Added a library for working with JSON.

src/lib/json/adl_serializer.hpp

@@ -0,0 +1,45 @@
+#pragma once
+
+#include <utility>
+
+#include <nlohmann/detail/conversions/from_json.hpp>
+#include <nlohmann/detail/conversions/to_json.hpp>
+
+namespace nlohmann
+{
+template<typename, typename>
+struct adl_serializer
+{
+    /*!
+    @brief convert a JSON value to any value type
+
+    This function is usually called by the `get()` function of the
+    @ref basic_json class (either explicit or via conversion operators).
+
+    @param[in] j         JSON value to read from
+    @param[in,out] val  value to write to
+    */
+    template<typename BasicJsonType, typename ValueType>
+    static void from_json(BasicJsonType&& j, ValueType& val) noexcept(
+        noexcept(::nlohmann::from_json(std::forward<BasicJsonType>(j), val)))
+    {
+        ::nlohmann::from_json(std::forward<BasicJsonType>(j), val);
+    }
+
+    /*!
+    @brief convert any value type to a JSON value
+
+    This function is usually called by the constructors of the @ref basic_json
+    class.
+
+    @param[in,out] j  JSON value to write to
+    @param[in] val     value to read from
+    */
+    template<typename BasicJsonType, typename ValueType>
+    static void to_json(BasicJsonType& j, ValueType&& val) noexcept(
+        noexcept(::nlohmann::to_json(j, std::forward<ValueType>(val))))
+    {
+        ::nlohmann::to_json(j, std::forward<ValueType>(val));
+    }
+};
+}

src/lib/json/detail/conversions/from_json.hpp

@@ -0,0 +1,321 @@
+#pragma once
+
+#include <algorithm> // transform
+#include <array> // array
+#include <ciso646> // and, not
+#include <forward_list> // forward_list
+#include <iterator> // inserter, front_inserter, end
+#include <string> // string
+#include <tuple> // tuple, make_tuple
+#include <type_traits> // is_arithmetic, is_same, is_enum, underlying_type, is_convertible
+#include <utility> // pair, declval
+#include <valarray> // valarray
+
+#include <nlohmann/detail/exceptions.hpp>
+#include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/meta.hpp>
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+// overloads for basic_json template parameters
+template<typename BasicJsonType, typename ArithmeticType,
+         enable_if_t<std::is_arithmetic<ArithmeticType>::value and
+                     not std::is_same<ArithmeticType, typename BasicJsonType::boolean_t>::value,
+                     int> = 0>
+void get_arithmetic_value(const BasicJsonType& j, ArithmeticType& val)
+{
+    switch (static_cast<value_t>(j))
+    {
+        case value_t::number_unsigned:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_unsigned_t*>());
+            break;
+        }
+        case value_t::number_integer:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_integer_t*>());
+            break;
+        }
+        case value_t::number_float:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_float_t*>());
+            break;
+        }
+
+        default:
+            JSON_THROW(type_error::create(302, "type must be number, but is " + std::string(j.type_name())));
+    }
+}
+
+template<typename BasicJsonType>
+void from_json(const BasicJsonType& j, typename BasicJsonType::boolean_t& b)
+{
+    if (JSON_UNLIKELY(not j.is_boolean()))
+    {
+        JSON_THROW(type_error::create(302, "type must be boolean, but is " + std::string(j.type_name())));
+    }
+    b = *j.template get_ptr<const typename BasicJsonType::boolean_t*>();
+}
+
+template<typename BasicJsonType>
+void from_json(const BasicJsonType& j, typename BasicJsonType::string_t& s)
+{
+    if (JSON_UNLIKELY(not j.is_string()))
+    {
+        JSON_THROW(type_error::create(302, "type must be string, but is " + std::string(j.type_name())));
+    }
+    s = *j.template get_ptr<const typename BasicJsonType::string_t*>();
+}
+
+template<typename BasicJsonType>
+void from_json(const BasicJsonType& j, typename BasicJsonType::number_float_t& val)
+{
+    get_arithmetic_value(j, val);
+}
+
+template<typename BasicJsonType>
+void from_json(const BasicJsonType& j, typename BasicJsonType::number_unsigned_t& val)
+{
+    get_arithmetic_value(j, val);
+}
+
+template<typename BasicJsonType>
+void from_json(const BasicJsonType& j, typename BasicJsonType::number_integer_t& val)
+{
+    get_arithmetic_value(j, val);
+}
+
+template<typename BasicJsonType, typename EnumType,
+         enable_if_t<std::is_enum<EnumType>::value, int> = 0>
+void from_json(const BasicJsonType& j, EnumType& e)
+{
+    typename std::underlying_type<EnumType>::type val;
+    get_arithmetic_value(j, val);
+    e = static_cast<EnumType>(val);
+}
+
+template<typename BasicJsonType>
+void from_json(const BasicJsonType& j, typename BasicJsonType::array_t& arr)
+{
+    if (JSON_UNLIKELY(not j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name())));
+    }
+    arr = *j.template get_ptr<const typename BasicJsonType::array_t*>();
+}
+
+// forward_list doesn't have an insert method
+template<typename BasicJsonType, typename T, typename Allocator,
+         enable_if_t<std::is_convertible<BasicJsonType, T>::value, int> = 0>
+void from_json(const BasicJsonType& j, std::forward_list<T, Allocator>& l)
+{
+    if (JSON_UNLIKELY(not j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name())));
+    }
+    std::transform(j.rbegin(), j.rend(),
+                   std::front_inserter(l), [](const BasicJsonType & i)
+    {
+        return i.template get<T>();
+    });
+}
+
+// valarray doesn't have an insert method
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_convertible<BasicJsonType, T>::value, int> = 0>
+void from_json(const BasicJsonType& j, std::valarray<T>& l)
+{
+    if (JSON_UNLIKELY(not j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name())));
+    }
+    l.resize(j.size());
+    std::copy(j.m_value.array->begin(), j.m_value.array->end(), std::begin(l));
+}
+
+template<typename BasicJsonType, typename CompatibleArrayType>
+void from_json_array_impl(const BasicJsonType& j, CompatibleArrayType& arr, priority_tag<0> /*unused*/)
+{
+    using std::end;
+
+    std::transform(j.begin(), j.end(),
+                   std::inserter(arr, end(arr)), [](const BasicJsonType & i)
+    {
+        // get<BasicJsonType>() returns *this, this won't call a from_json
+        // method when value_type is BasicJsonType
+        return i.template get<typename CompatibleArrayType::value_type>();
+    });
+}
+
+template<typename BasicJsonType, typename CompatibleArrayType>
+auto from_json_array_impl(const BasicJsonType& j, CompatibleArrayType& arr, priority_tag<1> /*unused*/)
+-> decltype(
+    arr.reserve(std::declval<typename CompatibleArrayType::size_type>()),
+    void())
+{
+    using std::end;
+
+    arr.reserve(j.size());
+    std::transform(j.begin(), j.end(),
+                   std::inserter(arr, end(arr)), [](const BasicJsonType & i)
+    {
+        // get<BasicJsonType>() returns *this, this won't call a from_json
+        // method when value_type is BasicJsonType
+        return i.template get<typename CompatibleArrayType::value_type>();
+    });
+}
+
+template<typename BasicJsonType, typename T, std::size_t N>
+void from_json_array_impl(const BasicJsonType& j, std::array<T, N>& arr, priority_tag<2> /*unused*/)
+{
+    for (std::size_t i = 0; i < N; ++i)
+    {
+        arr[i] = j.at(i).template get<T>();
+    }
+}
+
+template <
+    typename BasicJsonType, typename CompatibleArrayType,
+    enable_if_t <
+        is_compatible_array_type<BasicJsonType, CompatibleArrayType>::value and
+        not std::is_same<typename BasicJsonType::array_t,
+                         CompatibleArrayType>::value and
+        std::is_constructible <
+            BasicJsonType, typename CompatibleArrayType::value_type >::value,
+        int > = 0 >
+void from_json(const BasicJsonType& j, CompatibleArrayType& arr)
+{
+    if (JSON_UNLIKELY(not j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, "type must be array, but is " +
+                                      std::string(j.type_name())));
+    }
+
+    from_json_array_impl(j, arr, priority_tag<2> {});
+}
+
+template<typename BasicJsonType, typename CompatibleObjectType,
+         enable_if_t<is_compatible_object_type<BasicJsonType, CompatibleObjectType>::value, int> = 0>
+void from_json(const BasicJsonType& j, CompatibleObjectType& obj)
+{
+    if (JSON_UNLIKELY(not j.is_object()))
+    {
+        JSON_THROW(type_error::create(302, "type must be object, but is " + std::string(j.type_name())));
+    }
+
+    auto inner_object = j.template get_ptr<const typename BasicJsonType::object_t*>();
+    using value_type = typename CompatibleObjectType::value_type;
+    std::transform(
+        inner_object->begin(), inner_object->end(),
+        std::inserter(obj, obj.begin()),
+        [](typename BasicJsonType::object_t::value_type const & p)
+    {
+        return value_type(p.first, p.second.template get<typename CompatibleObjectType::mapped_type>());
+    });
+}
+
+// overload for arithmetic types, not chosen for basic_json template arguments
+// (BooleanType, etc..); note: Is it really necessary to provide explicit
+// overloads for boolean_t etc. in case of a custom BooleanType which is not
+// an arithmetic type?
+template<typename BasicJsonType, typename ArithmeticType,
+         enable_if_t <
+             std::is_arithmetic<ArithmeticType>::value and
+             not std::is_same<ArithmeticType, typename BasicJsonType::number_unsigned_t>::value and
+             not std::is_same<ArithmeticType, typename BasicJsonType::number_integer_t>::value and
+             not std::is_same<ArithmeticType, typename BasicJsonType::number_float_t>::value and
+             not std::is_same<ArithmeticType, typename BasicJsonType::boolean_t>::value,
+             int> = 0>
+void from_json(const BasicJsonType& j, ArithmeticType& val)
+{
+    switch (static_cast<value_t>(j))
+    {
+        case value_t::number_unsigned:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_unsigned_t*>());
+            break;
+        }
+        case value_t::number_integer:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_integer_t*>());
+            break;
+        }
+        case value_t::number_float:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_float_t*>());
+            break;
+        }
+        case value_t::boolean:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::boolean_t*>());
+            break;
+        }
+
+        default:
+            JSON_THROW(type_error::create(302, "type must be number, but is " + std::string(j.type_name())));
+    }
+}
+
+template<typename BasicJsonType, typename A1, typename A2>
+void from_json(const BasicJsonType& j, std::pair<A1, A2>& p)
+{
+    p = {j.at(0).template get<A1>(), j.at(1).template get<A2>()};
+}
+
+template<typename BasicJsonType, typename Tuple, std::size_t... Idx>
+void from_json_tuple_impl(const BasicJsonType& j, Tuple& t, index_sequence<Idx...>)
+{
+    t = std::make_tuple(j.at(Idx).template get<typename std::tuple_element<Idx, Tuple>::type>()...);
+}
+
+template<typename BasicJsonType, typename... Args>
+void from_json(const BasicJsonType& j, std::tuple<Args...>& t)
+{
+    from_json_tuple_impl(j, t, index_sequence_for<Args...> {});
+}
+
+struct from_json_fn
+{
+  private:
+    template<typename BasicJsonType, typename T>
+    auto call(const BasicJsonType& j, T& val, priority_tag<1> /*unused*/) const
+    noexcept(noexcept(from_json(j, val)))
+    -> decltype(from_json(j, val), void())
+    {
+        return from_json(j, val);
+    }
+
+    template<typename BasicJsonType, typename T>
+    void call(const BasicJsonType& /*unused*/, T& /*unused*/, priority_tag<0> /*unused*/) const noexcept
+    {
+        static_assert(sizeof(BasicJsonType) == 0,
+                      "could not find from_json() method in T's namespace");
+#ifdef _MSC_VER
+        // MSVC does not show a stacktrace for the above assert
+        using decayed = uncvref_t<T>;
+        static_assert(sizeof(typename decayed::force_msvc_stacktrace) == 0,
+                      "forcing MSVC stacktrace to show which T we're talking about.");
+#endif
+    }
+
+  public:
+    template<typename BasicJsonType, typename T>
+    void operator()(const BasicJsonType& j, T& val) const
+    noexcept(noexcept(std::declval<from_json_fn>().call(j, val, priority_tag<1> {})))
+    {
+        return call(j, val, priority_tag<1> {});
+    }
+};
+}
+
+/// namespace to hold default `from_json` function
+/// to see why this is required:
+/// http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2015/n4381.html
+namespace
+{
+constexpr const auto& from_json = detail::static_const<detail::from_json_fn>::value;
+}
+}

src/lib/json/detail/conversions/to_json.hpp

@@ -0,0 +1,338 @@
+#pragma once
+
+#include <ciso646> // or, and, not
+#include <iterator> // begin, end
+#include <tuple> // tuple, get
+#include <type_traits> // is_same, is_constructible, is_floating_point, is_enum, underlying_type
+#include <utility> // move, forward, declval, pair
+#include <valarray> // valarray
+#include <vector> // vector
+
+#include <nlohmann/detail/meta.hpp>
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+//////////////////
+// constructors //
+//////////////////
+
+template<value_t> struct external_constructor;
+
+template<>
+struct external_constructor<value_t::boolean>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::boolean_t b) noexcept
+    {
+        j.m_type = value_t::boolean;
+        j.m_value = b;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::string>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::string_t& s)
+    {
+        j.m_type = value_t::string;
+        j.m_value = s;
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::string_t&& s)
+    {
+        j.m_type = value_t::string;
+        j.m_value = std::move(s);
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::number_float>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::number_float_t val) noexcept
+    {
+        j.m_type = value_t::number_float;
+        j.m_value = val;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::number_unsigned>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::number_unsigned_t val) noexcept
+    {
+        j.m_type = value_t::number_unsigned;
+        j.m_value = val;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::number_integer>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::number_integer_t val) noexcept
+    {
+        j.m_type = value_t::number_integer;
+        j.m_value = val;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::array>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::array_t& arr)
+    {
+        j.m_type = value_t::array;
+        j.m_value = arr;
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::array_t&& arr)
+    {
+        j.m_type = value_t::array;
+        j.m_value = std::move(arr);
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType, typename CompatibleArrayType,
+             enable_if_t<not std::is_same<CompatibleArrayType, typename BasicJsonType::array_t>::value,
+                         int> = 0>
+    static void construct(BasicJsonType& j, const CompatibleArrayType& arr)
+    {
+        using std::begin;
+        using std::end;
+        j.m_type = value_t::array;
+        j.m_value.array = j.template create<typename BasicJsonType::array_t>(begin(arr), end(arr));
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const std::vector<bool>& arr)
+    {
+        j.m_type = value_t::array;
+        j.m_value = value_t::array;
+        j.m_value.array->reserve(arr.size());
+        for (const bool x : arr)
+        {
+            j.m_value.array->push_back(x);
+        }
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType, typename T,
+             enable_if_t<std::is_convertible<T, BasicJsonType>::value, int> = 0>
+    static void construct(BasicJsonType& j, const std::valarray<T>& arr)
+    {
+        j.m_type = value_t::array;
+        j.m_value = value_t::array;
+        j.m_value.array->resize(arr.size());
+        std::copy(std::begin(arr), std::end(arr), j.m_value.array->begin());
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::object>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::object_t& obj)
+    {
+        j.m_type = value_t::object;
+        j.m_value = obj;
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::object_t&& obj)
+    {
+        j.m_type = value_t::object;
+        j.m_value = std::move(obj);
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType, typename CompatibleObjectType,
+             enable_if_t<not std::is_same<CompatibleObjectType, typename BasicJsonType::object_t>::value, int> = 0>
+    static void construct(BasicJsonType& j, const CompatibleObjectType& obj)
+    {
+        using std::begin;
+        using std::end;
+
+        j.m_type = value_t::object;
+        j.m_value.object = j.template create<typename BasicJsonType::object_t>(begin(obj), end(obj));
+        j.assert_invariant();
+    }
+};
+
+/////////////
+// to_json //
+/////////////
+
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_same<T, typename BasicJsonType::boolean_t>::value, int> = 0>
+void to_json(BasicJsonType& j, T b) noexcept
+{
+    external_constructor<value_t::boolean>::construct(j, b);
+}
+
+template<typename BasicJsonType, typename CompatibleString,
+         enable_if_t<std::is_constructible<typename BasicJsonType::string_t, CompatibleString>::value, int> = 0>
+void to_json(BasicJsonType& j, const CompatibleString& s)
+{
+    external_constructor<value_t::string>::construct(j, s);
+}
+
+template<typename BasicJsonType>
+void to_json(BasicJsonType& j, typename BasicJsonType::string_t&& s)
+{
+    external_constructor<value_t::string>::construct(j, std::move(s));
+}
+
+template<typename BasicJsonType, typename FloatType,
+         enable_if_t<std::is_floating_point<FloatType>::value, int> = 0>
+void to_json(BasicJsonType& j, FloatType val) noexcept
+{
+    external_constructor<value_t::number_float>::construct(j, static_cast<typename BasicJsonType::number_float_t>(val));
+}
+
+template<typename BasicJsonType, typename CompatibleNumberUnsignedType,
+         enable_if_t<is_compatible_integer_type<typename BasicJsonType::number_unsigned_t, CompatibleNumberUnsignedType>::value, int> = 0>
+void to_json(BasicJsonType& j, CompatibleNumberUnsignedType val) noexcept
+{
+    external_constructor<value_t::number_unsigned>::construct(j, static_cast<typename BasicJsonType::number_unsigned_t>(val));
+}
+
+template<typename BasicJsonType, typename CompatibleNumberIntegerType,
+         enable_if_t<is_compatible_integer_type<typename BasicJsonType::number_integer_t, CompatibleNumberIntegerType>::value, int> = 0>
+void to_json(BasicJsonType& j, CompatibleNumberIntegerType val) noexcept
+{
+    external_constructor<value_t::number_integer>::construct(j, static_cast<typename BasicJsonType::number_integer_t>(val));
+}
+
+template<typename BasicJsonType, typename EnumType,
+         enable_if_t<std::is_enum<EnumType>::value, int> = 0>
+void to_json(BasicJsonType& j, EnumType e) noexcept
+{
+    using underlying_type = typename std::underlying_type<EnumType>::type;
+    external_constructor<value_t::number_integer>::construct(j, static_cast<underlying_type>(e));
+}
+
+template<typename BasicJsonType>
+void to_json(BasicJsonType& j, const std::vector<bool>& e)
+{
+    external_constructor<value_t::array>::construct(j, e);
+}
+
+template<typename BasicJsonType, typename CompatibleArrayType,
+         enable_if_t<is_compatible_array_type<BasicJsonType, CompatibleArrayType>::value or
+                     std::is_same<typename BasicJsonType::array_t, CompatibleArrayType>::value,
+                     int> = 0>
+void to_json(BasicJsonType& j, const CompatibleArrayType& arr)
+{
+    external_constructor<value_t::array>::construct(j, arr);
+}
+
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_convertible<T, BasicJsonType>::value, int> = 0>
+void to_json(BasicJsonType& j, std::valarray<T> arr)
+{
+    external_constructor<value_t::array>::construct(j, std::move(arr));
+}
+
+template<typename BasicJsonType>
+void to_json(BasicJsonType& j, typename BasicJsonType::array_t&& arr)
+{
+    external_constructor<value_t::array>::construct(j, std::move(arr));
+}
+
+template<typename BasicJsonType, typename CompatibleObjectType,
+         enable_if_t<is_compatible_object_type<BasicJsonType, CompatibleObjectType>::value, int> = 0>
+void to_json(BasicJsonType& j, const CompatibleObjectType& obj)
+{
+    external_constructor<value_t::object>::construct(j, obj);
+}
+
+template<typename BasicJsonType>
+void to_json(BasicJsonType& j, typename BasicJsonType::object_t&& obj)
+{
+    external_constructor<value_t::object>::construct(j, std::move(obj));
+}
+
+template<typename BasicJsonType, typename T, std::size_t N,
+         enable_if_t<not std::is_constructible<typename BasicJsonType::string_t, T (&)[N]>::value, int> = 0>
+void to_json(BasicJsonType& j, T (&arr)[N])
+{
+    external_constructor<value_t::array>::construct(j, arr);
+}
+
+template<typename BasicJsonType, typename... Args>
+void to_json(BasicJsonType& j, const std::pair<Args...>& p)
+{
+    j = {p.first, p.second};
+}
+
+template<typename BasicJsonType, typename Tuple, std::size_t... Idx>
+void to_json_tuple_impl(BasicJsonType& j, const Tuple& t, index_sequence<Idx...>)
+{
+    j = {std::get<Idx>(t)...};
+}
+
+template<typename BasicJsonType, typename... Args>
+void to_json(BasicJsonType& j, const std::tuple<Args...>& t)
+{
+    to_json_tuple_impl(j, t, index_sequence_for<Args...> {});
+}
+
+struct to_json_fn
+{
+  private:
+    template<typename BasicJsonType, typename T>
+    auto call(BasicJsonType& j, T&& val, priority_tag<1> /*unused*/) const noexcept(noexcept(to_json(j, std::forward<T>(val))))
+    -> decltype(to_json(j, std::forward<T>(val)), void())
+    {
+        return to_json(j, std::forward<T>(val));
+    }
+
+    template<typename BasicJsonType, typename T>
+    void call(BasicJsonType& /*unused*/, T&& /*unused*/, priority_tag<0> /*unused*/) const noexcept
+    {
+        static_assert(sizeof(BasicJsonType) == 0,
+                      "could not find to_json() method in T's namespace");
+
+#ifdef _MSC_VER
+        // MSVC does not show a stacktrace for the above assert
+        using decayed = uncvref_t<T>;
+        static_assert(sizeof(typename decayed::force_msvc_stacktrace) == 0,
+                      "forcing MSVC stacktrace to show which T we're talking about.");
+#endif
+    }
+
+  public:
+    template<typename BasicJsonType, typename T>
+    void operator()(BasicJsonType& j, T&& val) const
+    noexcept(noexcept(std::declval<to_json_fn>().call(j, std::forward<T>(val), priority_tag<1> {})))
+    {
+        return call(j, std::forward<T>(val), priority_tag<1> {});
+    }
+};
+}
+
+/// namespace to hold default `to_json` function
+namespace
+{
+constexpr const auto& to_json = detail::static_const<detail::to_json_fn>::value;
+}
+}

src/lib/json/detail/exceptions.hpp

@@ -0,0 +1,330 @@
+#pragma once
+
+#include <exception> // exception
+#include <stdexcept> // runtime_error
+#include <string> // to_string
+
+namespace nlohmann
+{
+namespace detail
+{
+////////////////
+// exceptions //
+////////////////
+
+/*!
+@brief general exception of the @ref basic_json class
+
+This class is an extension of `std::exception` objects with a member @a id for
+exception ids. It is used as the base class for all exceptions thrown by the
+@ref basic_json class. This class can hence be used as "wildcard" to catch
+exceptions.
+
+Subclasses:
+- @ref parse_error for exceptions indicating a parse error
+- @ref invalid_iterator for exceptions indicating errors with iterators
+- @ref type_error for exceptions indicating executing a member function with
+                  a wrong type
+- @ref out_of_range for exceptions indicating access out of the defined range
+- @ref other_error for exceptions indicating other library errors
+
+@internal
+@note To have nothrow-copy-constructible exceptions, we internally use
+      `std::runtime_error` which can cope with arbitrary-length error messages.
+      Intermediate strings are built with static functions and then passed to
+      the actual constructor.
+@endinternal
+
+@liveexample{The following code shows how arbitrary library exceptions can be
+caught.,exception}
+
+@since version 3.0.0
+*/
+class exception : public std::exception
+{
+  public:
+    /// returns the explanatory string
+    const char* what() const noexcept override
+    {
+        return m.what();
+    }
+
+    /// the id of the exception
+    const int id;
+
+  protected:
+    exception(int id_, const char* what_arg) : id(id_), m(what_arg) {}
+
+    static std::string name(const std::string& ename, int id_)
+    {
+        return "[json.exception." + ename + "." + std::to_string(id_) + "] ";
+    }
+
+  private:
+    /// an exception object as storage for error messages
+    std::runtime_error m;
+};
+
+/*!
+@brief exception indicating a parse error
+
+This exception is thrown by the library when a parse error occurs. Parse errors
+can occur during the deserialization of JSON text, CBOR, MessagePack, as well
+as when using JSON Patch.
+
+Member @a byte holds the byte index of the last read character in the input
+file.
+
+Exceptions have ids 1xx.
+
+name / id                      | example message | description
+------------------------------ | --------------- | -------------------------
+json.exception.parse_error.101 | parse error at 2: unexpected end of input; expected string literal | This error indicates a syntax error while deserializing a JSON text. The error message describes that an unexpected token (character) was encountered, and the member @a byte indicates the error position.
+json.exception.parse_error.102 | parse error at 14: missing or wrong low surrogate | JSON uses the `\uxxxx` format to describe Unicode characters. Code points above above 0xFFFF are split into two `\uxxxx` entries ("surrogate pairs"). This error indicates that the surrogate pair is incomplete or contains an invalid code point.
+json.exception.parse_error.103 | parse error: code points above 0x10FFFF are invalid | Unicode supports code points up to 0x10FFFF. Code points above 0x10FFFF are invalid.
+json.exception.parse_error.104 | parse error: JSON patch must be an array of objects | [RFC 6902](https://tools.ietf.org/html/rfc6902) requires a JSON Patch document to be a JSON document that represents an array of objects.
+json.exception.parse_error.105 | parse error: operation must have string member 'op' | An operation of a JSON Patch document must contain exactly one "op" member, whose value indicates the operation to perform. Its value must be one of "add", "remove", "replace", "move", "copy", or "test"; other values are errors.
+json.exception.parse_error.106 | parse error: array index '01' must not begin with '0' | An array index in a JSON Pointer ([RFC 6901](https://tools.ietf.org/html/rfc6901)) may be `0` or any number without a leading `0`.
+json.exception.parse_error.107 | parse error: JSON pointer must be empty or begin with '/' - was: 'foo' | A JSON Pointer must be a Unicode string containing a sequence of zero or more reference tokens, each prefixed by a `/` character.
+json.exception.parse_error.108 | parse error: escape character '~' must be followed with '0' or '1' | In a JSON Pointer, only `~0` and `~1` are valid escape sequences.
+json.exception.parse_error.109 | parse error: array index 'one' is not a number | A JSON Pointer array index must be a number.
+json.exception.parse_error.110 | parse error at 1: cannot read 2 bytes from vector | When parsing CBOR or MessagePack, the byte vector ends before the complete value has been read.
+json.exception.parse_error.112 | parse error at 1: error reading CBOR; last byte: 0xF8 | Not all types of CBOR or MessagePack are supported. This exception occurs if an unsupported byte was read.
+json.exception.parse_error.113 | parse error at 2: expected a CBOR string; last byte: 0x98 | While parsing a map key, a value that is not a string has been read.
+
+@note For an input with n bytes, 1 is the index of the first character and n+1
+      is the index of the terminating null byte or the end of file. This also
+      holds true when reading a byte vector (CBOR or MessagePack).
+
+@liveexample{The following code shows how a `parse_error` exception can be
+caught.,parse_error}
+
+@sa @ref exception for the base class of the library exceptions
+@sa @ref invalid_iterator for exceptions indicating errors with iterators
+@sa @ref type_error for exceptions indicating executing a member function with
+                    a wrong type
+@sa @ref out_of_range for exceptions indicating access out of the defined range
+@sa @ref other_error for exceptions indicating other library errors
+
+@since version 3.0.0
+*/
+class parse_error : public exception
+{
+  public:
+    /*!
+    @brief create a parse error exception
+    @param[in] id_       the id of the exception
+    @param[in] byte_     the byte index where the error occurred (or 0 if the
+                         position cannot be determined)
+    @param[in] what_arg  the explanatory string
+    @return parse_error object
+    */
+    static parse_error create(int id_, std::size_t byte_, const std::string& what_arg)
+    {
+        std::string w = exception::name("parse_error", id_) + "parse error" +
+                        (byte_ != 0 ? (" at " + std::to_string(byte_)) : "") +
+                        ": " + what_arg;
+        return parse_error(id_, byte_, w.c_str());
+    }
+
+    /*!
+    @brief byte index of the parse error
+
+    The byte index of the last read character in the input file.
+
+    @note For an input with n bytes, 1 is the index of the first character and
+          n+1 is the index of the terminating null byte or the end of file.
+          This also holds true when reading a byte vector (CBOR or MessagePack).
+    */
+    const std::size_t byte;
+
+  private:
+    parse_error(int id_, std::size_t byte_, const char* what_arg)
+        : exception(id_, what_arg), byte(byte_) {}
+};
+
+/*!
+@brief exception indicating errors with iterators
+
+This exception is thrown if iterators passed to a library function do not match
+the expected semantics.
+
+Exceptions have ids 2xx.
+
+name / id                           | example message | description
+----------------------------------- | --------------- | -------------------------
+json.exception.invalid_iterator.201 | iterators are not compatible | The iterators passed to constructor @ref basic_json(InputIT first, InputIT last) are not compatible, meaning they do not belong to the same container. Therefore, the range (@a first, @a last) is invalid.
+json.exception.invalid_iterator.202 | iterator does not fit current value | In an erase or insert function, the passed iterator @a pos does not belong to the JSON value for which the function was called. It hence does not define a valid position for the deletion/insertion.
+json.exception.invalid_iterator.203 | iterators do not fit current value | Either iterator passed to function @ref erase(IteratorType first, IteratorType last) does not belong to the JSON value from which values shall be erased. It hence does not define a valid range to delete values from.
+json.exception.invalid_iterator.204 | iterators out of range | When an iterator range for a primitive type (number, boolean, or string) is passed to a constructor or an erase function, this range has to be exactly (@ref begin(), @ref end()), because this is the only way the single stored value is expressed. All other ranges are invalid.
+json.exception.invalid_iterator.205 | iterator out of range | When an iterator for a primitive type (number, boolean, or string) is passed to an erase function, the iterator has to be the @ref begin() iterator, because it is the only way to address the stored value. All other iterators are invalid.
+json.exception.invalid_iterator.206 | cannot construct with iterators from null | The iterators passed to constructor @ref basic_json(InputIT first, InputIT last) belong to a JSON null value and hence to not define a valid range.
+json.exception.invalid_iterator.207 | cannot use key() for non-object iterators | The key() member function can only be used on iterators belonging to a JSON object, because other types do not have a concept of a key.
+json.exception.invalid_iterator.208 | cannot use operator[] for object iterators | The operator[] to specify a concrete offset cannot be used on iterators belonging to a JSON object, because JSON objects are unordered.
+json.exception.invalid_iterator.209 | cannot use offsets with object iterators | The offset operators (+, -, +=, -=) cannot be used on iterators belonging to a JSON object, because JSON objects are unordered.
+json.exception.invalid_iterator.210 | iterators do not fit | The iterator range passed to the insert function are not compatible, meaning they do not belong to the same container. Therefore, the range (@a first, @a last) is invalid.
+json.exception.invalid_iterator.211 | passed iterators may not belong to container | The iterator range passed to the insert function must not be a subrange of the container to insert to.
+json.exception.invalid_iterator.212 | cannot compare iterators of different containers | When two iterators are compared, they must belong to the same container.
+json.exception.invalid_iterator.213 | cannot compare order of object iterators | The order of object iterators cannot be compared, because JSON objects are unordered.
+json.exception.invalid_iterator.214 | cannot get value | Cannot get value for iterator: Either the iterator belongs to a null value or it is an iterator to a primitive type (number, boolean, or string), but the iterator is different to @ref begin().
+
+@liveexample{The following code shows how an `invalid_iterator` exception can be
+caught.,invalid_iterator}
+
+@sa @ref exception for the base class of the library exceptions
+@sa @ref parse_error for exceptions indicating a parse error
+@sa @ref type_error for exceptions indicating executing a member function with
+                    a wrong type
+@sa @ref out_of_range for exceptions indicating access out of the defined range
+@sa @ref other_error for exceptions indicating other library errors
+
+@since version 3.0.0
+*/
+class invalid_iterator : public exception
+{
+  public:
+    static invalid_iterator create(int id_, const std::string& what_arg)
+    {
+        std::string w = exception::name("invalid_iterator", id_) + what_arg;
+        return invalid_iterator(id_, w.c_str());
+    }
+
+  private:
+    invalid_iterator(int id_, const char* what_arg)
+        : exception(id_, what_arg) {}
+};
+
+/*!
+@brief exception indicating executing a member function with a wrong type
+
+This exception is thrown in case of a type error; that is, a library function is
+executed on a JSON value whose type does not match the expected semantics.
+
+Exceptions have ids 3xx.
+
+name / id                     | example message | description
+----------------------------- | --------------- | -------------------------
+json.exception.type_error.301 | cannot create object from initializer list | To create an object from an initializer list, the initializer list must consist only of a list of pairs whose first element is a string. When this constraint is violated, an array is created instead.
+json.exception.type_error.302 | type must be object, but is array | During implicit or explicit value conversion, the JSON type must be compatible to the target type. For instance, a JSON string can only be converted into string types, but not into numbers or boolean types.
+json.exception.type_error.303 | incompatible ReferenceType for get_ref, actual type is object | To retrieve a reference to a value stored in a @ref basic_json object with @ref get_ref, the type of the reference must match the value type. For instance, for a JSON array, the @a ReferenceType must be @ref array_t&.
+json.exception.type_error.304 | cannot use at() with string | The @ref at() member functions can only be executed for certain JSON types.
+json.exception.type_error.305 | cannot use operator[] with string | The @ref operator[] member functions can only be executed for certain JSON types.
+json.exception.type_error.306 | cannot use value() with string | The @ref value() member functions can only be executed for certain JSON types.
+json.exception.type_error.307 | cannot use erase() with string | The @ref erase() member functions can only be executed for certain JSON types.
+json.exception.type_error.308 | cannot use push_back() with string | The @ref push_back() and @ref operator+= member functions can only be executed for certain JSON types.
+json.exception.type_error.309 | cannot use insert() with | The @ref insert() member functions can only be executed for certain JSON types.
+json.exception.type_error.310 | cannot use swap() with number | The @ref swap() member functions can only be executed for certain JSON types.
+json.exception.type_error.311 | cannot use emplace_back() with string | The @ref emplace_back() member function can only be executed for certain JSON types.
+json.exception.type_error.312 | cannot use update() with string | The @ref update() member functions can only be executed for certain JSON types.
+json.exception.type_error.313 | invalid value to unflatten | The @ref unflatten function converts an object whose keys are JSON Pointers back into an arbitrary nested JSON value. The JSON Pointers must not overlap, because then the resulting value would not be well defined.
+json.exception.type_error.314 | only objects can be unflattened | The @ref unflatten function only works for an object whose keys are JSON Pointers.
+json.exception.type_error.315 | values in object must be primitive | The @ref unflatten function only works for an object whose keys are JSON Pointers and whose values are primitive.
+json.exception.type_error.316 | invalid UTF-8 byte at index 10: 0x7E | The @ref dump function only works with UTF-8 encoded strings; that is, if you assign a `std::string` to a JSON value, make sure it is UTF-8 encoded. |
+
+@liveexample{The following code shows how a `type_error` exception can be
+caught.,type_error}
+
+@sa @ref exception for the base class of the library exceptions
+@sa @ref parse_error for exceptions indicating a parse error
+@sa @ref invalid_iterator for exceptions indicating errors with iterators
+@sa @ref out_of_range for exceptions indicating access out of the defined range
+@sa @ref other_error for exceptions indicating other library errors
+
+@since version 3.0.0
+*/
+class type_error : public exception
+{
+  public:
+    static type_error create(int id_, const std::string& what_arg)
+    {
+        std::string w = exception::name("type_error", id_) + what_arg;
+        return type_error(id_, w.c_str());
+    }
+
+  private:
+    type_error(int id_, const char* what_arg) : exception(id_, what_arg) {}
+};
+
+/*!
+@brief exception indicating access out of the defined range
+
+This exception is thrown in case a library function is called on an input
+parameter that exceeds the expected range, for instance in case of array
+indices or nonexisting object keys.
+
+Exceptions have ids 4xx.
+
+name / id                       | example message | description
+------------------------------- | --------------- | -------------------------
+json.exception.out_of_range.401 | array index 3 is out of range | The provided array index @a i is larger than @a size-1.
+json.exception.out_of_range.402 | array index '-' (3) is out of range | The special array index `-` in a JSON Pointer never describes a valid element of the array, but the index past the end. That is, it can only be used to add elements at this position, but not to read it.
+json.exception.out_of_range.403 | key 'foo' not found | The provided key was not found in the JSON object.
+json.exception.out_of_range.404 | unresolved reference token 'foo' | A reference token in a JSON Pointer could not be resolved.
+json.exception.out_of_range.405 | JSON pointer has no parent | The JSON Patch operations 'remove' and 'add' can not be applied to the root element of the JSON value.
+json.exception.out_of_range.406 | number overflow parsing '10E1000' | A parsed number could not be stored as without changing it to NaN or INF.
+json.exception.out_of_range.407 | number overflow serializing '9223372036854775808' | UBJSON only supports integers numbers up to 9223372036854775807. |
+json.exception.out_of_range.408 | excessive array size: 8658170730974374167 | The size (following `#`) of an UBJSON array or object exceeds the maximal capacity. |
+
+@liveexample{The following code shows how an `out_of_range` exception can be
+caught.,out_of_range}
+
+@sa @ref exception for the base class of the library exceptions
+@sa @ref parse_error for exceptions indicating a parse error
+@sa @ref invalid_iterator for exceptions indicating errors with iterators
+@sa @ref type_error for exceptions indicating executing a member function with
+                    a wrong type
+@sa @ref other_error for exceptions indicating other library errors
+
+@since version 3.0.0
+*/
+class out_of_range : public exception
+{
+  public:
+    static out_of_range create(int id_, const std::string& what_arg)
+    {
+        std::string w = exception::name("out_of_range", id_) + what_arg;
+        return out_of_range(id_, w.c_str());
+    }
+
+  private:
+    out_of_range(int id_, const char* what_arg) : exception(id_, what_arg) {}
+};
+
+/*!
+@brief exception indicating other library errors
+
+This exception is thrown in case of errors that cannot be classified with the
+other exception types.
+
+Exceptions have ids 5xx.
+
+name / id                      | example message | description
+------------------------------ | --------------- | -------------------------
+json.exception.other_error.501 | unsuccessful: {"op":"test","path":"/baz", "value":"bar"} | A JSON Patch operation 'test' failed. The unsuccessful operation is also printed.
+
+@sa @ref exception for the base class of the library exceptions
+@sa @ref parse_error for exceptions indicating a parse error
+@sa @ref invalid_iterator for exceptions indicating errors with iterators
+@sa @ref type_error for exceptions indicating executing a member function with
+                    a wrong type
+@sa @ref out_of_range for exceptions indicating access out of the defined range
+
+@liveexample{The following code shows how an `other_error` exception can be
+caught.,other_error}
+
+@since version 3.0.0
+*/
+class other_error : public exception
+{
+  public:
+    static other_error create(int id_, const std::string& what_arg)
+    {
+        std::string w = exception::name("other_error", id_) + what_arg;
+        return other_error(id_, w.c_str());
+    }
+
+  private:
+    other_error(int id_, const char* what_arg) : exception(id_, what_arg) {}
+};
+}
+}

src/lib/json/detail/input/input_adapters.hpp

@@ -0,0 +1,262 @@
+#pragma once
+
+#include <algorithm> // min
+#include <array> // array
+#include <cassert> // assert
+#include <cstddef> // size_t
+#include <cstring> // strlen
+#include <ios> // streamsize, streamoff, streampos
+#include <istream> // istream
+#include <iterator> // begin, end, iterator_traits, random_access_iterator_tag, distance, next
+#include <memory> // shared_ptr, make_shared, addressof
+#include <numeric> // accumulate
+#include <string> // string, char_traits
+#include <type_traits> // enable_if, is_base_of, is_pointer, is_integral, remove_pointer
+#include <utility> // pair, declval
+
+#include <nlohmann/detail/macro_scope.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+////////////////////
+// input adapters //
+////////////////////
+
+/*!
+@brief abstract input adapter interface
+
+Produces a stream of std::char_traits<char>::int_type characters from a
+std::istream, a buffer, or some other input type.  Accepts the return of exactly
+one non-EOF character for future input.  The int_type characters returned
+consist of all valid char values as positive values (typically unsigned char),
+plus an EOF value outside that range, specified by the value of the function
+std::char_traits<char>::eof().  This value is typically -1, but could be any
+arbitrary value which is not a valid char value.
+*/
+struct input_adapter_protocol
+{
+    /// get a character [0,255] or std::char_traits<char>::eof().
+    virtual std::char_traits<char>::int_type get_character() = 0;
+    /// restore the last non-eof() character to input
+    virtual void unget_character() = 0;
+    virtual ~input_adapter_protocol() = default;
+};
+
+/// a type to simplify interfaces
+using input_adapter_t = std::shared_ptr<input_adapter_protocol>;
+
+/*!
+Input adapter for a (caching) istream. Ignores a UFT Byte Order Mark at
+beginning of input. Does not support changing the underlying std::streambuf
+in mid-input. Maintains underlying std::istream and std::streambuf to support
+subsequent use of standard std::istream operations to process any input
+characters following those used in parsing the JSON input.  Clears the
+std::istream flags; any input errors (e.g., EOF) will be detected by the first
+subsequent call for input from the std::istream.
+*/
+class input_stream_adapter : public input_adapter_protocol
+{
+  public:
+    ~input_stream_adapter() override
+    {
+        // clear stream flags; we use underlying streambuf I/O, do not
+        // maintain ifstream flags
+        is.clear();
+    }
+
+    explicit input_stream_adapter(std::istream& i)
+        : is(i), sb(*i.rdbuf())
+    {
+        // skip byte order mark
+        std::char_traits<char>::int_type c;
+        if ((c = get_character()) == 0xEF)
+        {
+            if ((c = get_character()) == 0xBB)
+            {
+                if ((c = get_character()) == 0xBF)
+                {
+                    return; // Ignore BOM
+                }
+                else if (c != std::char_traits<char>::eof())
+                {
+                    is.unget();
+                }
+                is.putback('\xBB');
+            }
+            else if (c != std::char_traits<char>::eof())
+            {
+                is.unget();
+            }
+            is.putback('\xEF');
+        }
+        else if (c != std::char_traits<char>::eof())
+        {
+            is.unget(); // no byte order mark; process as usual
+        }
+    }
+
+    // delete because of pointer members
+    input_stream_adapter(const input_stream_adapter&) = delete;
+    input_stream_adapter& operator=(input_stream_adapter&) = delete;
+
+    // std::istream/std::streambuf use std::char_traits<char>::to_int_type, to
+    // ensure that std::char_traits<char>::eof() and the character 0xFF do not
+    // end up as the same value, eg. 0xFFFFFFFF.
+    std::char_traits<char>::int_type get_character() override
+    {
+        return sb.sbumpc();
+    }
+
+    void unget_character() override
+    {
+        sb.sungetc();  // is.unget() avoided for performance
+    }
+
+  private:
+    /// the associated input stream
+    std::istream& is;
+    std::streambuf& sb;
+};
+
+/// input adapter for buffer input
+class input_buffer_adapter : public input_adapter_protocol
+{
+  public:
+    input_buffer_adapter(const char* b, const std::size_t l)
+        : cursor(b), limit(b + l), start(b)
+    {
+        // skip byte order mark
+        if (l >= 3 and b[0] == '\xEF' and b[1] == '\xBB' and b[2] == '\xBF')
+        {
+            cursor += 3;
+        }
+    }
+
+    // delete because of pointer members
+    input_buffer_adapter(const input_buffer_adapter&) = delete;
+    input_buffer_adapter& operator=(input_buffer_adapter&) = delete;
+
+    std::char_traits<char>::int_type get_character() noexcept override
+    {
+        if (JSON_LIKELY(cursor < limit))
+        {
+            return std::char_traits<char>::to_int_type(*(cursor++));
+        }
+
+        return std::char_traits<char>::eof();
+    }
+
+    void unget_character() noexcept override
+    {
+        if (JSON_LIKELY(cursor > start))
+        {
+            --cursor;
+        }
+    }
+
+  private:
+    /// pointer to the current character
+    const char* cursor;
+    /// pointer past the last character
+    const char* limit;
+    /// pointer to the first character
+    const char* start;
+};
+
+class input_adapter
+{
+  public:
+    // native support
+
+    /// input adapter for input stream
+    input_adapter(std::istream& i)
+        : ia(std::make_shared<input_stream_adapter>(i)) {}
+
+    /// input adapter for input stream
+    input_adapter(std::istream&& i)
+        : ia(std::make_shared<input_stream_adapter>(i)) {}
+
+    /// input adapter for buffer
+    template<typename CharT,
+             typename std::enable_if<
+                 std::is_pointer<CharT>::value and
+                 std::is_integral<typename std::remove_pointer<CharT>::type>::value and
+                 sizeof(typename std::remove_pointer<CharT>::type) == 1,
+                 int>::type = 0>
+    input_adapter(CharT b, std::size_t l)
+        : ia(std::make_shared<input_buffer_adapter>(reinterpret_cast<const char*>(b), l)) {}
+
+    // derived support
+
+    /// input adapter for string literal
+    template<typename CharT,
+             typename std::enable_if<
+                 std::is_pointer<CharT>::value and
+                 std::is_integral<typename std::remove_pointer<CharT>::type>::value and
+                 sizeof(typename std::remove_pointer<CharT>::type) == 1,
+                 int>::type = 0>
+    input_adapter(CharT b)
+        : input_adapter(reinterpret_cast<const char*>(b),
+                        std::strlen(reinterpret_cast<const char*>(b))) {}
+
+    /// input adapter for iterator range with contiguous storage
+    template<class IteratorType,
+             typename std::enable_if<
+                 std::is_same<typename std::iterator_traits<IteratorType>::iterator_category, std::random_access_iterator_tag>::value,
+                 int>::type = 0>
+    input_adapter(IteratorType first, IteratorType last)
+    {
+        // assertion to check that the iterator range is indeed contiguous,
+        // see http://stackoverflow.com/a/35008842/266378 for more discussion
+        assert(std::accumulate(
+                   first, last, std::pair<bool, int>(true, 0),
+                   [&first](std::pair<bool, int> res, decltype(*first) val)
+        {
+            res.first &= (val == *(std::next(std::addressof(*first), res.second++)));
+            return res;
+        }).first);
+
+        // assertion to check that each element is 1 byte long
+        static_assert(
+            sizeof(typename std::iterator_traits<IteratorType>::value_type) == 1,
+            "each element in the iterator range must have the size of 1 byte");
+
+        const auto len = static_cast<size_t>(std::distance(first, last));
+        if (JSON_LIKELY(len > 0))
+        {
+            // there is at least one element: use the address of first
+            ia = std::make_shared<input_buffer_adapter>(reinterpret_cast<const char*>(&(*first)), len);
+        }
+        else
+        {
+            // the address of first cannot be used: use nullptr
+            ia = std::make_shared<input_buffer_adapter>(nullptr, len);
+        }
+    }
+
+    /// input adapter for array
+    template<class T, std::size_t N>
+    input_adapter(T (&array)[N])
+        : input_adapter(std::begin(array), std::end(array)) {}
+
+    /// input adapter for contiguous container
+    template<class ContiguousContainer, typename
+             std::enable_if<not std::is_pointer<ContiguousContainer>::value and
+                            std::is_base_of<std::random_access_iterator_tag, typename std::iterator_traits<decltype(std::begin(std::declval<ContiguousContainer const>()))>::iterator_category>::value,
+                            int>::type = 0>
+    input_adapter(const ContiguousContainer& c)
+        : input_adapter(std::begin(c), std::end(c)) {}
+
+    operator input_adapter_t()
+    {
+        return ia;
+    }
+
+  private:
+    /// the actual adapter
+    input_adapter_t ia = nullptr;
+};
+}
+}

src/lib/json/detail/input/parser.hpp

@@ -0,0 +1,591 @@
+#pragma once
+
+#include <cassert> // assert
+#include <cmath> // isfinite
+#include <cstdint> // uint8_t
+#include <functional> // function
+#include <string> // string
+#include <utility> // move
+
+#include <nlohmann/detail/exceptions.hpp>
+#include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/input/input_adapters.hpp>
+#include <nlohmann/detail/input/lexer.hpp>
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+////////////
+// parser //
+////////////
+
+/*!
+@brief syntax analysis
+
+This class implements a recursive decent parser.
+*/
+template<typename BasicJsonType>
+class parser
+{
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using lexer_t = lexer<BasicJsonType>;
+    using token_type = typename lexer_t::token_type;
+
+  public:
+    enum class parse_event_t : uint8_t
+    {
+        /// the parser read `{` and started to process a JSON object
+        object_start,
+        /// the parser read `}` and finished processing a JSON object
+        object_end,
+        /// the parser read `[` and started to process a JSON array
+        array_start,
+        /// the parser read `]` and finished processing a JSON array
+        array_end,
+        /// the parser read a key of a value in an object
+        key,
+        /// the parser finished reading a JSON value
+        value
+    };
+
+    using parser_callback_t =
+        std::function<bool(int depth, parse_event_t event, BasicJsonType& parsed)>;
+
+    /// a parser reading from an input adapter
+    explicit parser(detail::input_adapter_t adapter,
+                    const parser_callback_t cb = nullptr,
+                    const bool allow_exceptions_ = true)
+        : callback(cb), m_lexer(adapter), allow_exceptions(allow_exceptions_)
+    {}
+
+    /*!
+    @brief public parser interface
+
+    @param[in] strict      whether to expect the last token to be EOF
+    @param[in,out] result  parsed JSON value
+
+    @throw parse_error.101 in case of an unexpected token
+    @throw parse_error.102 if to_unicode fails or surrogate error
+    @throw parse_error.103 if to_unicode fails
+    */
+    void parse(const bool strict, BasicJsonType& result)
+    {
+        // read first token
+        get_token();
+
+        parse_internal(true, result);
+        result.assert_invariant();
+
+        // in strict mode, input must be completely read
+        if (strict)
+        {
+            get_token();
+            expect(token_type::end_of_input);
+        }
+
+        // in case of an error, return discarded value
+        if (errored)
+        {
+            result = value_t::discarded;
+            return;
+        }
+
+        // set top-level value to null if it was discarded by the callback
+        // function
+        if (result.is_discarded())
+        {
+            result = nullptr;
+        }
+    }
+
+    /*!
+    @brief public accept interface
+
+    @param[in] strict  whether to expect the last token to be EOF
+    @return whether the input is a proper JSON text
+    */
+    bool accept(const bool strict = true)
+    {
+        // read first token
+        get_token();
+
+        if (not accept_internal())
+        {
+            return false;
+        }
+
+        // strict => last token must be EOF
+        return not strict or (get_token() == token_type::end_of_input);
+    }
+
+  private:
+    /*!
+    @brief the actual parser
+    @throw parse_error.101 in case of an unexpected token
+    @throw parse_error.102 if to_unicode fails or surrogate error
+    @throw parse_error.103 if to_unicode fails
+    */
+    void parse_internal(bool keep, BasicJsonType& result)
+    {
+        // never parse after a parse error was detected
+        assert(not errored);
+
+        // start with a discarded value
+        if (not result.is_discarded())
+        {
+            result.m_value.destroy(result.m_type);
+            result.m_type = value_t::discarded;
+        }
+
+        switch (last_token)
+        {
+            case token_type::begin_object:
+            {
+                if (keep)
+                {
+                    if (callback)
+                    {
+                        keep = callback(depth++, parse_event_t::object_start, result);
+                    }
+
+                    if (not callback or keep)
+                    {
+                        // explicitly set result to object to cope with {}
+                        result.m_type = value_t::object;
+                        result.m_value = value_t::object;
+                    }
+                }
+
+                // read next token
+                get_token();
+
+                // closing } -> we are done
+                if (last_token == token_type::end_object)
+                {
+                    if (keep and callback and not callback(--depth, parse_event_t::object_end, result))
+                    {
+                        result.m_value.destroy(result.m_type);
+                        result.m_type = value_t::discarded;
+                    }
+                    break;
+                }
+
+                // parse values
+                string_t key;
+                BasicJsonType value;
+                while (true)
+                {
+                    // store key
+                    if (not expect(token_type::value_string))
+                    {
+                        return;
+                    }
+                    key = m_lexer.move_string();
+
+                    bool keep_tag = false;
+                    if (keep)
+                    {
+                        if (callback)
+                        {
+                            BasicJsonType k(key);
+                            keep_tag = callback(depth, parse_event_t::key, k);
+                        }
+                        else
+                        {
+                            keep_tag = true;
+                        }
+                    }
+
+                    // parse separator (:)
+                    get_token();
+                    if (not expect(token_type::name_separator))
+                    {
+                        return;
+                    }
+
+                    // parse and add value
+                    get_token();
+                    value.m_value.destroy(value.m_type);
+                    value.m_type = value_t::discarded;
+                    parse_internal(keep, value);
+
+                    if (JSON_UNLIKELY(errored))
+                    {
+                        return;
+                    }
+
+                    if (keep and keep_tag and not value.is_discarded())
+                    {
+                        result.m_value.object->emplace(std::move(key), std::move(value));
+                    }
+
+                    // comma -> next value
+                    get_token();
+                    if (last_token == token_type::value_separator)
+                    {
+                        get_token();
+                        continue;
+                    }
+
+                    // closing }
+                    if (not expect(token_type::end_object))
+                    {
+                        return;
+                    }
+                    break;
+                }
+
+                if (keep and callback and not callback(--depth, parse_event_t::object_end, result))
+                {
+                    result.m_value.destroy(result.m_type);
+                    result.m_type = value_t::discarded;
+                }
+                break;
+            }
+
+            case token_type::begin_array:
+            {
+                if (keep)
+                {
+                    if (callback)
+                    {
+                        keep = callback(depth++, parse_event_t::array_start, result);
+                    }
+
+                    if (not callback or keep)
+                    {
+                        // explicitly set result to array to cope with []
+                        result.m_type = value_t::array;
+                        result.m_value = value_t::array;
+                    }
+                }
+
+                // read next token
+                get_token();
+
+                // closing ] -> we are done
+                if (last_token == token_type::end_array)
+                {
+                    if (callback and not callback(--depth, parse_event_t::array_end, result))
+                    {
+                        result.m_value.destroy(result.m_type);
+                        result.m_type = value_t::discarded;
+                    }
+                    break;
+                }
+
+                // parse values
+                BasicJsonType value;
+                while (true)
+                {
+                    // parse value
+                    value.m_value.destroy(value.m_type);
+                    value.m_type = value_t::discarded;
+                    parse_internal(keep, value);
+
+                    if (JSON_UNLIKELY(errored))
+                    {
+                        return;
+                    }
+
+                    if (keep and not value.is_discarded())
+                    {
+                        result.m_value.array->push_back(std::move(value));
+                    }
+
+                    // comma -> next value
+                    get_token();
+                    if (last_token == token_type::value_separator)
+                    {
+                        get_token();
+                        continue;
+                    }
+
+                    // closing ]
+                    if (not expect(token_type::end_array))
+                    {
+                        return;
+                    }
+                    break;
+                }
+
+                if (keep and callback and not callback(--depth, parse_event_t::array_end, result))
+                {
+                    result.m_value.destroy(result.m_type);
+                    result.m_type = value_t::discarded;
+                }
+                break;
+            }
+
+            case token_type::literal_null:
+            {
+                result.m_type = value_t::null;
+                break;
+            }
+
+            case token_type::value_string:
+            {
+                result.m_type = value_t::string;
+                result.m_value = m_lexer.move_string();
+                break;
+            }
+
+            case token_type::literal_true:
+            {
+                result.m_type = value_t::boolean;
+                result.m_value = true;
+                break;
+            }
+
+            case token_type::literal_false:
+            {
+                result.m_type = value_t::boolean;
+                result.m_value = false;
+                break;
+            }
+
+            case token_type::value_unsigned:
+            {
+                result.m_type = value_t::number_unsigned;
+                result.m_value = m_lexer.get_number_unsigned();
+                break;
+            }
+
+            case token_type::value_integer:
+            {
+                result.m_type = value_t::number_integer;
+                result.m_value = m_lexer.get_number_integer();
+                break;
+            }
+
+            case token_type::value_float:
+            {
+                result.m_type = value_t::number_float;
+                result.m_value = m_lexer.get_number_float();
+
+                // throw in case of infinity or NAN
+                if (JSON_UNLIKELY(not std::isfinite(result.m_value.number_float)))
+                {
+                    if (allow_exceptions)
+                    {
+                        JSON_THROW(out_of_range::create(406, "number overflow parsing '" +
+                                                        m_lexer.get_token_string() + "'"));
+                    }
+                    expect(token_type::uninitialized);
+                }
+                break;
+            }
+
+            case token_type::parse_error:
+            {
+                // using "uninitialized" to avoid "expected" message
+                if (not expect(token_type::uninitialized))
+                {
+                    return;
+                }
+                break; // LCOV_EXCL_LINE
+            }
+
+            default:
+            {
+                // the last token was unexpected; we expected a value
+                if (not expect(token_type::literal_or_value))
+                {
+                    return;
+                }
+                break; // LCOV_EXCL_LINE
+            }
+        }
+
+        if (keep and callback and not callback(depth, parse_event_t::value, result))
+        {
+            result.m_value.destroy(result.m_type);
+            result.m_type = value_t::discarded;
+        }
+    }
+
+    /*!
+    @brief the actual acceptor
+
+    @invariant 1. The last token is not yet processed. Therefore, the caller
+                  of this function must make sure a token has been read.
+               2. When this function returns, the last token is processed.
+                  That is, the last read character was already considered.
+
+    This invariant makes sure that no token needs to be "unput".
+    */
+    bool accept_internal()
+    {
+        switch (last_token)
+        {
+            case token_type::begin_object:
+            {
+                // read next token
+                get_token();
+
+                // closing } -> we are done
+                if (last_token == token_type::end_object)
+                {
+                    return true;
+                }
+
+                // parse values
+                while (true)
+                {
+                    // parse key
+                    if (last_token != token_type::value_string)
+                    {
+                        return false;
+                    }
+
+                    // parse separator (:)
+                    get_token();
+                    if (last_token != token_type::name_separator)
+                    {
+                        return false;
+                    }
+
+                    // parse value
+                    get_token();
+                    if (not accept_internal())
+                    {
+                        return false;
+                    }
+
+                    // comma -> next value
+                    get_token();
+                    if (last_token == token_type::value_separator)
+                    {
+                        get_token();
+                        continue;
+                    }
+
+                    // closing }
+                    return (last_token == token_type::end_object);
+                }
+            }
+
+            case token_type::begin_array:
+            {
+                // read next token
+                get_token();
+
+                // closing ] -> we are done
+                if (last_token == token_type::end_array)
+                {
+                    return true;
+                }
+
+                // parse values
+                while (true)
+                {
+                    // parse value
+                    if (not accept_internal())
+                    {
+                        return false;
+                    }
+
+                    // comma -> next value
+                    get_token();
+                    if (last_token == token_type::value_separator)
+                    {
+                        get_token();
+                        continue;
+                    }
+
+                    // closing ]
+                    return (last_token == token_type::end_array);
+                }
+            }
+
+            case token_type::value_float:
+            {
+                // reject infinity or NAN
+                return std::isfinite(m_lexer.get_number_float());
+            }
+
+            case token_type::literal_false:
+            case token_type::literal_null:
+            case token_type::literal_true:
+            case token_type::value_integer:
+            case token_type::value_string:
+            case token_type::value_unsigned:
+                return true;
+
+            default: // the last token was unexpected
+                return false;
+        }
+    }
+
+    /// get next token from lexer
+    token_type get_token()
+    {
+        return (last_token = m_lexer.scan());
+    }
+
+    /*!
+    @throw parse_error.101 if expected token did not occur
+    */
+    bool expect(token_type t)
+    {
+        if (JSON_UNLIKELY(t != last_token))
+        {
+            errored = true;
+            expected = t;
+            if (allow_exceptions)
+            {
+                throw_exception();
+            }
+            else
+            {
+                return false;
+            }
+        }
+
+        return true;
+    }
+
+    [[noreturn]] void throw_exception() const
+    {
+        std::string error_msg = "syntax error - ";
+        if (last_token == token_type::parse_error)
+        {
+            error_msg += std::string(m_lexer.get_error_message()) + "; last read: '" +
+                         m_lexer.get_token_string() + "'";
+        }
+        else
+        {
+            error_msg += "unexpected " + std::string(lexer_t::token_type_name(last_token));
+        }
+
+        if (expected != token_type::uninitialized)
+        {
+            error_msg += "; expected " + std::string(lexer_t::token_type_name(expected));
+        }
+
+        JSON_THROW(parse_error::create(101, m_lexer.get_position(), error_msg));
+    }
+
+  private:
+    /// current level of recursion
+    int depth = 0;
+    /// callback function
+    const parser_callback_t callback = nullptr;
+    /// the type of the last read token
+    token_type last_token = token_type::uninitialized;
+    /// the lexer
+    lexer_t m_lexer;
+    /// whether a syntax error occurred
+    bool errored = false;
+    /// possible reason for the syntax error
+    token_type expected = token_type::uninitialized;
+    /// whether to throw exceptions in case of errors
+    const bool allow_exceptions = true;
+};
+}
+}

src/lib/json/detail/iterators/internal_iterator.hpp

@@ -0,0 +1,25 @@
+#pragma once
+
+#include <nlohmann/detail/iterators/primitive_iterator.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+/*!
+@brief an iterator value
+
+@note This structure could easily be a union, but MSVC currently does not allow
+unions members with complex constructors, see https://github.com/nlohmann/json/pull/105.
+*/
+template<typename BasicJsonType> struct internal_iterator
+{
+    /// iterator for JSON objects
+    typename BasicJsonType::object_t::iterator object_iterator {};
+    /// iterator for JSON arrays
+    typename BasicJsonType::array_t::iterator array_iterator {};
+    /// generic iterator for all other types
+    primitive_iterator_t primitive_iterator {};
+};
+}
+}

src/lib/json/detail/iterators/iter_impl.hpp

@@ -0,0 +1,614 @@
+#pragma once
+
+#include <ciso646> // not
+#include <iterator> // iterator, random_access_iterator_tag, bidirectional_iterator_tag, advance, next
+#include <type_traits> // conditional, is_const, remove_const
+
+#include <nlohmann/detail/exceptions.hpp>
+#include <nlohmann/detail/iterators/internal_iterator.hpp>
+#include <nlohmann/detail/iterators/primitive_iterator.hpp>
+#include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/meta.hpp>
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+// forward declare, to be able to friend it later on
+template<typename IteratorType> class iteration_proxy;
+
+/*!
+@brief a template for a bidirectional iterator for the @ref basic_json class
+
+This class implements a both iterators (iterator and const_iterator) for the
+@ref basic_json class.
+
+@note An iterator is called *initialized* when a pointer to a JSON value has
+      been set (e.g., by a constructor or a copy assignment). If the iterator is
+      default-constructed, it is *uninitialized* and most methods are undefined.
+      **The library uses assertions to detect calls on uninitialized iterators.**
+
+@requirement The class satisfies the following concept requirements:
+-
+[BidirectionalIterator](http://en.cppreference.com/w/cpp/concept/BidirectionalIterator):
+  The iterator that can be moved can be moved in both directions (i.e.
+  incremented and decremented).
+
+@since version 1.0.0, simplified in version 2.0.9, change to bidirectional
+       iterators in version 3.0.0 (see https://github.com/nlohmann/json/issues/593)
+*/
+template<typename BasicJsonType>
+class iter_impl
+{
+    /// allow basic_json to access private members
+    friend iter_impl<typename std::conditional<std::is_const<BasicJsonType>::value, typename std::remove_const<BasicJsonType>::type, const BasicJsonType>::type>;
+    friend BasicJsonType;
+    friend iteration_proxy<iter_impl>;
+
+    using object_t = typename BasicJsonType::object_t;
+    using array_t = typename BasicJsonType::array_t;
+    // make sure BasicJsonType is basic_json or const basic_json
+    static_assert(is_basic_json<typename std::remove_const<BasicJsonType>::type>::value,
+                  "iter_impl only accepts (const) basic_json");
+
+  public:
+
+    /// The std::iterator class template (used as a base class to provide typedefs) is deprecated in C++17.
+    /// The C++ Standard has never required user-defined iterators to derive from std::iterator.
+    /// A user-defined iterator should provide publicly accessible typedefs named
+    /// iterator_category, value_type, difference_type, pointer, and reference.
+    /// Note that value_type is required to be non-const, even for constant iterators.
+    using iterator_category = std::bidirectional_iterator_tag;
+
+    /// the type of the values when the iterator is dereferenced
+    using value_type = typename BasicJsonType::value_type;
+    /// a type to represent differences between iterators
+    using difference_type = typename BasicJsonType::difference_type;
+    /// defines a pointer to the type iterated over (value_type)
+    using pointer = typename std::conditional<std::is_const<BasicJsonType>::value,
+          typename BasicJsonType::const_pointer,
+          typename BasicJsonType::pointer>::type;
+    /// defines a reference to the type iterated over (value_type)
+    using reference =
+        typename std::conditional<std::is_const<BasicJsonType>::value,
+        typename BasicJsonType::const_reference,
+        typename BasicJsonType::reference>::type;
+
+    /// default constructor
+    iter_impl() = default;
+
+    /*!
+    @brief constructor for a given JSON instance
+    @param[in] object  pointer to a JSON object for this iterator
+    @pre object != nullptr
+    @post The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    explicit iter_impl(pointer object) noexcept : m_object(object)
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                m_it.object_iterator = typename object_t::iterator();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_it.array_iterator = typename array_t::iterator();
+                break;
+            }
+
+            default:
+            {
+                m_it.primitive_iterator = primitive_iterator_t();
+                break;
+            }
+        }
+    }
+
+    /*!
+    @note The conventional copy constructor and copy assignment are implicitly
+          defined. Combined with the following converting constructor and
+          assignment, they support: (1) copy from iterator to iterator, (2)
+          copy from const iterator to const iterator, and (3) conversion from
+          iterator to const iterator. However conversion from const iterator
+          to iterator is not defined.
+    */
+
+    /*!
+    @brief converting constructor
+    @param[in] other  non-const iterator to copy from
+    @note It is not checked whether @a other is initialized.
+    */
+    iter_impl(const iter_impl<typename std::remove_const<BasicJsonType>::type>& other) noexcept
+        : m_object(other.m_object), m_it(other.m_it) {}
+
+    /*!
+    @brief converting assignment
+    @param[in,out] other  non-const iterator to copy from
+    @return const/non-const iterator
+    @note It is not checked whether @a other is initialized.
+    */
+    iter_impl& operator=(const iter_impl<typename std::remove_const<BasicJsonType>::type>& other) noexcept
+    {
+        m_object = other.m_object;
+        m_it = other.m_it;
+        return *this;
+    }
+
+  private:
+    /*!
+    @brief set the iterator to the first value
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    void set_begin() noexcept
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                m_it.object_iterator = m_object->m_value.object->begin();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_it.array_iterator = m_object->m_value.array->begin();
+                break;
+            }
+
+            case value_t::null:
+            {
+                // set to end so begin()==end() is true: null is empty
+                m_it.primitive_iterator.set_end();
+                break;
+            }
+
+            default:
+            {
+                m_it.primitive_iterator.set_begin();
+                break;
+            }
+        }
+    }
+
+    /*!
+    @brief set the iterator past the last value
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    void set_end() noexcept
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                m_it.object_iterator = m_object->m_value.object->end();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_it.array_iterator = m_object->m_value.array->end();
+                break;
+            }
+
+            default:
+            {
+                m_it.primitive_iterator.set_end();
+                break;
+            }
+        }
+    }
+
+  public:
+    /*!
+    @brief return a reference to the value pointed to by the iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    reference operator*() const
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                assert(m_it.object_iterator != m_object->m_value.object->end());
+                return m_it.object_iterator->second;
+            }
+
+            case value_t::array:
+            {
+                assert(m_it.array_iterator != m_object->m_value.array->end());
+                return *m_it.array_iterator;
+            }
+
+            case value_t::null:
+                JSON_THROW(invalid_iterator::create(214, "cannot get value"));
+
+            default:
+            {
+                if (JSON_LIKELY(m_it.primitive_iterator.is_begin()))
+                {
+                    return *m_object;
+                }
+
+                JSON_THROW(invalid_iterator::create(214, "cannot get value"));
+            }
+        }
+    }
+
+    /*!
+    @brief dereference the iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    pointer operator->() const
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                assert(m_it.object_iterator != m_object->m_value.object->end());
+                return &(m_it.object_iterator->second);
+            }
+
+            case value_t::array:
+            {
+                assert(m_it.array_iterator != m_object->m_value.array->end());
+                return &*m_it.array_iterator;
+            }
+
+            default:
+            {
+                if (JSON_LIKELY(m_it.primitive_iterator.is_begin()))
+                {
+                    return m_object;
+                }
+
+                JSON_THROW(invalid_iterator::create(214, "cannot get value"));
+            }
+        }
+    }
+
+    /*!
+    @brief post-increment (it++)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl const operator++(int)
+    {
+        auto result = *this;
+        ++(*this);
+        return result;
+    }
+
+    /*!
+    @brief pre-increment (++it)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator++()
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                std::advance(m_it.object_iterator, 1);
+                break;
+            }
+
+            case value_t::array:
+            {
+                std::advance(m_it.array_iterator, 1);
+                break;
+            }
+
+            default:
+            {
+                ++m_it.primitive_iterator;
+                break;
+            }
+        }
+
+        return *this;
+    }
+
+    /*!
+    @brief post-decrement (it--)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl const operator--(int)
+    {
+        auto result = *this;
+        --(*this);
+        return result;
+    }
+
+    /*!
+    @brief pre-decrement (--it)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator--()
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+            {
+                std::advance(m_it.object_iterator, -1);
+                break;
+            }
+
+            case value_t::array:
+            {
+                std::advance(m_it.array_iterator, -1);
+                break;
+            }
+
+            default:
+            {
+                --m_it.primitive_iterator;
+                break;
+            }
+        }
+
+        return *this;
+    }
+
+    /*!
+    @brief  comparison: equal
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    bool operator==(const iter_impl& other) const
+    {
+        // if objects are not the same, the comparison is undefined
+        if (JSON_UNLIKELY(m_object != other.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers"));
+        }
+
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+                return (m_it.object_iterator == other.m_it.object_iterator);
+
+            case value_t::array:
+                return (m_it.array_iterator == other.m_it.array_iterator);
+
+            default:
+                return (m_it.primitive_iterator == other.m_it.primitive_iterator);
+        }
+    }
+
+    /*!
+    @brief  comparison: not equal
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    bool operator!=(const iter_impl& other) const
+    {
+        return not operator==(other);
+    }
+
+    /*!
+    @brief  comparison: smaller
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    bool operator<(const iter_impl& other) const
+    {
+        // if objects are not the same, the comparison is undefined
+        if (JSON_UNLIKELY(m_object != other.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers"));
+        }
+
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(213, "cannot compare order of object iterators"));
+
+            case value_t::array:
+                return (m_it.array_iterator < other.m_it.array_iterator);
+
+            default:
+                return (m_it.primitive_iterator < other.m_it.primitive_iterator);
+        }
+    }
+
+    /*!
+    @brief  comparison: less than or equal
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    bool operator<=(const iter_impl& other) const
+    {
+        return not other.operator < (*this);
+    }
+
+    /*!
+    @brief  comparison: greater than
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    bool operator>(const iter_impl& other) const
+    {
+        return not operator<=(other);
+    }
+
+    /*!
+    @brief  comparison: greater than or equal
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    bool operator>=(const iter_impl& other) const
+    {
+        return not operator<(other);
+    }
+
+    /*!
+    @brief  add to iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator+=(difference_type i)
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators"));
+
+            case value_t::array:
+            {
+                std::advance(m_it.array_iterator, i);
+                break;
+            }
+
+            default:
+            {
+                m_it.primitive_iterator += i;
+                break;
+            }
+        }
+
+        return *this;
+    }
+
+    /*!
+    @brief  subtract from iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator-=(difference_type i)
+    {
+        return operator+=(-i);
+    }
+
+    /*!
+    @brief  add to iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl operator+(difference_type i) const
+    {
+        auto result = *this;
+        result += i;
+        return result;
+    }
+
+    /*!
+    @brief  addition of distance and iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    friend iter_impl operator+(difference_type i, const iter_impl& it)
+    {
+        auto result = it;
+        result += i;
+        return result;
+    }
+
+    /*!
+    @brief  subtract from iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl operator-(difference_type i) const
+    {
+        auto result = *this;
+        result -= i;
+        return result;
+    }
+
+    /*!
+    @brief  return difference
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    difference_type operator-(const iter_impl& other) const
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators"));
+
+            case value_t::array:
+                return m_it.array_iterator - other.m_it.array_iterator;
+
+            default:
+                return m_it.primitive_iterator - other.m_it.primitive_iterator;
+        }
+    }
+
+    /*!
+    @brief  access to successor
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    reference operator[](difference_type n) const
+    {
+        assert(m_object != nullptr);
+
+        switch (m_object->m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(208, "cannot use operator[] for object iterators"));
+
+            case value_t::array:
+                return *std::next(m_it.array_iterator, n);
+
+            case value_t::null:
+                JSON_THROW(invalid_iterator::create(214, "cannot get value"));
+
+            default:
+            {
+                if (JSON_LIKELY(m_it.primitive_iterator.get_value() == -n))
+                {
+                    return *m_object;
+                }
+
+                JSON_THROW(invalid_iterator::create(214, "cannot get value"));
+            }
+        }
+    }
+
+    /*!
+    @brief  return the key of an object iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    typename object_t::key_type key() const
+    {
+        assert(m_object != nullptr);
+
+        if (JSON_LIKELY(m_object->is_object()))
+        {
+            return m_it.object_iterator->first;
+        }
+
+        JSON_THROW(invalid_iterator::create(207, "cannot use key() for non-object iterators"));
+    }
+
+    /*!
+    @brief  return the value of an iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    reference value() const
+    {
+        return operator*();
+    }
+
+  private:
+    /// associated JSON instance
+    pointer m_object = nullptr;
+    /// the actual iterator of the associated instance
+    internal_iterator<typename std::remove_const<BasicJsonType>::type> m_it;
+};
+}
+}

src/lib/json/detail/iterators/iteration_proxy.hpp

@@ -0,0 +1,98 @@
+#pragma once
+
+#include <cstddef> // size_t
+#include <string> // string, to_string
+
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+/// proxy class for the items() function
+template<typename IteratorType> class iteration_proxy
+{
+  private:
+    /// helper class for iteration
+    class iteration_proxy_internal
+    {
+      private:
+        /// the iterator
+        IteratorType anchor;
+        /// an index for arrays (used to create key names)
+        std::size_t array_index = 0;
+
+      public:
+        explicit iteration_proxy_internal(IteratorType it) noexcept : anchor(it) {}
+
+        /// dereference operator (needed for range-based for)
+        iteration_proxy_internal& operator*()
+        {
+            return *this;
+        }
+
+        /// increment operator (needed for range-based for)
+        iteration_proxy_internal& operator++()
+        {
+            ++anchor;
+            ++array_index;
+
+            return *this;
+        }
+
+        /// inequality operator (needed for range-based for)
+        bool operator!=(const iteration_proxy_internal& o) const noexcept
+        {
+            return anchor != o.anchor;
+        }
+
+        /// return key of the iterator
+        std::string key() const
+        {
+            assert(anchor.m_object != nullptr);
+
+            switch (anchor.m_object->type())
+            {
+                // use integer array index as key
+                case value_t::array:
+                    return std::to_string(array_index);
+
+                // use key from the object
+                case value_t::object:
+                    return anchor.key();
+
+                // use an empty key for all primitive types
+                default:
+                    return "";
+            }
+        }
+
+        /// return value of the iterator
+        typename IteratorType::reference value() const
+        {
+            return anchor.value();
+        }
+    };
+
+    /// the container to iterate
+    typename IteratorType::reference container;
+
+  public:
+    /// construct iteration proxy from a container
+    explicit iteration_proxy(typename IteratorType::reference cont) noexcept
+        : container(cont) {}
+
+    /// return iterator begin (needed for range-based for)
+    iteration_proxy_internal begin() noexcept
+    {
+        return iteration_proxy_internal(container.begin());
+    }
+
+    /// return iterator end (needed for range-based for)
+    iteration_proxy_internal end() noexcept
+    {
+        return iteration_proxy_internal(container.end());
+    }
+};
+}
+}

src/lib/json/detail/iterators/json_reverse_iterator.hpp

@@ -0,0 +1,119 @@
+#pragma once
+
+#include <cstddef> // ptrdiff_t
+#include <iterator> // reverse_iterator
+#include <utility> // declval
+
+namespace nlohmann
+{
+namespace detail
+{
+//////////////////////
+// reverse_iterator //
+//////////////////////
+
+/*!
+@brief a template for a reverse iterator class
+
+@tparam Base the base iterator type to reverse. Valid types are @ref
+iterator (to create @ref reverse_iterator) and @ref const_iterator (to
+create @ref const_reverse_iterator).
+
+@requirement The class satisfies the following concept requirements:
+-
+[BidirectionalIterator](http://en.cppreference.com/w/cpp/concept/BidirectionalIterator):
+  The iterator that can be moved can be moved in both directions (i.e.
+  incremented and decremented).
+- [OutputIterator](http://en.cppreference.com/w/cpp/concept/OutputIterator):
+  It is possible to write to the pointed-to element (only if @a Base is
+  @ref iterator).
+
+@since version 1.0.0
+*/
+template<typename Base>
+class json_reverse_iterator : public std::reverse_iterator<Base>
+{
+  public:
+    using difference_type = std::ptrdiff_t;
+    /// shortcut to the reverse iterator adapter
+    using base_iterator = std::reverse_iterator<Base>;
+    /// the reference type for the pointed-to element
+    using reference = typename Base::reference;
+
+    /// create reverse iterator from iterator
+    json_reverse_iterator(const typename base_iterator::iterator_type& it) noexcept
+        : base_iterator(it) {}
+
+    /// create reverse iterator from base class
+    json_reverse_iterator(const base_iterator& it) noexcept : base_iterator(it) {}
+
+    /// post-increment (it++)
+    json_reverse_iterator const operator++(int)
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator++(1));
+    }
+
+    /// pre-increment (++it)
+    json_reverse_iterator& operator++()
+    {
+        return static_cast<json_reverse_iterator&>(base_iterator::operator++());
+    }
+
+    /// post-decrement (it--)
+    json_reverse_iterator const operator--(int)
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator--(1));
+    }
+
+    /// pre-decrement (--it)
+    json_reverse_iterator& operator--()
+    {
+        return static_cast<json_reverse_iterator&>(base_iterator::operator--());
+    }
+
+    /// add to iterator
+    json_reverse_iterator& operator+=(difference_type i)
+    {
+        return static_cast<json_reverse_iterator&>(base_iterator::operator+=(i));
+    }
+
+    /// add to iterator
+    json_reverse_iterator operator+(difference_type i) const
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator+(i));
+    }
+
+    /// subtract from iterator
+    json_reverse_iterator operator-(difference_type i) const
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator-(i));
+    }
+
+    /// return difference
+    difference_type operator-(const json_reverse_iterator& other) const
+    {
+        return base_iterator(*this) - base_iterator(other);
+    }
+
+    /// access to successor
+    reference operator[](difference_type n) const
+    {
+        return *(this->operator+(n));
+    }
+
+    /// return the key of an object iterator
+    auto key() const -> decltype(std::declval<Base>().key())
+    {
+        auto it = --this->base();
+        return it.key();
+    }
+
+    /// return the value of an iterator
+    reference value() const
+    {
+        auto it = --this->base();
+        return it.operator * ();
+    }
+};
+}
+}

src/lib/json/detail/iterators/primitive_iterator.hpp

@@ -0,0 +1,120 @@
+#pragma once
+
+#include <cstddef> // ptrdiff_t
+#include <limits>  // numeric_limits
+
+namespace nlohmann
+{
+namespace detail
+{
+/*
+@brief an iterator for primitive JSON types
+
+This class models an iterator for primitive JSON types (boolean, number,
+string). It's only purpose is to allow the iterator/const_iterator classes
+to "iterate" over primitive values. Internally, the iterator is modeled by
+a `difference_type` variable. Value begin_value (`0`) models the begin,
+end_value (`1`) models past the end.
+*/
+class primitive_iterator_t
+{
+  private:
+    using difference_type = std::ptrdiff_t;
+    static constexpr difference_type begin_value = 0;
+    static constexpr difference_type end_value = begin_value + 1;
+
+    /// iterator as signed integer type
+    difference_type m_it = (std::numeric_limits<std::ptrdiff_t>::min)();
+
+  public:
+    constexpr difference_type get_value() const noexcept
+    {
+        return m_it;
+    }
+
+    /// set iterator to a defined beginning
+    void set_begin() noexcept
+    {
+        m_it = begin_value;
+    }
+
+    /// set iterator to a defined past the end
+    void set_end() noexcept
+    {
+        m_it = end_value;
+    }
+
+    /// return whether the iterator can be dereferenced
+    constexpr bool is_begin() const noexcept
+    {
+        return m_it == begin_value;
+    }
+
+    /// return whether the iterator is at end
+    constexpr bool is_end() const noexcept
+    {
+        return m_it == end_value;
+    }
+
+    friend constexpr bool operator==(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept
+    {
+        return lhs.m_it == rhs.m_it;
+    }
+
+    friend constexpr bool operator<(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept
+    {
+        return lhs.m_it < rhs.m_it;
+    }
+
+    primitive_iterator_t operator+(difference_type n) noexcept
+    {
+        auto result = *this;
+        result += n;
+        return result;
+    }
+
+    friend constexpr difference_type operator-(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept
+    {
+        return lhs.m_it - rhs.m_it;
+    }
+
+    primitive_iterator_t& operator++() noexcept
+    {
+        ++m_it;
+        return *this;
+    }
+
+    primitive_iterator_t const operator++(int) noexcept
+    {
+        auto result = *this;
+        m_it++;
+        return result;
+    }
+
+    primitive_iterator_t& operator--() noexcept
+    {
+        --m_it;
+        return *this;
+    }
+
+    primitive_iterator_t const operator--(int) noexcept
+    {
+        auto result = *this;
+        m_it--;
+        return result;
+    }
+
+    primitive_iterator_t& operator+=(difference_type n) noexcept
+    {
+        m_it += n;
+        return *this;
+    }
+
+    primitive_iterator_t& operator-=(difference_type n) noexcept
+    {
+        m_it -= n;
+        return *this;
+    }
+};
+}
+}

src/lib/json/detail/json_pointer.hpp

@@ -0,0 +1,696 @@
+#pragma once
+
+#include <cassert> // assert
+#include <numeric> // accumulate
+#include <string> // string
+#include <vector> // vector
+
+#include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/exceptions.hpp>
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+template<typename BasicJsonType>
+class json_pointer
+{
+    // allow basic_json to access private members
+    NLOHMANN_BASIC_JSON_TPL_DECLARATION
+    friend class basic_json;
+
+  public:
+    /*!
+    @brief create JSON pointer
+
+    Create a JSON pointer according to the syntax described in
+    [Section 3 of RFC6901](https://tools.ietf.org/html/rfc6901#section-3).
+
+    @param[in] s  string representing the JSON pointer; if omitted, the empty
+                  string is assumed which references the whole JSON value
+
+    @throw parse_error.107 if the given JSON pointer @a s is nonempty and does
+                           not begin with a slash (`/`); see example below
+
+    @throw parse_error.108 if a tilde (`~`) in the given JSON pointer @a s is
+    not followed by `0` (representing `~`) or `1` (representing `/`); see
+    example below
+
+    @liveexample{The example shows the construction several valid JSON pointers
+    as well as the exceptional behavior.,json_pointer}
+
+    @since version 2.0.0
+    */
+    explicit json_pointer(const std::string& s = "")
+        : reference_tokens(split(s))
+    {}
+
+    /*!
+    @brief return a string representation of the JSON pointer
+
+    @invariant For each JSON pointer `ptr`, it holds:
+    @code {.cpp}
+    ptr == json_pointer(ptr.to_string());
+    @endcode
+
+    @return a string representation of the JSON pointer
+
+    @liveexample{The example shows the result of `to_string`.,
+    json_pointer__to_string}
+
+    @since version 2.0.0
+    */
+    std::string to_string() const noexcept
+    {
+        return std::accumulate(reference_tokens.begin(), reference_tokens.end(),
+                               std::string{},
+                               [](const std::string & a, const std::string & b)
+        {
+            return a + "/" + escape(b);
+        });
+    }
+
+    /// @copydoc to_string()
+    operator std::string() const
+    {
+        return to_string();
+    }
+
+    /*!
+    @param[in] s  reference token to be converted into an array index
+
+    @return integer representation of @a s
+
+    @throw out_of_range.404 if string @a s could not be converted to an integer
+    */
+    static int array_index(const std::string& s)
+    {
+        std::size_t processed_chars = 0;
+        const int res = std::stoi(s, &processed_chars);
+
+        // check if the string was completely read
+        if (JSON_UNLIKELY(processed_chars != s.size()))
+        {
+            JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + s + "'"));
+        }
+
+        return res;
+    }
+
+  private:
+    /*!
+    @brief remove and return last reference pointer
+    @throw out_of_range.405 if JSON pointer has no parent
+    */
+    std::string pop_back()
+    {
+        if (JSON_UNLIKELY(is_root()))
+        {
+            JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent"));
+        }
+
+        auto last = reference_tokens.back();
+        reference_tokens.pop_back();
+        return last;
+    }
+
+    /// return whether pointer points to the root document
+    bool is_root() const
+    {
+        return reference_tokens.empty();
+    }
+
+    json_pointer top() const
+    {
+        if (JSON_UNLIKELY(is_root()))
+        {
+            JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent"));
+        }
+
+        json_pointer result = *this;
+        result.reference_tokens = {reference_tokens[0]};
+        return result;
+    }
+
+    /*!
+    @brief create and return a reference to the pointed to value
+
+    @complexity Linear in the number of reference tokens.
+
+    @throw parse_error.109 if array index is not a number
+    @throw type_error.313 if value cannot be unflattened
+    */
+    BasicJsonType& get_and_create(BasicJsonType& j) const
+    {
+        using size_type = typename BasicJsonType::size_type;
+        auto result = &j;
+
+        // in case no reference tokens exist, return a reference to the JSON value
+        // j which will be overwritten by a primitive value
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (result->m_type)
+            {
+                case detail::value_t::null:
+                {
+                    if (reference_token == "0")
+                    {
+                        // start a new array if reference token is 0
+                        result = &result->operator[](0);
+                    }
+                    else
+                    {
+                        // start a new object otherwise
+                        result = &result->operator[](reference_token);
+                    }
+                    break;
+                }
+
+                case detail::value_t::object:
+                {
+                    // create an entry in the object
+                    result = &result->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    // create an entry in the array
+                    JSON_TRY
+                    {
+                        result = &result->operator[](static_cast<size_type>(array_index(reference_token)));
+                    }
+                    JSON_CATCH(std::invalid_argument&)
+                    {
+                        JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number"));
+                    }
+                    break;
+                }
+
+                /*
+                The following code is only reached if there exists a reference
+                token _and_ the current value is primitive. In this case, we have
+                an error situation, because primitive values may only occur as
+                single value; that is, with an empty list of reference tokens.
+                */
+                default:
+                    JSON_THROW(detail::type_error::create(313, "invalid value to unflatten"));
+            }
+        }
+
+        return *result;
+    }
+
+    /*!
+    @brief return a reference to the pointed to value
+
+    @note This version does not throw if a value is not present, but tries to
+          create nested values instead. For instance, calling this function
+          with pointer `"/this/that"` on a null value is equivalent to calling
+          `operator[]("this").operator[]("that")` on that value, effectively
+          changing the null value to an object.
+
+    @param[in] ptr  a JSON value
+
+    @return reference to the JSON value pointed to by the JSON pointer
+
+    @complexity Linear in the length of the JSON pointer.
+
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    BasicJsonType& get_unchecked(BasicJsonType* ptr) const
+    {
+        using size_type = typename BasicJsonType::size_type;
+        for (const auto& reference_token : reference_tokens)
+        {
+            // convert null values to arrays or objects before continuing
+            if (ptr->m_type == detail::value_t::null)
+            {
+                // check if reference token is a number
+                const bool nums =
+                    std::all_of(reference_token.begin(), reference_token.end(),
+                                [](const char x)
+                {
+                    return (x >= '0' and x <= '9');
+                });
+
+                // change value to array for numbers or "-" or to object otherwise
+                *ptr = (nums or reference_token == "-")
+                       ? detail::value_t::array
+                       : detail::value_t::object;
+            }
+
+            switch (ptr->m_type)
+            {
+                case detail::value_t::object:
+                {
+                    // use unchecked object access
+                    ptr = &ptr->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    // error condition (cf. RFC 6901, Sect. 4)
+                    if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0'))
+                    {
+                        JSON_THROW(detail::parse_error::create(106, 0,
+                                                               "array index '" + reference_token +
+                                                               "' must not begin with '0'"));
+                    }
+
+                    if (reference_token == "-")
+                    {
+                        // explicitly treat "-" as index beyond the end
+                        ptr = &ptr->operator[](ptr->m_value.array->size());
+                    }
+                    else
+                    {
+                        // convert array index to number; unchecked access
+                        JSON_TRY
+                        {
+                            ptr = &ptr->operator[](
+                                static_cast<size_type>(array_index(reference_token)));
+                        }
+                        JSON_CATCH(std::invalid_argument&)
+                        {
+                            JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number"));
+                        }
+                    }
+                    break;
+                }
+
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'"));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.402  if the array index '-' is used
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    BasicJsonType& get_checked(BasicJsonType* ptr) const
+    {
+        using size_type = typename BasicJsonType::size_type;
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->m_type)
+            {
+                case detail::value_t::object:
+                {
+                    // note: at performs range check
+                    ptr = &ptr->at(reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" always fails the range check
+                        JSON_THROW(detail::out_of_range::create(402,
+                                                                "array index '-' (" + std::to_string(ptr->m_value.array->size()) +
+                                                                ") is out of range"));
+                    }
+
+                    // error condition (cf. RFC 6901, Sect. 4)
+                    if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0'))
+                    {
+                        JSON_THROW(detail::parse_error::create(106, 0,
+                                                               "array index '" + reference_token +
+                                                               "' must not begin with '0'"));
+                    }
+
+                    // note: at performs range check
+                    JSON_TRY
+                    {
+                        ptr = &ptr->at(static_cast<size_type>(array_index(reference_token)));
+                    }
+                    JSON_CATCH(std::invalid_argument&)
+                    {
+                        JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number"));
+                    }
+                    break;
+                }
+
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'"));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @brief return a const reference to the pointed to value
+
+    @param[in] ptr  a JSON value
+
+    @return const reference to the JSON value pointed to by the JSON
+    pointer
+
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.402  if the array index '-' is used
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    const BasicJsonType& get_unchecked(const BasicJsonType* ptr) const
+    {
+        using size_type = typename BasicJsonType::size_type;
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->m_type)
+            {
+                case detail::value_t::object:
+                {
+                    // use unchecked object access
+                    ptr = &ptr->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" cannot be used for const access
+                        JSON_THROW(detail::out_of_range::create(402,
+                                                                "array index '-' (" + std::to_string(ptr->m_value.array->size()) +
+                                                                ") is out of range"));
+                    }
+
+                    // error condition (cf. RFC 6901, Sect. 4)
+                    if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0'))
+                    {
+                        JSON_THROW(detail::parse_error::create(106, 0,
+                                                               "array index '" + reference_token +
+                                                               "' must not begin with '0'"));
+                    }
+
+                    // use unchecked array access
+                    JSON_TRY
+                    {
+                        ptr = &ptr->operator[](
+                            static_cast<size_type>(array_index(reference_token)));
+                    }
+                    JSON_CATCH(std::invalid_argument&)
+                    {
+                        JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number"));
+                    }
+                    break;
+                }
+
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'"));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.402  if the array index '-' is used
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    const BasicJsonType& get_checked(const BasicJsonType* ptr) const
+    {
+        using size_type = typename BasicJsonType::size_type;
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->m_type)
+            {
+                case detail::value_t::object:
+                {
+                    // note: at performs range check
+                    ptr = &ptr->at(reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" always fails the range check
+                        JSON_THROW(detail::out_of_range::create(402,
+                                                                "array index '-' (" + std::to_string(ptr->m_value.array->size()) +
+                                                                ") is out of range"));
+                    }
+
+                    // error condition (cf. RFC 6901, Sect. 4)
+                    if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0'))
+                    {
+                        JSON_THROW(detail::parse_error::create(106, 0,
+                                                               "array index '" + reference_token +
+                                                               "' must not begin with '0'"));
+                    }
+
+                    // note: at performs range check
+                    JSON_TRY
+                    {
+                        ptr = &ptr->at(static_cast<size_type>(array_index(reference_token)));
+                    }
+                    JSON_CATCH(std::invalid_argument&)
+                    {
+                        JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number"));
+                    }
+                    break;
+                }
+
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'"));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @brief split the string input to reference tokens
+
+    @note This function is only called by the json_pointer constructor.
+          All exceptions below are documented there.
+
+    @throw parse_error.107  if the pointer is not empty or begins with '/'
+    @throw parse_error.108  if character '~' is not followed by '0' or '1'
+    */
+    static std::vector<std::string> split(const std::string& reference_string)
+    {
+        std::vector<std::string> result;
+
+        // special case: empty reference string -> no reference tokens
+        if (reference_string.empty())
+        {
+            return result;
+        }
+
+        // check if nonempty reference string begins with slash
+        if (JSON_UNLIKELY(reference_string[0] != '/'))
+        {
+            JSON_THROW(detail::parse_error::create(107, 1,
+                                                   "JSON pointer must be empty or begin with '/' - was: '" +
+                                                   reference_string + "'"));
+        }
+
+        // extract the reference tokens:
+        // - slash: position of the last read slash (or end of string)
+        // - start: position after the previous slash
+        for (
+            // search for the first slash after the first character
+            std::size_t slash = reference_string.find_first_of('/', 1),
+            // set the beginning of the first reference token
+            start = 1;
+            // we can stop if start == string::npos+1 = 0
+            start != 0;
+            // set the beginning of the next reference token
+            // (will eventually be 0 if slash == std::string::npos)
+            start = slash + 1,
+            // find next slash
+            slash = reference_string.find_first_of('/', start))
+        {
+            // use the text between the beginning of the reference token
+            // (start) and the last slash (slash).
+            auto reference_token = reference_string.substr(start, slash - start);
+
+            // check reference tokens are properly escaped
+            for (std::size_t pos = reference_token.find_first_of('~');
+                    pos != std::string::npos;
+                    pos = reference_token.find_first_of('~', pos + 1))
+            {
+                assert(reference_token[pos] == '~');
+
+                // ~ must be followed by 0 or 1
+                if (JSON_UNLIKELY(pos == reference_token.size() - 1 or
+                                  (reference_token[pos + 1] != '0' and
+                                   reference_token[pos + 1] != '1')))
+                {
+                    JSON_THROW(detail::parse_error::create(108, 0, "escape character '~' must be followed with '0' or '1'"));
+                }
+            }
+
+            // finally, store the reference token
+            unescape(reference_token);
+            result.push_back(reference_token);
+        }
+
+        return result;
+    }
+
+    /*!
+    @brief replace all occurrences of a substring by another string
+
+    @param[in,out] s  the string to manipulate; changed so that all
+                   occurrences of @a f are replaced with @a t
+    @param[in]     f  the substring to replace with @a t
+    @param[in]     t  the string to replace @a f
+
+    @pre The search string @a f must not be empty. **This precondition is
+    enforced with an assertion.**
+
+    @since version 2.0.0
+    */
+    static void replace_substring(std::string& s, const std::string& f,
+                                  const std::string& t)
+    {
+        assert(not f.empty());
+        for (auto pos = s.find(f);                // find first occurrence of f
+                pos != std::string::npos;         // make sure f was found
+                s.replace(pos, f.size(), t),      // replace with t, and
+                pos = s.find(f, pos + t.size()))  // find next occurrence of f
+        {}
+    }
+
+    /// escape "~"" to "~0" and "/" to "~1"
+    static std::string escape(std::string s)
+    {
+        replace_substring(s, "~", "~0");
+        replace_substring(s, "/", "~1");
+        return s;
+    }
+
+    /// unescape "~1" to tilde and "~0" to slash (order is important!)
+    static void unescape(std::string& s)
+    {
+        replace_substring(s, "~1", "/");
+        replace_substring(s, "~0", "~");
+    }
+
+    /*!
+    @param[in] reference_string  the reference string to the current value
+    @param[in] value             the value to consider
+    @param[in,out] result        the result object to insert values to
+
+    @note Empty objects or arrays are flattened to `null`.
+    */
+    static void flatten(const std::string& reference_string,
+                        const BasicJsonType& value,
+                        BasicJsonType& result)
+    {
+        switch (value.m_type)
+        {
+            case detail::value_t::array:
+            {
+                if (value.m_value.array->empty())
+                {
+                    // flatten empty array as null
+                    result[reference_string] = nullptr;
+                }
+                else
+                {
+                    // iterate array and use index as reference string
+                    for (std::size_t i = 0; i < value.m_value.array->size(); ++i)
+                    {
+                        flatten(reference_string + "/" + std::to_string(i),
+                                value.m_value.array->operator[](i), result);
+                    }
+                }
+                break;
+            }
+
+            case detail::value_t::object:
+            {
+                if (value.m_value.object->empty())
+                {
+                    // flatten empty object as null
+                    result[reference_string] = nullptr;
+                }
+                else
+                {
+                    // iterate object and use keys as reference string
+                    for (const auto& element : *value.m_value.object)
+                    {
+                        flatten(reference_string + "/" + escape(element.first), element.second, result);
+                    }
+                }
+                break;
+            }
+
+            default:
+            {
+                // add primitive value with its reference string
+                result[reference_string] = value;
+                break;
+            }
+        }
+    }
+
+    /*!
+    @param[in] value  flattened JSON
+
+    @return unflattened JSON
+
+    @throw parse_error.109 if array index is not a number
+    @throw type_error.314  if value is not an object
+    @throw type_error.315  if object values are not primitive
+    @throw type_error.313  if value cannot be unflattened
+    */
+    static BasicJsonType
+    unflatten(const BasicJsonType& value)
+    {
+        if (JSON_UNLIKELY(not value.is_object()))
+        {
+            JSON_THROW(detail::type_error::create(314, "only objects can be unflattened"));
+        }
+
+        BasicJsonType result;
+
+        // iterate the JSON object values
+        for (const auto& element : *value.m_value.object)
+        {
+            if (JSON_UNLIKELY(not element.second.is_primitive()))
+            {
+                JSON_THROW(detail::type_error::create(315, "values in object must be primitive"));
+            }
+
+            // assign value to reference pointed to by JSON pointer; Note that if
+            // the JSON pointer is "" (i.e., points to the whole value), function
+            // get_and_create returns a reference to result itself. An assignment
+            // will then create a primitive value.
+            json_pointer(element.first).get_and_create(result) = element.second;
+        }
+
+        return result;
+    }
+
+    friend bool operator==(json_pointer const& lhs,
+                           json_pointer const& rhs) noexcept
+    {
+        return (lhs.reference_tokens == rhs.reference_tokens);
+    }
+
+    friend bool operator!=(json_pointer const& lhs,
+                           json_pointer const& rhs) noexcept
+    {
+        return not (lhs == rhs);
+    }
+
+    /// the reference tokens
+    std::vector<std::string> reference_tokens;
+};
+}

src/lib/json/detail/json_ref.hpp

@@ -0,0 +1,63 @@
+#pragma once
+
+#include <initializer_list>
+#include <utility>
+
+namespace nlohmann
+{
+namespace detail
+{
+template<typename BasicJsonType>
+class json_ref
+{
+  public:
+    using value_type = BasicJsonType;
+
+    json_ref(value_type&& value)
+        : owned_value(std::move(value)), value_ref(&owned_value), is_rvalue(true)
+    {}
+
+    json_ref(const value_type& value)
+        : value_ref(const_cast<value_type*>(&value)), is_rvalue(false)
+    {}
+
+    json_ref(std::initializer_list<json_ref> init)
+        : owned_value(init), value_ref(&owned_value), is_rvalue(true)
+    {}
+
+    template<class... Args>
+    json_ref(Args&& ... args)
+        : owned_value(std::forward<Args>(args)...), value_ref(&owned_value), is_rvalue(true)
+    {}
+
+    // class should be movable only
+    json_ref(json_ref&&) = default;
+    json_ref(const json_ref&) = delete;
+    json_ref& operator=(const json_ref&) = delete;
+
+    value_type moved_or_copied() const
+    {
+        if (is_rvalue)
+        {
+            return std::move(*value_ref);
+        }
+        return *value_ref;
+    }
+
+    value_type const& operator*() const
+    {
+        return *static_cast<value_type const*>(value_ref);
+    }
+
+    value_type const* operator->() const
+    {
+        return static_cast<value_type const*>(value_ref);
+    }
+
+  private:
+    mutable value_type owned_value = nullptr;
+    value_type* value_ref = nullptr;
+    const bool is_rvalue;
+};
+}
+}

src/lib/json/detail/macro_scope.hpp

@@ -0,0 +1,115 @@
+#pragma once
+
+// This file contains all internal macro definitions
+// You MUST include macro_unscope.hpp at the end of json.hpp to undef all of them
+
+// exclude unsupported compilers
+#if defined(__clang__)
+    #if (__clang_major__ * 10000 + __clang_minor__ * 100 + __clang_patchlevel__) < 30400
+        #error "unsupported Clang version - see https://github.com/nlohmann/json#supported-compilers"
+    #endif
+#elif defined(__GNUC__) && !(defined(__ICC) || defined(__INTEL_COMPILER))
+    #if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) < 40900
+        #error "unsupported GCC version - see https://github.com/nlohmann/json#supported-compilers"
+    #endif
+#endif
+
+// disable float-equal warnings on GCC/clang
+#if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__)
+    #pragma GCC diagnostic push
+    #pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+
+// disable documentation warnings on clang
+#if defined(__clang__)
+    #pragma GCC diagnostic push
+    #pragma GCC diagnostic ignored "-Wdocumentation"
+#endif
+
+// allow for portable deprecation warnings
+#if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__)
+    #define JSON_DEPRECATED __attribute__((deprecated))
+#elif defined(_MSC_VER)
+    #define JSON_DEPRECATED __declspec(deprecated)
+#else
+    #define JSON_DEPRECATED
+#endif
+
+// allow to disable exceptions
+#if (defined(__cpp_exceptions) || defined(__EXCEPTIONS) || defined(_CPPUNWIND)) && !defined(JSON_NOEXCEPTION)
+    #define JSON_THROW(exception) throw exception
+    #define JSON_TRY try
+    #define JSON_CATCH(exception) catch(exception)
+#else
+    #define JSON_THROW(exception) std::abort()
+    #define JSON_TRY if(true)
+    #define JSON_CATCH(exception) if(false)
+#endif
+
+// override exception macros
+#if defined(JSON_THROW_USER)
+    #undef JSON_THROW
+    #define JSON_THROW JSON_THROW_USER
+#endif
+#if defined(JSON_TRY_USER)
+    #undef JSON_TRY
+    #define JSON_TRY JSON_TRY_USER
+#endif
+#if defined(JSON_CATCH_USER)
+    #undef JSON_CATCH
+    #define JSON_CATCH JSON_CATCH_USER
+#endif
+
+// manual branch prediction
+#if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__)
+    #define JSON_LIKELY(x)      __builtin_expect(!!(x), 1)
+    #define JSON_UNLIKELY(x)    __builtin_expect(!!(x), 0)
+#else
+    #define JSON_LIKELY(x)      x
+    #define JSON_UNLIKELY(x)    x
+#endif
+
+// C++ language standard detection
+#if (defined(__cplusplus) && __cplusplus >= 201703L) || (defined(_HAS_CXX17) && _HAS_CXX17 == 1) // fix for issue #464
+    #define JSON_HAS_CPP_17
+    #define JSON_HAS_CPP_14
+#elif (defined(__cplusplus) && __cplusplus >= 201402L) || (defined(_HAS_CXX14) && _HAS_CXX14 == 1)
+    #define JSON_HAS_CPP_14
+#endif
+
+// Ugly macros to avoid uglier copy-paste when specializing basic_json. They
+// may be removed in the future once the class is split.
+
+#define NLOHMANN_BASIC_JSON_TPL_DECLARATION                                \
+    template<template<typename, typename, typename...> class ObjectType,   \
+             template<typename, typename...> class ArrayType,              \
+             class StringType, class BooleanType, class NumberIntegerType, \
+             class NumberUnsignedType, class NumberFloatType,              \
+             template<typename> class AllocatorType,                       \
+             template<typename, typename = void> class JSONSerializer>
+
+#define NLOHMANN_BASIC_JSON_TPL                                            \
+    basic_json<ObjectType, ArrayType, StringType, BooleanType,             \
+    NumberIntegerType, NumberUnsignedType, NumberFloatType,                \
+    AllocatorType, JSONSerializer>
+
+/*!
+@brief Helper to determine whether there's a key_type for T.
+
+This helper is used to tell associative containers apart from other containers
+such as sequence containers. For instance, `std::map` passes the test as it
+contains a `mapped_type`, whereas `std::vector` fails the test.
+
+@sa http://stackoverflow.com/a/7728728/266378
+@since version 1.0.0, overworked in version 2.0.6
+*/
+#define NLOHMANN_JSON_HAS_HELPER(type)                                        \
+    template<typename T> struct has_##type {                                  \
+    private:                                                                  \
+        template<typename U, typename = typename U::type>                     \
+        static int detect(U &&);                                              \
+        static void detect(...);                                              \
+    public:                                                                   \
+        static constexpr bool value =                                         \
+                std::is_integral<decltype(detect(std::declval<T>()))>::value; \
+    }

src/lib/json/detail/macro_unscope.hpp

@@ -0,0 +1,22 @@
+#pragma once
+
+// restore GCC/clang diagnostic settings
+#if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__)
+    #pragma GCC diagnostic pop
+#endif
+#if defined(__clang__)
+    #pragma GCC diagnostic pop
+#endif
+
+// clean up
+#undef JSON_CATCH
+#undef JSON_THROW
+#undef JSON_TRY
+#undef JSON_LIKELY
+#undef JSON_UNLIKELY
+#undef JSON_DEPRECATED
+#undef JSON_HAS_CPP_14
+#undef JSON_HAS_CPP_17
+#undef NLOHMANN_BASIC_JSON_TPL_DECLARATION
+#undef NLOHMANN_BASIC_JSON_TPL
+#undef NLOHMANN_JSON_HAS_HELPER

src/lib/json/detail/meta.hpp

@@ -0,0 +1,258 @@
+#pragma once
+
+#include <ciso646> // not
+#include <cstddef> // size_t
+#include <limits> // numeric_limits
+#include <type_traits> // conditional, enable_if, false_type, integral_constant, is_constructible, is_integral, is_same, remove_cv, remove_reference, true_type
+#include <utility> // declval
+
+#include <nlohmann/json_fwd.hpp>
+#include <nlohmann/detail/macro_scope.hpp>
+
+namespace nlohmann
+{
+/*!
+@brief detail namespace with internal helper functions
+
+This namespace collects functions that should not be exposed,
+implementations of some @ref basic_json methods, and meta-programming helpers.
+
+@since version 2.1.0
+*/
+namespace detail
+{
+/////////////
+// helpers //
+/////////////
+
+template<typename> struct is_basic_json : std::false_type {};
+
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+struct is_basic_json<NLOHMANN_BASIC_JSON_TPL> : std::true_type {};
+
+// alias templates to reduce boilerplate
+template<bool B, typename T = void>
+using enable_if_t = typename std::enable_if<B, T>::type;
+
+template<typename T>
+using uncvref_t = typename std::remove_cv<typename std::remove_reference<T>::type>::type;
+
+// implementation of C++14 index_sequence and affiliates
+// source: https://stackoverflow.com/a/32223343
+template<std::size_t... Ints>
+struct index_sequence
+{
+    using type = index_sequence;
+    using value_type = std::size_t;
+    static constexpr std::size_t size() noexcept
+    {
+        return sizeof...(Ints);
+    }
+};
+
+template<class Sequence1, class Sequence2>
+struct merge_and_renumber;
+
+template<std::size_t... I1, std::size_t... I2>
+struct merge_and_renumber<index_sequence<I1...>, index_sequence<I2...>>
+        : index_sequence < I1..., (sizeof...(I1) + I2)... > {};
+
+template<std::size_t N>
+struct make_index_sequence
+    : merge_and_renumber < typename make_index_sequence < N / 2 >::type,
+      typename make_index_sequence < N - N / 2 >::type > {};
+
+template<> struct make_index_sequence<0> : index_sequence<> {};
+template<> struct make_index_sequence<1> : index_sequence<0> {};
+
+template<typename... Ts>
+using index_sequence_for = make_index_sequence<sizeof...(Ts)>;
+
+/*
+Implementation of two C++17 constructs: conjunction, negation. This is needed
+to avoid evaluating all the traits in a condition
+
+For example: not std::is_same<void, T>::value and has_value_type<T>::value
+will not compile when T = void (on MSVC at least). Whereas
+conjunction<negation<std::is_same<void, T>>, has_value_type<T>>::value will
+stop evaluating if negation<...>::value == false
+
+Please note that those constructs must be used with caution, since symbols can
+become very long quickly (which can slow down compilation and cause MSVC
+internal compiler errors). Only use it when you have to (see example ahead).
+*/
+template<class...> struct conjunction : std::true_type {};
+template<class B1> struct conjunction<B1> : B1 {};
+template<class B1, class... Bn>
+struct conjunction<B1, Bn...> : std::conditional<bool(B1::value), conjunction<Bn...>, B1>::type {};
+
+template<class B> struct negation : std::integral_constant<bool, not B::value> {};
+
+// dispatch utility (taken from ranges-v3)
+template<unsigned N> struct priority_tag : priority_tag < N - 1 > {};
+template<> struct priority_tag<0> {};
+
+////////////////////////
+// has_/is_ functions //
+////////////////////////
+
+// source: https://stackoverflow.com/a/37193089/4116453
+
+template <typename T, typename = void>
+struct is_complete_type : std::false_type {};
+
+template <typename T>
+struct is_complete_type<T, decltype(void(sizeof(T)))> : std::true_type {};
+
+NLOHMANN_JSON_HAS_HELPER(mapped_type);
+NLOHMANN_JSON_HAS_HELPER(key_type);
+NLOHMANN_JSON_HAS_HELPER(value_type);
+NLOHMANN_JSON_HAS_HELPER(iterator);
+
+template<bool B, class RealType, class CompatibleObjectType>
+struct is_compatible_object_type_impl : std::false_type {};
+
+template<class RealType, class CompatibleObjectType>
+struct is_compatible_object_type_impl<true, RealType, CompatibleObjectType>
+{
+    static constexpr auto value =
+        std::is_constructible<typename RealType::key_type, typename CompatibleObjectType::key_type>::value and
+        std::is_constructible<typename RealType::mapped_type, typename CompatibleObjectType::mapped_type>::value;
+};
+
+template<class BasicJsonType, class CompatibleObjectType>
+struct is_compatible_object_type
+{
+    static auto constexpr value = is_compatible_object_type_impl <
+                                  conjunction<negation<std::is_same<void, CompatibleObjectType>>,
+                                  has_mapped_type<CompatibleObjectType>,
+                                  has_key_type<CompatibleObjectType>>::value,
+                                  typename BasicJsonType::object_t, CompatibleObjectType >::value;
+};
+
+template<typename BasicJsonType, typename T>
+struct is_basic_json_nested_type
+{
+    static auto constexpr value = std::is_same<T, typename BasicJsonType::iterator>::value or
+                                  std::is_same<T, typename BasicJsonType::const_iterator>::value or
+                                  std::is_same<T, typename BasicJsonType::reverse_iterator>::value or
+                                  std::is_same<T, typename BasicJsonType::const_reverse_iterator>::value;
+};
+
+template<class BasicJsonType, class CompatibleArrayType>
+struct is_compatible_array_type
+{
+    static auto constexpr value =
+        conjunction<negation<std::is_same<void, CompatibleArrayType>>,
+        negation<is_compatible_object_type<
+        BasicJsonType, CompatibleArrayType>>,
+        negation<std::is_constructible<typename BasicJsonType::string_t,
+        CompatibleArrayType>>,
+        negation<is_basic_json_nested_type<BasicJsonType, CompatibleArrayType>>,
+        has_value_type<CompatibleArrayType>,
+        has_iterator<CompatibleArrayType>>::value;
+};
+
+template<bool, typename, typename>
+struct is_compatible_integer_type_impl : std::false_type {};
+
+template<typename RealIntegerType, typename CompatibleNumberIntegerType>
+struct is_compatible_integer_type_impl<true, RealIntegerType, CompatibleNumberIntegerType>
+{
+    // is there an assert somewhere on overflows?
+    using RealLimits = std::numeric_limits<RealIntegerType>;
+    using CompatibleLimits = std::numeric_limits<CompatibleNumberIntegerType>;
+
+    static constexpr auto value =
+        std::is_constructible<RealIntegerType, CompatibleNumberIntegerType>::value and
+        CompatibleLimits::is_integer and
+        RealLimits::is_signed == CompatibleLimits::is_signed;
+};
+
+template<typename RealIntegerType, typename CompatibleNumberIntegerType>
+struct is_compatible_integer_type
+{
+    static constexpr auto value =
+        is_compatible_integer_type_impl <
+        std::is_integral<CompatibleNumberIntegerType>::value and
+        not std::is_same<bool, CompatibleNumberIntegerType>::value,
+        RealIntegerType, CompatibleNumberIntegerType > ::value;
+};
+
+// trait checking if JSONSerializer<T>::from_json(json const&, udt&) exists
+template<typename BasicJsonType, typename T>
+struct has_from_json
+{
+  private:
+    // also check the return type of from_json
+    template<typename U, typename = enable_if_t<std::is_same<void, decltype(uncvref_t<U>::from_json(
+                 std::declval<BasicJsonType>(), std::declval<T&>()))>::value>>
+    static int detect(U&&);
+    static void detect(...);
+
+  public:
+    static constexpr bool value = std::is_integral<decltype(
+                                      detect(std::declval<typename BasicJsonType::template json_serializer<T, void>>()))>::value;
+};
+
+// This trait checks if JSONSerializer<T>::from_json(json const&) exists
+// this overload is used for non-default-constructible user-defined-types
+template<typename BasicJsonType, typename T>
+struct has_non_default_from_json
+{
+  private:
+    template <
+        typename U,
+        typename = enable_if_t<std::is_same<
+                                   T, decltype(uncvref_t<U>::from_json(std::declval<BasicJsonType>()))>::value >>
+    static int detect(U&&);
+    static void detect(...);
+
+  public:
+    static constexpr bool value = std::is_integral<decltype(detect(
+                                      std::declval<typename BasicJsonType::template json_serializer<T, void>>()))>::value;
+};
+
+// This trait checks if BasicJsonType::json_serializer<T>::to_json exists
+template<typename BasicJsonType, typename T>
+struct has_to_json
+{
+  private:
+    template<typename U, typename = decltype(uncvref_t<U>::to_json(
+                 std::declval<BasicJsonType&>(), std::declval<T>()))>
+    static int detect(U&&);
+    static void detect(...);
+
+  public:
+    static constexpr bool value = std::is_integral<decltype(detect(
+                                      std::declval<typename BasicJsonType::template json_serializer<T, void>>()))>::value;
+};
+
+template <typename BasicJsonType, typename CompatibleCompleteType>
+struct is_compatible_complete_type
+{
+    static constexpr bool value =
+        not std::is_base_of<std::istream, CompatibleCompleteType>::value and
+        not is_basic_json<CompatibleCompleteType>::value and
+        not is_basic_json_nested_type<BasicJsonType, CompatibleCompleteType>::value and
+        has_to_json<BasicJsonType, CompatibleCompleteType>::value;
+};
+
+template <typename BasicJsonType, typename CompatibleType>
+struct is_compatible_type
+    : conjunction<is_complete_type<CompatibleType>,
+      is_compatible_complete_type<BasicJsonType, CompatibleType>>
+{
+};
+
+// taken from ranges-v3
+template<typename T>
+struct static_const
+{
+    static constexpr T value{};
+};
+
+template<typename T>
+constexpr T static_const<T>::value;
+}
+}

src/lib/json/detail/output/binary_writer.hpp

@@ -0,0 +1,919 @@
+#pragma once
+
+#include <algorithm> // reverse
+#include <array> // array
+#include <cstdint> // uint8_t, uint16_t, uint32_t, uint64_t
+#include <cstring> // memcpy
+#include <limits> // numeric_limits
+
+#include <nlohmann/detail/input/binary_reader.hpp>
+#include <nlohmann/detail/output/output_adapters.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+///////////////////
+// binary writer //
+///////////////////
+
+/*!
+@brief serialization to CBOR and MessagePack values
+*/
+template<typename BasicJsonType, typename CharType>
+class binary_writer
+{
+  public:
+    /*!
+    @brief create a binary writer
+
+    @param[in] adapter  output adapter to write to
+    */
+    explicit binary_writer(output_adapter_t<CharType> adapter) : oa(adapter)
+    {
+        assert(oa);
+    }
+
+    /*!
+    @brief[in] j  JSON value to serialize
+    */
+    void write_cbor(const BasicJsonType& j)
+    {
+        switch (j.type())
+        {
+            case value_t::null:
+            {
+                oa->write_character(static_cast<CharType>(0xF6));
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                oa->write_character(j.m_value.boolean
+                                    ? static_cast<CharType>(0xF5)
+                                    : static_cast<CharType>(0xF4));
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                if (j.m_value.number_integer >= 0)
+                {
+                    // CBOR does not differentiate between positive signed
+                    // integers and unsigned integers. Therefore, we used the
+                    // code from the value_t::number_unsigned case here.
+                    if (j.m_value.number_integer <= 0x17)
+                    {
+                        write_number(static_cast<uint8_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer <= (std::numeric_limits<uint8_t>::max)())
+                    {
+                        oa->write_character(static_cast<CharType>(0x18));
+                        write_number(static_cast<uint8_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer <= (std::numeric_limits<uint16_t>::max)())
+                    {
+                        oa->write_character(static_cast<CharType>(0x19));
+                        write_number(static_cast<uint16_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer <= (std::numeric_limits<uint32_t>::max)())
+                    {
+                        oa->write_character(static_cast<CharType>(0x1A));
+                        write_number(static_cast<uint32_t>(j.m_value.number_integer));
+                    }
+                    else
+                    {
+                        oa->write_character(static_cast<CharType>(0x1B));
+                        write_number(static_cast<uint64_t>(j.m_value.number_integer));
+                    }
+                }
+                else
+                {
+                    // The conversions below encode the sign in the first
+                    // byte, and the value is converted to a positive number.
+                    const auto positive_number = -1 - j.m_value.number_integer;
+                    if (j.m_value.number_integer >= -24)
+                    {
+                        write_number(static_cast<uint8_t>(0x20 + positive_number));
+                    }
+                    else if (positive_number <= (std::numeric_limits<uint8_t>::max)())
+                    {
+                        oa->write_character(static_cast<CharType>(0x38));
+                        write_number(static_cast<uint8_t>(positive_number));
+                    }
+                    else if (positive_number <= (std::numeric_limits<uint16_t>::max)())
+                    {
+                        oa->write_character(static_cast<CharType>(0x39));
+                        write_number(static_cast<uint16_t>(positive_number));
+                    }
+                    else if (positive_number <= (std::numeric_limits<uint32_t>::max)())
+                    {
+                        oa->write_character(static_cast<CharType>(0x3A));
+                        write_number(static_cast<uint32_t>(positive_number));
+                    }
+                    else
+                    {
+                        oa->write_character(static_cast<CharType>(0x3B));
+                        write_number(static_cast<uint64_t>(positive_number));
+                    }
+                }
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                if (j.m_value.number_unsigned <= 0x17)
+                {
+                    write_number(static_cast<uint8_t>(j.m_value.number_unsigned));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x18));
+                    write_number(static_cast<uint8_t>(j.m_value.number_unsigned));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x19));
+                    write_number(static_cast<uint16_t>(j.m_value.number_unsigned));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x1A));
+                    write_number(static_cast<uint32_t>(j.m_value.number_unsigned));
+                }
+                else
+                {
+                    oa->write_character(static_cast<CharType>(0x1B));
+                    write_number(static_cast<uint64_t>(j.m_value.number_unsigned));
+                }
+                break;
+            }
+
+            case value_t::number_float: // Double-Precision Float
+            {
+                oa->write_character(static_cast<CharType>(0xFB));
+                write_number(j.m_value.number_float);
+                break;
+            }
+
+            case value_t::string:
+            {
+                // step 1: write control byte and the string length
+                const auto N = j.m_value.string->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<uint8_t>(0x60 + N));
+                }
+                else if (N <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x78));
+                    write_number(static_cast<uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x79));
+                    write_number(static_cast<uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x7A));
+                    write_number(static_cast<uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<uint64_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x7B));
+                    write_number(static_cast<uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write the string
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_value.string->c_str()),
+                    j.m_value.string->size());
+                break;
+            }
+
+            case value_t::array:
+            {
+                // step 1: write control byte and the array size
+                const auto N = j.m_value.array->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<uint8_t>(0x80 + N));
+                }
+                else if (N <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x98));
+                    write_number(static_cast<uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x99));
+                    write_number(static_cast<uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x9A));
+                    write_number(static_cast<uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<uint64_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0x9B));
+                    write_number(static_cast<uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write each element
+                for (const auto& el : *j.m_value.array)
+                {
+                    write_cbor(el);
+                }
+                break;
+            }
+
+            case value_t::object:
+            {
+                // step 1: write control byte and the object size
+                const auto N = j.m_value.object->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<uint8_t>(0xA0 + N));
+                }
+                else if (N <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0xB8));
+                    write_number(static_cast<uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0xB9));
+                    write_number(static_cast<uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0xBA));
+                    write_number(static_cast<uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<uint64_t>::max)())
+                {
+                    oa->write_character(static_cast<CharType>(0xBB));
+                    write_number(static_cast<uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write each element
+                for (const auto& el : *j.m_value.object)
+                {
+                    write_cbor(el.first);
+                    write_cbor(el.second);
+                }
+                break;
+            }
+
+            default:
+                break;
+        }
+    }
+
+    /*!
+    @brief[in] j  JSON value to serialize
+    */
+    void write_msgpack(const BasicJsonType& j)
+    {
+        switch (j.type())
+        {
+            case value_t::null: // nil
+            {
+                oa->write_character(static_cast<CharType>(0xC0));
+                break;
+            }
+
+            case value_t::boolean: // true and false
+            {
+                oa->write_character(j.m_value.boolean
+                                    ? static_cast<CharType>(0xC3)
+                                    : static_cast<CharType>(0xC2));
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                if (j.m_value.number_integer >= 0)
+                {
+                    // MessagePack does not differentiate between positive
+                    // signed integers and unsigned integers. Therefore, we used
+                    // the code from the value_t::number_unsigned case here.
+                    if (j.m_value.number_unsigned < 128)
+                    {
+                        // positive fixnum
+                        write_number(static_cast<uint8_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_unsigned <= (std::numeric_limits<uint8_t>::max)())
+                    {
+                        // uint 8
+                        oa->write_character(static_cast<CharType>(0xCC));
+                        write_number(static_cast<uint8_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_unsigned <= (std::numeric_limits<uint16_t>::max)())
+                    {
+                        // uint 16
+                        oa->write_character(static_cast<CharType>(0xCD));
+                        write_number(static_cast<uint16_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_unsigned <= (std::numeric_limits<uint32_t>::max)())
+                    {
+                        // uint 32
+                        oa->write_character(static_cast<CharType>(0xCE));
+                        write_number(static_cast<uint32_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_unsigned <= (std::numeric_limits<uint64_t>::max)())
+                    {
+                        // uint 64
+                        oa->write_character(static_cast<CharType>(0xCF));
+                        write_number(static_cast<uint64_t>(j.m_value.number_integer));
+                    }
+                }
+                else
+                {
+                    if (j.m_value.number_integer >= -32)
+                    {
+                        // negative fixnum
+                        write_number(static_cast<int8_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer >= (std::numeric_limits<int8_t>::min)() and
+                             j.m_value.number_integer <= (std::numeric_limits<int8_t>::max)())
+                    {
+                        // int 8
+                        oa->write_character(static_cast<CharType>(0xD0));
+                        write_number(static_cast<int8_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer >= (std::numeric_limits<int16_t>::min)() and
+                             j.m_value.number_integer <= (std::numeric_limits<int16_t>::max)())
+                    {
+                        // int 16
+                        oa->write_character(static_cast<CharType>(0xD1));
+                        write_number(static_cast<int16_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer >= (std::numeric_limits<int32_t>::min)() and
+                             j.m_value.number_integer <= (std::numeric_limits<int32_t>::max)())
+                    {
+                        // int 32
+                        oa->write_character(static_cast<CharType>(0xD2));
+                        write_number(static_cast<int32_t>(j.m_value.number_integer));
+                    }
+                    else if (j.m_value.number_integer >= (std::numeric_limits<int64_t>::min)() and
+                             j.m_value.number_integer <= (std::numeric_limits<int64_t>::max)())
+                    {
+                        // int 64
+                        oa->write_character(static_cast<CharType>(0xD3));
+                        write_number(static_cast<int64_t>(j.m_value.number_integer));
+                    }
+                }
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                if (j.m_value.number_unsigned < 128)
+                {
+                    // positive fixnum
+                    write_number(static_cast<uint8_t>(j.m_value.number_integer));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    // uint 8
+                    oa->write_character(static_cast<CharType>(0xCC));
+                    write_number(static_cast<uint8_t>(j.m_value.number_integer));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    // uint 16
+                    oa->write_character(static_cast<CharType>(0xCD));
+                    write_number(static_cast<uint16_t>(j.m_value.number_integer));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    // uint 32
+                    oa->write_character(static_cast<CharType>(0xCE));
+                    write_number(static_cast<uint32_t>(j.m_value.number_integer));
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint64_t>::max)())
+                {
+                    // uint 64
+                    oa->write_character(static_cast<CharType>(0xCF));
+                    write_number(static_cast<uint64_t>(j.m_value.number_integer));
+                }
+                break;
+            }
+
+            case value_t::number_float: // float 64
+            {
+                oa->write_character(static_cast<CharType>(0xCB));
+                write_number(j.m_value.number_float);
+                break;
+            }
+
+            case value_t::string:
+            {
+                // step 1: write control byte and the string length
+                const auto N = j.m_value.string->size();
+                if (N <= 31)
+                {
+                    // fixstr
+                    write_number(static_cast<uint8_t>(0xA0 | N));
+                }
+                else if (N <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    // str 8
+                    oa->write_character(static_cast<CharType>(0xD9));
+                    write_number(static_cast<uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    // str 16
+                    oa->write_character(static_cast<CharType>(0xDA));
+                    write_number(static_cast<uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    // str 32
+                    oa->write_character(static_cast<CharType>(0xDB));
+                    write_number(static_cast<uint32_t>(N));
+                }
+
+                // step 2: write the string
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_value.string->c_str()),
+                    j.m_value.string->size());
+                break;
+            }
+
+            case value_t::array:
+            {
+                // step 1: write control byte and the array size
+                const auto N = j.m_value.array->size();
+                if (N <= 15)
+                {
+                    // fixarray
+                    write_number(static_cast<uint8_t>(0x90 | N));
+                }
+                else if (N <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    // array 16
+                    oa->write_character(static_cast<CharType>(0xDC));
+                    write_number(static_cast<uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    // array 32
+                    oa->write_character(static_cast<CharType>(0xDD));
+                    write_number(static_cast<uint32_t>(N));
+                }
+
+                // step 2: write each element
+                for (const auto& el : *j.m_value.array)
+                {
+                    write_msgpack(el);
+                }
+                break;
+            }
+
+            case value_t::object:
+            {
+                // step 1: write control byte and the object size
+                const auto N = j.m_value.object->size();
+                if (N <= 15)
+                {
+                    // fixmap
+                    write_number(static_cast<uint8_t>(0x80 | (N & 0xF)));
+                }
+                else if (N <= (std::numeric_limits<uint16_t>::max)())
+                {
+                    // map 16
+                    oa->write_character(static_cast<CharType>(0xDE));
+                    write_number(static_cast<uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<uint32_t>::max)())
+                {
+                    // map 32
+                    oa->write_character(static_cast<CharType>(0xDF));
+                    write_number(static_cast<uint32_t>(N));
+                }
+
+                // step 2: write each element
+                for (const auto& el : *j.m_value.object)
+                {
+                    write_msgpack(el.first);
+                    write_msgpack(el.second);
+                }
+                break;
+            }
+
+            default:
+                break;
+        }
+    }
+
+    /*!
+    @param[in] j  JSON value to serialize
+    @param[in] use_count   whether to use '#' prefixes (optimized format)
+    @param[in] use_type    whether to use '$' prefixes (optimized format)
+    @param[in] add_prefix  whether prefixes need to be used for this value
+    */
+    void write_ubjson(const BasicJsonType& j, const bool use_count,
+                      const bool use_type, const bool add_prefix = true)
+    {
+        switch (j.type())
+        {
+            case value_t::null:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(static_cast<CharType>('Z'));
+                }
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                if (add_prefix)
+                    oa->write_character(j.m_value.boolean
+                                        ? static_cast<CharType>('T')
+                                        : static_cast<CharType>('F'));
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                write_number_with_ubjson_prefix(j.m_value.number_integer, add_prefix);
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                write_number_with_ubjson_prefix(j.m_value.number_unsigned, add_prefix);
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                write_number_with_ubjson_prefix(j.m_value.number_float, add_prefix);
+                break;
+            }
+
+            case value_t::string:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(static_cast<CharType>('S'));
+                }
+                write_number_with_ubjson_prefix(j.m_value.string->size(), true);
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_value.string->c_str()),
+                    j.m_value.string->size());
+                break;
+            }
+
+            case value_t::array:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(static_cast<CharType>('['));
+                }
+
+                bool prefix_required = true;
+                if (use_type and not j.m_value.array->empty())
+                {
+                    assert(use_count);
+                    const char first_prefix = ubjson_prefix(j.front());
+                    const bool same_prefix = std::all_of(j.begin() + 1, j.end(),
+                                                         [this, first_prefix](const BasicJsonType & v)
+                    {
+                        return ubjson_prefix(v) == first_prefix;
+                    });
+
+                    if (same_prefix)
+                    {
+                        prefix_required = false;
+                        oa->write_character(static_cast<CharType>('$'));
+                        oa->write_character(static_cast<CharType>(first_prefix));
+                    }
+                }
+
+                if (use_count)
+                {
+                    oa->write_character(static_cast<CharType>('#'));
+                    write_number_with_ubjson_prefix(j.m_value.array->size(), true);
+                }
+
+                for (const auto& el : *j.m_value.array)
+                {
+                    write_ubjson(el, use_count, use_type, prefix_required);
+                }
+
+                if (not use_count)
+                {
+                    oa->write_character(static_cast<CharType>(']'));
+                }
+
+                break;
+            }
+
+            case value_t::object:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(static_cast<CharType>('{'));
+                }
+
+                bool prefix_required = true;
+                if (use_type and not j.m_value.object->empty())
+                {
+                    assert(use_count);
+                    const char first_prefix = ubjson_prefix(j.front());
+                    const bool same_prefix = std::all_of(j.begin(), j.end(),
+                                                         [this, first_prefix](const BasicJsonType & v)
+                    {
+                        return ubjson_prefix(v) == first_prefix;
+                    });
+
+                    if (same_prefix)
+                    {
+                        prefix_required = false;
+                        oa->write_character(static_cast<CharType>('$'));
+                        oa->write_character(static_cast<CharType>(first_prefix));
+                    }
+                }
+
+                if (use_count)
+                {
+                    oa->write_character(static_cast<CharType>('#'));
+                    write_number_with_ubjson_prefix(j.m_value.object->size(), true);
+                }
+
+                for (const auto& el : *j.m_value.object)
+                {
+                    write_number_with_ubjson_prefix(el.first.size(), true);
+                    oa->write_characters(
+                        reinterpret_cast<const CharType*>(el.first.c_str()),
+                        el.first.size());
+                    write_ubjson(el.second, use_count, use_type, prefix_required);
+                }
+
+                if (not use_count)
+                {
+                    oa->write_character(static_cast<CharType>('}'));
+                }
+
+                break;
+            }
+
+            default:
+                break;
+        }
+    }
+
+  private:
+    /*
+    @brief write a number to output input
+
+    @param[in] n number of type @a NumberType
+    @tparam NumberType the type of the number
+
+    @note This function needs to respect the system's endianess, because bytes
+          in CBOR, MessagePack, and UBJSON are stored in network order (big
+          endian) and therefore need reordering on little endian systems.
+    */
+    template<typename NumberType>
+    void write_number(const NumberType n)
+    {
+        // step 1: write number to array of length NumberType
+        std::array<CharType, sizeof(NumberType)> vec;
+        std::memcpy(vec.data(), &n, sizeof(NumberType));
+
+        // step 2: write array to output (with possible reordering)
+        if (is_little_endian)
+        {
+            // reverse byte order prior to conversion if necessary
+            std::reverse(vec.begin(), vec.end());
+        }
+
+        oa->write_characters(vec.data(), sizeof(NumberType));
+    }
+
+    // UBJSON: write number (floating point)
+    template<typename NumberType, typename std::enable_if<
+                 std::is_floating_point<NumberType>::value, int>::type = 0>
+    void write_number_with_ubjson_prefix(const NumberType n,
+                                         const bool add_prefix)
+    {
+        if (add_prefix)
+        {
+            oa->write_character(static_cast<CharType>('D'));  // float64
+        }
+        write_number(n);
+    }
+
+    // UBJSON: write number (unsigned integer)
+    template<typename NumberType, typename std::enable_if<
+                 std::is_unsigned<NumberType>::value, int>::type = 0>
+    void write_number_with_ubjson_prefix(const NumberType n,
+                                         const bool add_prefix)
+    {
+        if (n <= static_cast<uint64_t>((std::numeric_limits<int8_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('i'));  // int8
+            }
+            write_number(static_cast<uint8_t>(n));
+        }
+        else if (n <= (std::numeric_limits<uint8_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('U'));  // uint8
+            }
+            write_number(static_cast<uint8_t>(n));
+        }
+        else if (n <= static_cast<uint64_t>((std::numeric_limits<int16_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('I'));  // int16
+            }
+            write_number(static_cast<int16_t>(n));
+        }
+        else if (n <= static_cast<uint64_t>((std::numeric_limits<int32_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('l'));  // int32
+            }
+            write_number(static_cast<int32_t>(n));
+        }
+        else if (n <= static_cast<uint64_t>((std::numeric_limits<int64_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('L'));  // int64
+            }
+            write_number(static_cast<int64_t>(n));
+        }
+        else
+        {
+            JSON_THROW(out_of_range::create(407, "number overflow serializing " + std::to_string(n)));
+        }
+    }
+
+    // UBJSON: write number (signed integer)
+    template<typename NumberType, typename std::enable_if<
+                 std::is_signed<NumberType>::value and
+                 not std::is_floating_point<NumberType>::value, int>::type = 0>
+    void write_number_with_ubjson_prefix(const NumberType n,
+                                         const bool add_prefix)
+    {
+        if ((std::numeric_limits<int8_t>::min)() <= n and n <= (std::numeric_limits<int8_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('i'));  // int8
+            }
+            write_number(static_cast<int8_t>(n));
+        }
+        else if (static_cast<int64_t>((std::numeric_limits<uint8_t>::min)()) <= n and n <= static_cast<int64_t>((std::numeric_limits<uint8_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('U'));  // uint8
+            }
+            write_number(static_cast<uint8_t>(n));
+        }
+        else if ((std::numeric_limits<int16_t>::min)() <= n and n <= (std::numeric_limits<int16_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('I'));  // int16
+            }
+            write_number(static_cast<int16_t>(n));
+        }
+        else if ((std::numeric_limits<int32_t>::min)() <= n and n <= (std::numeric_limits<int32_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('l'));  // int32
+            }
+            write_number(static_cast<int32_t>(n));
+        }
+        else if ((std::numeric_limits<int64_t>::min)() <= n and n <= (std::numeric_limits<int64_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(static_cast<CharType>('L'));  // int64
+            }
+            write_number(static_cast<int64_t>(n));
+        }
+        // LCOV_EXCL_START
+        else
+        {
+            JSON_THROW(out_of_range::create(407, "number overflow serializing " + std::to_string(n)));
+        }
+        // LCOV_EXCL_STOP
+    }
+
+    /*!
+    @brief determine the type prefix of container values
+
+    @note This function does not need to be 100% accurate when it comes to
+          integer limits. In case a number exceeds the limits of int64_t,
+          this will be detected by a later call to function
+          write_number_with_ubjson_prefix. Therefore, we return 'L' for any
+          value that does not fit the previous limits.
+    */
+    char ubjson_prefix(const BasicJsonType& j) const noexcept
+    {
+        switch (j.type())
+        {
+            case value_t::null:
+                return 'Z';
+
+            case value_t::boolean:
+                return j.m_value.boolean ? 'T' : 'F';
+
+            case value_t::number_integer:
+            {
+                if ((std::numeric_limits<int8_t>::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits<int8_t>::max)())
+                {
+                    return 'i';
+                }
+                else if ((std::numeric_limits<uint8_t>::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    return 'U';
+                }
+                else if ((std::numeric_limits<int16_t>::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits<int16_t>::max)())
+                {
+                    return 'I';
+                }
+                else if ((std::numeric_limits<int32_t>::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits<int32_t>::max)())
+                {
+                    return 'l';
+                }
+                else  // no check and assume int64_t (see note above)
+                {
+                    return 'L';
+                }
+            }
+
+            case value_t::number_unsigned:
+            {
+                if (j.m_value.number_unsigned <= (std::numeric_limits<int8_t>::max)())
+                {
+                    return 'i';
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<uint8_t>::max)())
+                {
+                    return 'U';
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<int16_t>::max)())
+                {
+                    return 'I';
+                }
+                else if (j.m_value.number_unsigned <= (std::numeric_limits<int32_t>::max)())
+                {
+                    return 'l';
+                }
+                else  // no check and assume int64_t (see note above)
+                {
+                    return 'L';
+                }
+            }
+
+            case value_t::number_float:
+                return 'D';
+
+            case value_t::string:
+                return 'S';
+
+            case value_t::array:
+                return '[';
+
+            case value_t::object:
+                return '{';
+
+            default:  // discarded values
+                return 'N';
+        }
+    }
+
+  private:
+    /// whether we can assume little endianess
+    const bool is_little_endian = binary_reader<BasicJsonType>::little_endianess();
+
+    /// the output
+    output_adapter_t<CharType> oa = nullptr;
+};
+}
+}

src/lib/json/detail/output/output_adapters.hpp

@@ -0,0 +1,113 @@
+#pragma once
+
+#include <algorithm> // copy
+#include <cstddef> // size_t
+#include <ios> // streamsize
+#include <iterator> // back_inserter
+#include <memory> // shared_ptr, make_shared
+#include <ostream> // basic_ostream
+#include <string> // basic_string
+#include <vector> // vector
+
+namespace nlohmann
+{
+namespace detail
+{
+/// abstract output adapter interface
+template<typename CharType> struct output_adapter_protocol
+{
+    virtual void write_character(CharType c) = 0;
+    virtual void write_characters(const CharType* s, std::size_t length) = 0;
+    virtual ~output_adapter_protocol() = default;
+};
+
+/// a type to simplify interfaces
+template<typename CharType>
+using output_adapter_t = std::shared_ptr<output_adapter_protocol<CharType>>;
+
+/// output adapter for byte vectors
+template<typename CharType>
+class output_vector_adapter : public output_adapter_protocol<CharType>
+{
+  public:
+    explicit output_vector_adapter(std::vector<CharType>& vec) : v(vec) {}
+
+    void write_character(CharType c) override
+    {
+        v.push_back(c);
+    }
+
+    void write_characters(const CharType* s, std::size_t length) override
+    {
+        std::copy(s, s + length, std::back_inserter(v));
+    }
+
+  private:
+    std::vector<CharType>& v;
+};
+
+/// output adapter for output streams
+template<typename CharType>
+class output_stream_adapter : public output_adapter_protocol<CharType>
+{
+  public:
+    explicit output_stream_adapter(std::basic_ostream<CharType>& s) : stream(s) {}
+
+    void write_character(CharType c) override
+    {
+        stream.put(c);
+    }
+
+    void write_characters(const CharType* s, std::size_t length) override
+    {
+        stream.write(s, static_cast<std::streamsize>(length));
+    }
+
+  private:
+    std::basic_ostream<CharType>& stream;
+};
+
+/// output adapter for basic_string
+template<typename CharType, typename StringType = std::basic_string<CharType>>
+class output_string_adapter : public output_adapter_protocol<CharType>
+{
+  public:
+    explicit output_string_adapter(StringType& s) : str(s) {}
+
+    void write_character(CharType c) override
+    {
+        str.push_back(c);
+    }
+
+    void write_characters(const CharType* s, std::size_t length) override
+    {
+        str.append(s, length);
+    }
+
+  private:
+    StringType& str;
+};
+
+template<typename CharType, typename StringType = std::basic_string<CharType>>
+class output_adapter
+{
+  public:
+    output_adapter(std::vector<CharType>& vec)
+        : oa(std::make_shared<output_vector_adapter<CharType>>(vec)) {}
+
+    output_adapter(std::basic_ostream<CharType>& s)
+        : oa(std::make_shared<output_stream_adapter<CharType>>(s)) {}
+
+    output_adapter(StringType& s)
+        : oa(std::make_shared<output_string_adapter<CharType, StringType>>(s)) {}
+
+    operator output_adapter_t<CharType>()
+    {
+        return oa;
+    }
+
+  private:
+    output_adapter_t<CharType> oa = nullptr;
+};
+}
+}

src/lib/json/detail/output/serializer.hpp

@@ -0,0 +1,634 @@
+#pragma once
+
+#include <algorithm> // reverse, remove, fill, find, none_of
+#include <array> // array
+#include <cassert> // assert
+#include <ciso646> // and, or
+#include <clocale> // localeconv, lconv
+#include <cmath> // labs, isfinite, isnan, signbit
+#include <cstddef> // size_t, ptrdiff_t
+#include <cstdint> // uint8_t
+#include <cstdio> // snprintf
+#include <iomanip> // setfill
+#include <iterator> // next
+#include <limits> // numeric_limits
+#include <string> // string
+#include <sstream> // stringstream
+#include <type_traits> // is_same
+
+#include <nlohmann/detail/exceptions.hpp>
+#include <nlohmann/detail/conversions/to_chars.hpp>
+#include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/meta.hpp>
+#include <nlohmann/detail/output/output_adapters.hpp>
+#include <nlohmann/detail/value_t.hpp>
+
+namespace nlohmann
+{
+namespace detail
+{
+///////////////////
+// serialization //
+///////////////////
+
+template<typename BasicJsonType>
+class serializer
+{
+    using string_t = typename BasicJsonType::string_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    static constexpr uint8_t UTF8_ACCEPT = 0;
+    static constexpr uint8_t UTF8_REJECT = 1;
+
+  public:
+    /*!
+    @param[in] s  output stream to serialize to
+    @param[in] ichar  indentation character to use
+    */
+    serializer(output_adapter_t<char> s, const char ichar)
+        : o(std::move(s)), loc(std::localeconv()),
+          thousands_sep(loc->thousands_sep == nullptr ? '\0' : * (loc->thousands_sep)),
+          decimal_point(loc->decimal_point == nullptr ? '\0' : * (loc->decimal_point)),
+          indent_char(ichar), indent_string(512, indent_char)
+    {}
+
+    // delete because of pointer members
+    serializer(const serializer&) = delete;
+    serializer& operator=(const serializer&) = delete;
+
+    /*!
+    @brief internal implementation of the serialization function
+
+    This function is called by the public member function dump and organizes
+    the serialization internally. The indentation level is propagated as
+    additional parameter. In case of arrays and objects, the function is
+    called recursively.
+
+    - strings and object keys are escaped using `escape_string()`
+    - integer numbers are converted implicitly via `operator<<`
+    - floating-point numbers are converted to a string using `"%g"` format
+
+    @param[in] val             value to serialize
+    @param[in] pretty_print    whether the output shall be pretty-printed
+    @param[in] indent_step     the indent level
+    @param[in] current_indent  the current indent level (only used internally)
+    */
+    void dump(const BasicJsonType& val, const bool pretty_print,
+              const bool ensure_ascii,
+              const unsigned int indent_step,
+              const unsigned int current_indent = 0)
+    {
+        switch (val.m_type)
+        {
+            case value_t::object:
+            {
+                if (val.m_value.object->empty())
+                {
+                    o->write_characters("{}", 2);
+                    return;
+                }
+
+                if (pretty_print)
+                {
+                    o->write_characters("{\n", 2);
+
+                    // variable to hold indentation for recursive calls
+                    const auto new_indent = current_indent + indent_step;
+                    if (JSON_UNLIKELY(indent_string.size() < new_indent))
+                    {
+                        indent_string.resize(indent_string.size() * 2, ' ');
+                    }
+
+                    // first n-1 elements
+                    auto i = val.m_value.object->cbegin();
+                    for (std::size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i)
+                    {
+                        o->write_characters(indent_string.c_str(), new_indent);
+                        o->write_character('\"');
+                        dump_escaped(i->first, ensure_ascii);
+                        o->write_characters("\": ", 3);
+                        dump(i->second, true, ensure_ascii, indent_step, new_indent);
+                        o->write_characters(",\n", 2);
+                    }
+
+                    // last element
+                    assert(i != val.m_value.object->cend());
+                    assert(std::next(i) == val.m_value.object->cend());
+                    o->write_characters(indent_string.c_str(), new_indent);
+                    o->write_character('\"');
+                    dump_escaped(i->first, ensure_ascii);
+                    o->write_characters("\": ", 3);
+                    dump(i->second, true, ensure_ascii, indent_step, new_indent);
+
+                    o->write_character('\n');
+                    o->write_characters(indent_string.c_str(), current_indent);
+                    o->write_character('}');
+                }
+                else
+                {
+                    o->write_character('{');
+
+                    // first n-1 elements
+                    auto i = val.m_value.object->cbegin();
+                    for (std::size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i)
+                    {
+                        o->write_character('\"');
+                        dump_escaped(i->first, ensure_ascii);
+                        o->write_characters("\":", 2);
+                        dump(i->second, false, ensure_ascii, indent_step, current_indent);
+                        o->write_character(',');
+                    }
+
+                    // last element
+                    assert(i != val.m_value.object->cend());
+                    assert(std::next(i) == val.m_value.object->cend());
+                    o->write_character('\"');
+                    dump_escaped(i->first, ensure_ascii);
+                    o->write_characters("\":", 2);
+                    dump(i->second, false, ensure_ascii, indent_step, current_indent);
+
+                    o->write_character('}');
+                }
+
+                return;
+            }
+
+            case value_t::array:
+            {
+                if (val.m_value.array->empty())
+                {
+                    o->write_characters("[]", 2);
+                    return;
+                }
+
+                if (pretty_print)
+                {
+                    o->write_characters("[\n", 2);
+
+                    // variable to hold indentation for recursive calls
+                    const auto new_indent = current_indent + indent_step;
+                    if (JSON_UNLIKELY(indent_string.size() < new_indent))
+                    {
+                        indent_string.resize(indent_string.size() * 2, ' ');
+                    }
+
+                    // first n-1 elements
+                    for (auto i = val.m_value.array->cbegin();
+                            i != val.m_value.array->cend() - 1; ++i)
+                    {
+                        o->write_characters(indent_string.c_str(), new_indent);
+                        dump(*i, true, ensure_ascii, indent_step, new_indent);
+                        o->write_characters(",\n", 2);
+                    }
+
+                    // last element
+                    assert(not val.m_value.array->empty());
+                    o->write_characters(indent_string.c_str(), new_indent);
+                    dump(val.m_value.array->back(), true, ensure_ascii, indent_step, new_indent);
+
+                    o->write_character('\n');
+                    o->write_characters(indent_string.c_str(), current_indent);
+                    o->write_character(']');
+                }
+                else
+                {
+                    o->write_character('[');
+
+                    // first n-1 elements
+                    for (auto i = val.m_value.array->cbegin();
+                            i != val.m_value.array->cend() - 1; ++i)
+                    {
+                        dump(*i, false, ensure_ascii, indent_step, current_indent);
+                        o->write_character(',');
+                    }
+
+                    // last element
+                    assert(not val.m_value.array->empty());
+                    dump(val.m_value.array->back(), false, ensure_ascii, indent_step, current_indent);
+
+                    o->write_character(']');
+                }
+
+                return;
+            }
+
+            case value_t::string:
+            {
+                o->write_character('\"');
+                dump_escaped(*val.m_value.string, ensure_ascii);
+                o->write_character('\"');
+                return;
+            }
+
+            case value_t::boolean:
+            {
+                if (val.m_value.boolean)
+                {
+                    o->write_characters("true", 4);
+                }
+                else
+                {
+                    o->write_characters("false", 5);
+                }
+                return;
+            }
+
+            case value_t::number_integer:
+            {
+                dump_integer(val.m_value.number_integer);
+                return;
+            }
+
+            case value_t::number_unsigned:
+            {
+                dump_integer(val.m_value.number_unsigned);
+                return;
+            }
+
+            case value_t::number_float:
+            {
+                dump_float(val.m_value.number_float);
+                return;
+            }
+
+            case value_t::discarded:
+            {
+                o->write_characters("<discarded>", 11);
+                return;
+            }
+
+            case value_t::null:
+            {
+                o->write_characters("null", 4);
+                return;
+            }
+        }
+    }
+
+  private:
+    /*!
+    @brief dump escaped string
+
+    Escape a string by replacing certain special characters by a sequence of an
+    escape character (backslash) and another character and other control
+    characters by a sequence of "\u" followed by a four-digit hex
+    representation. The escaped string is written to output stream @a o.
+
+    @param[in] s  the string to escape
+    @param[in] ensure_ascii  whether to escape non-ASCII characters with
+                             \uXXXX sequences
+
+    @complexity Linear in the length of string @a s.
+    */
+    void dump_escaped(const string_t& s, const bool ensure_ascii)
+    {
+        uint32_t codepoint;
+        uint8_t state = UTF8_ACCEPT;
+        std::size_t bytes = 0;  // number of bytes written to string_buffer
+
+        for (std::size_t i = 0; i < s.size(); ++i)
+        {
+            const auto byte = static_cast<uint8_t>(s[i]);
+
+            switch (decode(state, codepoint, byte))
+            {
+                case UTF8_ACCEPT:  // decode found a new code point
+                {
+                    switch (codepoint)
+                    {
+                        case 0x08: // backspace
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'b';
+                            break;
+                        }
+
+                        case 0x09: // horizontal tab
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 't';
+                            break;
+                        }
+
+                        case 0x0A: // newline
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'n';
+                            break;
+                        }
+
+                        case 0x0C: // formfeed
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'f';
+                            break;
+                        }
+
+                        case 0x0D: // carriage return
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'r';
+                            break;
+                        }
+
+                        case 0x22: // quotation mark
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = '\"';
+                            break;
+                        }
+
+                        case 0x5C: // reverse solidus
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = '\\';
+                            break;
+                        }
+
+                        default:
+                        {
+                            // escape control characters (0x00..0x1F) or, if
+                            // ensure_ascii parameter is used, non-ASCII characters
+                            if ((codepoint <= 0x1F) or (ensure_ascii and (codepoint >= 0x7F)))
+                            {
+                                if (codepoint <= 0xFFFF)
+                                {
+                                    std::snprintf(string_buffer.data() + bytes, 7, "\\u%04x",
+                                                  static_cast<uint16_t>(codepoint));
+                                    bytes += 6;
+                                }
+                                else
+                                {
+                                    std::snprintf(string_buffer.data() + bytes, 13, "\\u%04x\\u%04x",
+                                                  static_cast<uint16_t>(0xD7C0 + (codepoint >> 10)),
+                                                  static_cast<uint16_t>(0xDC00 + (codepoint & 0x3FF)));
+                                    bytes += 12;
+                                }
+                            }
+                            else
+                            {
+                                // copy byte to buffer (all previous bytes
+                                // been copied have in default case above)
+                                string_buffer[bytes++] = s[i];
+                            }
+                            break;
+                        }
+                    }
+
+                    // write buffer and reset index; there must be 13 bytes
+                    // left, as this is the maximal number of bytes to be
+                    // written ("\uxxxx\uxxxx\0") for one code point
+                    if (string_buffer.size() - bytes < 13)
+                    {
+                        o->write_characters(string_buffer.data(), bytes);
+                        bytes = 0;
+                    }
+                    break;
+                }
+
+                case UTF8_REJECT:  // decode found invalid UTF-8 byte
+                {
+                    std::stringstream ss;
+                    ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << static_cast<int>(byte);
+                    JSON_THROW(type_error::create(316, "invalid UTF-8 byte at index " + std::to_string(i) + ": 0x" + ss.str()));
+                }
+
+                default:  // decode found yet incomplete multi-byte code point
+                {
+                    if (not ensure_ascii)
+                    {
+                        // code point will not be escaped - copy byte to buffer
+                        string_buffer[bytes++] = s[i];
+                    }
+                    break;
+                }
+            }
+        }
+
+        if (JSON_LIKELY(state == UTF8_ACCEPT))
+        {
+            // write buffer
+            if (bytes > 0)
+            {
+                o->write_characters(string_buffer.data(), bytes);
+            }
+        }
+        else
+        {
+            // we finish reading, but do not accept: string was incomplete
+            std::stringstream ss;
+            ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << static_cast<int>(static_cast<uint8_t>(s.back()));
+            JSON_THROW(type_error::create(316, "incomplete UTF-8 string; last byte: 0x" + ss.str()));
+        }
+    }
+
+    /*!
+    @brief dump an integer
+
+    Dump a given integer to output stream @a o. Works internally with
+    @a number_buffer.
+
+    @param[in] x  integer number (signed or unsigned) to dump
+    @tparam NumberType either @a number_integer_t or @a number_unsigned_t
+    */
+    template<typename NumberType, detail::enable_if_t<
+                 std::is_same<NumberType, number_unsigned_t>::value or
+                 std::is_same<NumberType, number_integer_t>::value,
+                 int> = 0>
+    void dump_integer(NumberType x)
+    {
+        // special case for "0"
+        if (x == 0)
+        {
+            o->write_character('0');
+            return;
+        }
+
+        const bool is_negative = (x <= 0) and (x != 0);  // see issue #755
+        std::size_t i = 0;
+
+        while (x != 0)
+        {
+            // spare 1 byte for '\0'
+            assert(i < number_buffer.size() - 1);
+
+            const auto digit = std::labs(static_cast<long>(x % 10));
+            number_buffer[i++] = static_cast<char>('0' + digit);
+            x /= 10;
+        }
+
+        if (is_negative)
+        {
+            // make sure there is capacity for the '-'
+            assert(i < number_buffer.size() - 2);
+            number_buffer[i++] = '-';
+        }
+
+        std::reverse(number_buffer.begin(), number_buffer.begin() + i);
+        o->write_characters(number_buffer.data(), i);
+    }
+
+    /*!
+    @brief dump a floating-point number
+
+    Dump a given floating-point number to output stream @a o. Works internally
+    with @a number_buffer.
+
+    @param[in] x  floating-point number to dump
+    */
+    void dump_float(number_float_t x)
+    {
+        // NaN / inf
+        if (not std::isfinite(x))
+        {
+            o->write_characters("null", 4);
+            return;
+        }
+
+        // If number_float_t is an IEEE-754 single or double precision number,
+        // use the Grisu2 algorithm to produce short numbers which are
+        // guaranteed to round-trip, using strtof and strtod, resp.
+        //
+        // NB: The test below works if <long double> == <double>.
+        static constexpr bool is_ieee_single_or_double
+            = (std::numeric_limits<number_float_t>::is_iec559 and std::numeric_limits<number_float_t>::digits == 24 and std::numeric_limits<number_float_t>::max_exponent == 128) or
+              (std::numeric_limits<number_float_t>::is_iec559 and std::numeric_limits<number_float_t>::digits == 53 and std::numeric_limits<number_float_t>::max_exponent == 1024);
+
+        dump_float(x, std::integral_constant<bool, is_ieee_single_or_double>());
+    }
+
+    void dump_float(number_float_t x, std::true_type /*is_ieee_single_or_double*/)
+    {
+        char* begin = number_buffer.data();
+        char* end = ::nlohmann::detail::to_chars(begin, begin + number_buffer.size(), x);
+
+        o->write_characters(begin, static_cast<size_t>(end - begin));
+    }
+
+    void dump_float(number_float_t x, std::false_type /*is_ieee_single_or_double*/)
+    {
+        // get number of digits for a float -> text -> float round-trip
+        static constexpr auto d = std::numeric_limits<number_float_t>::max_digits10;
+
+        // the actual conversion
+        std::ptrdiff_t len = snprintf(number_buffer.data(), number_buffer.size(), "%.*g", d, x);
+
+        // negative value indicates an error
+        assert(len > 0);
+        // check if buffer was large enough
+        assert(static_cast<std::size_t>(len) < number_buffer.size());
+
+        // erase thousands separator
+        if (thousands_sep != '\0')
+        {
+            const auto end = std::remove(number_buffer.begin(),
+                                         number_buffer.begin() + len, thousands_sep);
+            std::fill(end, number_buffer.end(), '\0');
+            assert((end - number_buffer.begin()) <= len);
+            len = (end - number_buffer.begin());
+        }
+
+        // convert decimal point to '.'
+        if (decimal_point != '\0' and decimal_point != '.')
+        {
+            const auto dec_pos = std::find(number_buffer.begin(), number_buffer.end(), decimal_point);
+            if (dec_pos != number_buffer.end())
+            {
+                *dec_pos = '.';
+            }
+        }
+
+        o->write_characters(number_buffer.data(), static_cast<std::size_t>(len));
+
+        // determine if need to append ".0"
+        const bool value_is_int_like =
+            std::none_of(number_buffer.begin(), number_buffer.begin() + len + 1,
+                         [](char c)
+        {
+            return (c == '.' or c == 'e');
+        });
+
+        if (value_is_int_like)
+        {
+            o->write_characters(".0", 2);
+        }
+    }
+
+    /*!
+    @brief check whether a string is UTF-8 encoded
+
+    The function checks each byte of a string whether it is UTF-8 encoded. The
+    result of the check is stored in the @a state parameter. The function must
+    be called initially with state 0 (accept). State 1 means the string must
+    be rejected, because the current byte is not allowed. If the string is
+    completely processed, but the state is non-zero, the string ended
+    prematurely; that is, the last byte indicated more bytes should have
+    followed.
+
+    @param[in,out] state  the state of the decoding
+    @param[in,out] codep  codepoint (valid only if resulting state is UTF8_ACCEPT)
+    @param[in] byte       next byte to decode
+    @return               new state
+
+    @note The function has been edited: a std::array is used.
+
+    @copyright Copyright (c) 2008-2009 Bjoern Hoehrmann <bjoern@hoehrmann.de>
+    @sa http://bjoern.hoehrmann.de/utf-8/decoder/dfa/
+    */
+    static uint8_t decode(uint8_t& state, uint32_t& codep, const uint8_t byte) noexcept
+    {
+        static const std::array<uint8_t, 400> utf8d =
+        {
+            {
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 00..1F
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 20..3F
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 40..5F
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 60..7F
+                1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, // 80..9F
+                7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, // A0..BF
+                8, 8, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // C0..DF
+                0xA, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x4, 0x3, 0x3, // E0..EF
+                0xB, 0x6, 0x6, 0x6, 0x5, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, // F0..FF
+                0x0, 0x1, 0x2, 0x3, 0x5, 0x8, 0x7, 0x1, 0x1, 0x1, 0x4, 0x6, 0x1, 0x1, 0x1, 0x1, // s0..s0
+                1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, // s1..s2
+                1, 2, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, // s3..s4
+                1, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, // s5..s6
+                1, 3, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 // s7..s8
+            }
+        };
+
+        const uint8_t type = utf8d[byte];
+
+        codep = (state != UTF8_ACCEPT)
+                ? (byte & 0x3fu) | (codep << 6)
+                : static_cast<uint32_t>(0xff >> type) & (byte);
+
+        state = utf8d[256u + state * 16u + type];
+        return state;
+    }
+
+  private:
+    /// the output of the serializer
+    output_adapter_t<char> o = nullptr;
+
+    /// a (hopefully) large enough character buffer
+    std::array<char, 64> number_buffer{{}};
+
+    /// the locale
+    const std::lconv* loc = nullptr;
+    /// the locale's thousand separator character
+    const char thousands_sep = '\0';
+    /// the locale's decimal point character
+    const char decimal_point = '\0';
+
+    /// string buffer
+    std::array<char, 512> string_buffer{{}};
+
+    /// the indentation character
+    const char indent_char;
+    /// the indentation string
+    string_t indent_string;
+};
+}
+}

src/lib/json/detail/value_t.hpp

@@ -0,0 +1,76 @@
+#pragma once
+
+#include <array> // array
+#include <ciso646> // and
+#include <cstddef> // size_t
+#include <cstdint> // uint8_t
+
+namespace nlohmann
+{
+namespace detail
+{
+///////////////////////////
+// JSON type enumeration //
+///////////////////////////
+
+/*!
+@brief the JSON type enumeration
+
+This enumeration collects the different JSON types. It is internally used to
+distinguish the stored values, and the functions @ref basic_json::is_null(),
+@ref basic_json::is_object(), @ref basic_json::is_array(),
+@ref basic_json::is_string(), @ref basic_json::is_boolean(),
+@ref basic_json::is_number() (with @ref basic_json::is_number_integer(),
+@ref basic_json::is_number_unsigned(), and @ref basic_json::is_number_float()),
+@ref basic_json::is_discarded(), @ref basic_json::is_primitive(), and
+@ref basic_json::is_structured() rely on it.
+
+@note There are three enumeration entries (number_integer, number_unsigned, and
+number_float), because the library distinguishes these three types for numbers:
+@ref basic_json::number_unsigned_t is used for unsigned integers,
+@ref basic_json::number_integer_t is used for signed integers, and
+@ref basic_json::number_float_t is used for floating-point numbers or to
+approximate integers which do not fit in the limits of their respective type.
+
+@sa @ref basic_json::basic_json(const value_t value_type) -- create a JSON
+value with the default value for a given type
+
+@since version 1.0.0
+*/
+enum class value_t : std::uint8_t
+{
+    null,             ///< null value
+    object,           ///< object (unordered set of name/value pairs)
+    array,            ///< array (ordered collection of values)
+    string,           ///< string value
+    boolean,          ///< boolean value
+    number_integer,   ///< number value (signed integer)
+    number_unsigned,  ///< number value (unsigned integer)
+    number_float,     ///< number value (floating-point)
+    discarded         ///< discarded by the the parser callback function
+};
+
+/*!
+@brief comparison operator for JSON types
+
+Returns an ordering that is similar to Python:
+- order: null < boolean < number < object < array < string
+- furthermore, each type is not smaller than itself
+- discarded values are not comparable
+
+@since version 1.0.0
+*/
+inline bool operator<(const value_t lhs, const value_t rhs) noexcept
+{
+    static constexpr std::array<std::uint8_t, 8> order = {{
+            0 /* null */, 3 /* object */, 4 /* array */, 5 /* string */,
+            1 /* boolean */, 2 /* integer */, 2 /* unsigned */, 2 /* float */
+        }
+    };
+
+    const auto l_index = static_cast<std::size_t>(lhs);
+    const auto r_index = static_cast<std::size_t>(rhs);
+    return l_index < order.size() and r_index < order.size() and order[l_index] < order[r_index];
+}
+}
+}

src/lib/json/json_fwd.hpp

@@ -0,0 +1,64 @@
+#ifndef NLOHMANN_JSON_FWD_HPP
+#define NLOHMANN_JSON_FWD_HPP
+
+#include <cstdint> // int64_t, uint64_t
+#include <map> // map
+#include <memory> // allocator
+#include <string> // string
+#include <vector> // vector
+
+/*!
+@brief namespace for Niels Lohmann
+@see https://github.com/nlohmann
+@since version 1.0.0
+*/
+namespace nlohmann
+{
+/*!
+@brief default JSONSerializer template argument
+
+This serializer ignores the template arguments and uses ADL
+([argument-dependent lookup](http://en.cppreference.com/w/cpp/language/adl))
+for serialization.
+*/
+template<typename = void, typename = void>
+struct adl_serializer;
+
+template<template<typename U, typename V, typename... Args> class ObjectType =
+         std::map,
+         template<typename U, typename... Args> class ArrayType = std::vector,
+         class StringType = std::string, class BooleanType = bool,
+         class NumberIntegerType = std::int64_t,
+         class NumberUnsignedType = std::uint64_t,
+         class NumberFloatType = double,
+         template<typename U> class AllocatorType = std::allocator,
+         template<typename T, typename SFINAE = void> class JSONSerializer =
+         adl_serializer>
+class basic_json;
+
+/*!
+@brief JSON Pointer
+
+A JSON pointer defines a string syntax for identifying a specific value
+within a JSON document. It can be used with functions `at` and
+`operator[]`. Furthermore, JSON pointers are the base for JSON patches.
+
+@sa [RFC 6901](https://tools.ietf.org/html/rfc6901)
+
+@since version 2.0.0
+*/
+template<typename BasicJsonType>
+class json_pointer;
+
+/*!
+@brief default JSON class
+
+This type is the default specialization of the @ref basic_json class which
+uses the standard template types.
+
+@since version 1.0.0
+*/
+using json = basic_json<>;
+}
+
+#endif