jsmn (pronounced like 'jasmine') is a minimalistic JSON parser in C. It can be easily integrated into resource-limited or embedded projects. This version is based on Serge Zaitsev's original single-file jsmn.h with several key differences:

• jsmn is split into two files, jsmn.h and jsmn.c for ease of integration into workflows that need a two-file layout.
• Each parsed token carries a pointer and a length field to its underlying JSON string.
• jsmn_token_ref() provides a safe accessor to the parsed tokens.
• A suite of predicates -- jsmn_token_is_array(), jsmn_token_is_number(), etc -- help categorize parsed tokens.

You can find more information about JSON format at json.org

zserge's original sources are available at https://github.com/zserge/jsmn

The web page with information on the original jsmn design can be found at http://zserge.com/jsmn.html

Most JSON parsers offer you a suite of functions to load JSON data, parse it and extract any value by its name. jsmn takes the view this this is overkill and instead parses the JSON data into individual token strings, leaving it to the user to perform the conversion from string to the desired type.

jsmn is designed to be robust (it should work fine even with erroneous data), fast (it should parse data on the fly), portable (no superfluous dependencies or non-standard C extensions). And of course, simplicity is a key feature - simple code style, simple algorithm, simple integration into other projects.

• compatible with C89
• minimal stdxxx dependencies
• highly portable (tested on x86/amd64, ARM, AVR)
• extremely small code footprint
• no dynamic memory allocation
• incremental single-pass parsing
• (increasingly) complete coverage with unit tests

The rudimentary jsmn object is a token. Let's consider a JSON string:

'{ "name" : "Jack", "age" : 27 }'

It holds the following tokens:

• Object: { "name" : "Jack", "age" : 27} (the whole object)
• Strings: "name", "Jack", "age" (keys and some values)
• Number: 27

In jsmn, tokens do not hold parsed data, but point to token boundaries in JSON string instead. In the example above jsmn will create tokens like:

• Object: { "na..., length = 30
• String: name" : ..., length = 4
• String: Jack", ..., length = 4
• String: age" : ..., length = 3
• Number: 27 }, length = 2

Every jsmn token has a type, which indicates the type of corresponding JSON token. jsmn supports the following primitive token types:

• Object - a container of key-value pairs, e.g.: { "foo":"bar", "x":0.3 }
• Array - a sequence of values, e.g.: [ 1, 2, 3 ]
• String - a quoted sequence of chars, e.g.: "foo"
• Primitive - a number, a boolean (true, false) or null

Besides start of string and length, jsmn tokens for complex types (like arrays or objects) also contain the number of child items so you can easily follow object hierarchy.

This approach provides enough information for parsing any JSON data and makes it possible to use zero-copy techniques.

Download jsmn.h and jsmn.c and incorporate them into your project.

#include "jsmn.h"

...
jsmn_parser_t p;
jsmn_token_t t[128]; /* We expect no more than 128 JSON tokens */

jsmn_init(&p, t, 128);
r = jsmn_parse(&p, s, strlen(s)); // "s" is the char array holding the json content

Token types are described by jsmn_token_type_t:

typedef enum {
	JSMN_UNDEFINED = 0,
	JSMN_OBJECT = 1 << 0,
	JSMN_ARRAY = 1 << 1,
	JSMN_STRING = 1 << 2,
	JSMN_PRIMITIVE = 1 << 3
} jsmn_token_type_t;

Note: Unlike JSON data types, primitive tokens are not divided into numbers, booleans and null, because one can easily tell the type using the first character:

• 't', 'f'- boolean
• 'n' - null
• '-', '0'..'9' - number

Token is an object of jsmn_token_t type:

typedef struct {
  jsmn_token_type_t type;
  const char *start;     // start of token
  int length;            // length of token
  int size;              // number of nested tokens within OBJECT or ARRAy
#ifdef JSMN_PARENT_LINKS
  int parent;            // index to token that contains this token
#endif
} jsmn_token_t;

Note: string tokens point to the first character after the opening quote and the previous symbol before final quote. This was made to simplify string extraction from JSON data.

All the parsing is done by a jsmn_parser_t object. You can parse a JSON string as follows:

    #include "jsmn.h"

	// you must provide a parser object and an array of tokens
	jsmn_parser_t parser;
	jsmn_token_t tokens[10];

    // initialize the parser and call jsmn_parse() on the string to be parsed.
	jsmn_init(&parser, tokens, sizeof(tokens)/sizeof(jsmn_token_t));
	int r = jsmn_parse(&parser, js, strlen(js));

	if (r < 0) {
		// there was some sort of error.  See JSMN_ERROR_NOMEM, etc, in jsmn.h
	} else {
		// r represents how many tokens were parsed.
	}

This will create a parser, and then it tries to parse up to 10 JSON tokens from the js string.

A non-negative return value of jsmn_parse is the number of tokens actually used by the parser. Passing NULL instead of the tokens array will not store parsing results, but instead will return the number of tokens needed to parse the given string. This can be useful if your platforms supports malloc() and you don't yet know how many tokens to allocate.

If something goes wrong, jsmn_parser() will return one of these negative values:

• JSMN_ERROR_INVAL - bad token, JSON string is corrupted
• JSMN_ERROR_NOMEM - not enough tokens, JSON string is too large
• JSMN_ERROR_PART - JSON string is too short, expecting more JSON data

If you get JSMN_ERROR_NOMEM, you can re-allocate more tokens and call jsmn_parse once more.

If you read json data from the stream, you can periodically call jsmn_parse and check if return value is JSMN_ERROR_PART. You will get this error until you reach the end of JSON data.

This software is distributed under MIT license, so feel free to integrate it in your commercial products.