There are myriads of [JSON](http://json.org) libraries out there, and each may even have its reason to exist. Our class had these design goals:
- **Intuitive syntax**. In languages such as Python, JSON feels like a first class data type. We used all the operator magic of modern C++ to achieve the same feeling in your code. Check out the [examples below](#examples) and you know, what I mean.
- **Trivial integration**. Our whole code consists of a single header file `json.hpp`. That's it. No library, no subproject, no dependencies, no complex build system. The class is written in vanilla C++11. All in all, everything should require no adjustment of your compiler flags or project settings.
- **Serious testing**. Our class is heavily [unit-tested](https://github.com/nlohmann/json/blob/master/test/json_unit.cc) and covers [100%](https://coveralls.io/r/nlohmann/json) of the code, including all exceptional behavior. Furthermore, we checked with [Valgrind](http://valgrind.org) that there are no memory leaks.
Other aspects were not so important to us:
- **Memory efficiency**. Each JSON object has an overhead of one pointer (the maximal size of a union) and one enumeration element (1 byte). The default generalization uses the following C++ data types: `std::string` for strings, `int64_t` or `double` for numbers, `std::map` for objects, `std::vector` for arrays, and `bool` for Booleans. However, you can template the generalized class `basic_json` to your needs.
- **Speed**. We currently implement the parser as naive [recursive descent parser](http://en.wikipedia.org/wiki/Recursive_descent_parser) with hand coded string handling. It is fast enough, but a [LALR-parser](http://en.wikipedia.org/wiki/LALR_parser) with a decent regular expression processor should be even faster (but would consist of more files which makes the integration harder).
- **Rigorous Unicode compliance**. We did our best to implement some robust Unicode support. There are still some issues with escaping, and if you run into a problem, please [tell me](https://github.com/nlohmann/json/issues).
## Updates since last version
As of February 2015, the following updates were made to the library
- *Changed:* In the generic class `basic_json`, all JSON value types (array, object, string, bool, integer number, and floating-point) are now **templated**. That is, you can choose whether you like a `std::list` for your arrays or an `std::unordered_map` for your objects. The specialization `json` sets some reasonable defaults.
- *Changed:* The library now consists of a **single header**, called `json.hpp`. Consequently, build systems such as Automake or CMake are not any longer required.
- *Changed:* The **deserialization** is now supported by a lexer generated with [re2c](http://re2c.org) from file [`src/json.hpp.re2c`](https://github.com/nlohmann/json/blob/master/src/json.hpp.re2c). As a result, we follow the JSON specification more strictly. Note neither the tool re2c nor its input are required to use the class.
- *Added:* The library now satisfies the [**ReversibleContainer**](http://en.cppreference.com/w/cpp/concept/ReversibleContainer) requirement. It hence provides four different iterators (`iterator`, `const_iterator`, `reverse_iterator`, and `const_reverse_iterator`), comparison functions, `swap()`, `size()`, `max_size()`, and `empty()` member functions.
- *Added*: The class uses **user-defined allocators** which default to `std::allocator`, but can be templated via parameter `Allocator`.
- *Added:* To simplify pretty-printing, the `std::setw`**stream manipulator** has been overloaded to set the desired indentation. Pretty-printing a JSON object `j` is as simple as `std::cout << std::setw(4) << j << '\n'`.
- *Changed*: The type `json::value_t::number` is now called `json::value_t::number_integer` to be more symmetric compared to `json::value_t::number_float`.
## Integration
The single required source, `json.hpp` file is in the `src` directory. All you need to do is add
```cpp
#include "json.hpp"
// for convenience
using json = nlohmann::json;
```
to the files you want to use JSON objects. That's it. Do not forget to set the necessary switches to enable C++11 (e.g., `-std=c++11` for GCC and Clang).
Unfortunately, Microsoft Visual C++ support at the moment. MSVC 2013 and MSVC 2014 have been reported not to be able to compile the class. Note that I will not accept pull requests that "fix" the code by polluting it with preprocessor directives to mitigate MSVC's lacking C++11 support.
Note that in all these cases, you never need to "tell" the compiler which JSON value you want to use. If you want to be explicit or express some edge cases, the functions `json::array` and `json::object` will help:
We designed the JSON class to behave just like an STL container. In fact, it satisfies the [**ReversibleContainer**](http://en.cppreference.com/w/cpp/concept/ReversibleContainer) requirement.
Any sequence container (`std::array`, `std::vector`, `std::deque`, `std::forward_list`, `std::list`) whose values can be used to construct JSON types (e.g., integers, floating point numbers, Booleans, string types, or again STL containers described in this section) can be used to create a JSON array. The same holds for similar associative containers (`std::set`, `std::multiset`, `std::unordered_set`, `std::unordered_multiset`), but in these cases the order of the elements of the array depends how the elements are ordered in the respective STL container.
json j_umset(c_umset); // both entries for "one" are used
// maybe ["one", "two", "one", "four"]
```
Likewise, any associative key-value containers (`std::map`, `std::multimap`, `std::unordered_map`, `std::unordered_multimap`) whose keys are can construct an `std::string` and whose values can be used to construct JSON types (see examples above) can be used to to create a JSON object. Note that in case of multimaps only one key is used in the JSON object and the value depends on the internal order of the STL container.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
## Thanks
I deeply appreciate the help of the following people.
- [Teemperor](https://github.com/Teemperor) implemented CMake support and lcov integration, realized escape and Unicode handling in the string parser, and fixed the JSON serialization.
- [elliotgoodrich](https://github.com/elliotgoodrich) fixed an issue with double deletion in the iterator classes.
- [kirkshoop](https://github.com/kirkshoop) made the iterators of the class composable to other libraries.
- [wancw](https://github.com/wanwc) fixed a bug that hindered the class to compile with Clang.
- Tomas Åblad found a bug in the iterator implementation.
- [Aaron Burghardt](https://github.com/aburgh) implemented code to parse streams incrementally. Furthermore, he greatly improved the parser class by allowing the definition of a filter function to discard undesired elements while parsing.