef json_bigz_t (*json_bigint_from_str_t)(const char *value, const json_memory_funcs_t *memfuncs)

The returned value should be a newly-allocated big number type. If the string value can not be converted then NULL should be returned.

json_bigint_from_int_t

A typedef for a function pointer that will convert a standard native type of json_int_t into a big number value. It has the signture:

typedef json_bigz_t (*json_bigint_from_int_t)(json_int_t value, const json_memory_funcs_t *memfuncs)

The returned value should be a newly-allocated big number type.

To register the callback functions, they must first be assembled into a structure of type json_bigint_funcs_t or json_bigreal_funcs_t, for example:

json_bigint_funcs_t callbacks;
callbacks.copy_fn = my_copy_function;
callbacks.delete_fn = my_delete_function;
callbacks.compare_fn = my_compare_function;
callbacks.to_string_fn = my_to_string_function;
callbacks.from_string_fn = my_from_string_function;
callbacks.from_int_fn = my_from_int_function;

Callbacks for big reals are similar except the last member is named from_real_fn instead of from_int_fn.

Then the whole set of callbacks are registered with a single API call.

void json_set_biginteger_funcs(const json_bigint_funcs_t* functions)

Registers the set of callback functions to use for manipulating big integer values. Passing NULL will unregister any callback functions and disable the use of big integers. This function should be called before any other functions that may involve big integers, such as loading JSON.

void json_set_bigreal_funcs(const json_bigreal_funcs_t* functions)

Registers the set of callback functions to use for manipulating big real values. Passing NULL will unregister any callback functions and disable the use of big reals. This function should be called before any other functions that may involve big integers, such as loading JSON.

Example using long double

The following example code shows how to use the big number extensions to support the native C type of long double for big real numbers. This is essentially the same as what is in the supplied json_bignum_ldbl.c sample:

#include <string.h>
#include <stdlib.h>
#include <float.h>
#include <errno.h>

#define JSON_BIGR_TYPE long double
#include <jansson.h>

/* These typedefs are now in effect:
     json_bigr_t        =>  long double *
     json_bigr_const_t  =>  long double const *
*/

static int json_bigreal_ldbl_compare(json_bigr_const_t r1, json_bigr_const_t r2,
                                     const json_memory_funcs_t *memfuncs)
{
    const long double * f1 = r1;
    const long double * f2 = r2;

    if( *f1 == *f2 ) return 0;
    if( *f1 < *f2 )  return -1;
    return 1;
}

static json_bigr_t json_bigreal_ldbl_copy(json_bigr_const_t r,
                                          const json_memory_funcs_t *memfuncs)
{
    const long double * f1 = r;
    long double * f2;

    f2 = memfuncs->malloc_fn( sizeof(long double) );
    if(!f2)
        return NULL;
    *f2 = *f1;
    return f2;
}

static void json_bigreal_ldbl_delete(json_bigr_t r,
                                     const json_memory_funcs_t *memfuncs)
{
    long double * f = r;

    memfuncs->free_fn( f );
    return;
}

static json_bigr_t json_bigreal_ldbl_from_real(double value,
                                               const json_memory_funcs_t *memfuncs)
{
    long double * f;

    f = memfuncs->malloc_fn( sizeof(long double) );
    if(!f)
        return NULL;
    *f = value;
    return f;
}

static json_bigr_t json_bigreal_ldbl_from_str(const char *value,
                                              const json_memory_funcs_t *memfuncs)
{
    long double f0;
    long double *f1;

    errno = 0;
    f0 = strtold( value, NULL );

    f1 = json_bigreal_ldbl_copy( &f0, memfuncs );
    memfuncs->overwrite_fn( &f0, sizeof(long double) );
    return f1;
}

static int json_bigreal_ldbl_to_str(json_bigr_const_t r, char *buffer, size_t size,
                                    const json_memory_funcs_t *memfuncs)
{
    const long double *f = r;
    int outsize;

    outsize = snprintf( buffer, size, "%.*Lg", LDBL_DIG, *f );
    return outsize;
}

int json_use_ldbl_for_bigreals()
{
    static json_bigreal_funcs_t funcs;
    funcs.copy_fn        = json_bigreal_ldbl_copy;
    funcs.delete_fn      = json_bigreal_ldbl_delete;
    funcs.compare_fn     = json_bigreal_ldbl_compare;
    funcs.to_string_fn   = json_bigreal_ldbl_to_str;
    funcs.from_string_fn = json_bigreal_ldbl_from_str;
    funcs.from_real_fn   = json_bigreal_ldbl_from_real;

    json_set_bigreal_funcs( &funcs );
    return 0;
}

int main()
{
    json_use_ldbl_for_bigreals();
    ...
}