1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
|
# Custom Wrappers
Custom wrappers are used to change the serialization/deserialization on basic types (e.g. int, double, std::string).
The example below shows how numbers can be read in via strings by providing a wrapper to indicate the proper serialization/deserialization.
```c++
template <class T>
struct quoted_helper
{
T& val;
};
namespace glz
{
template <class T>
struct from<JSON, quoted_helper<T>>
{
template <auto Opts>
static void op(auto&& value, auto&& ctx, auto&& it, auto&& end)
{
skip_ws<Opts>(ctx, it, end);
if (match<'"'>(ctx, it)) {
return;
}
parse<JSON>::op<Opts>(value.val, ctx, it, end);
if (match<'"'>(ctx, it)) {
return;
}
}
};
template <class T>
struct to<JSON, quoted_helper<T>>
{
template <auto Opts>
static void op(auto&& value, is_context auto&& ctx, auto&& b, auto&& ix) noexcept
{
dump<'"'>(b, ix);
serialize<JSON>::op<Opts>(value.val, ctx, b, ix);
dump<'"'>(b, ix);
}
};
}
template <auto MemPtr>
constexpr decltype(auto) qouted()
{
return [](auto&& val) { return quoted_helper<std::decay_t<decltype(val.*MemPtr)>>{val.*MemPtr}; };
}
struct A
{
double x;
};
template <>
struct glz::meta<A>
{
static constexpr auto value = object("x", qouted<&A::x>());
// or...
//static constexpr auto value = object("x", [](auto&& val) { return quoted_helper(val.x); });
};
void example() {
A a{3.14};
std::string buffer{};
expect(not glz::write_json(a, buffer));
expect(buffer == R"({"x":"3.14"})");
buffer = R"({"x":"999.2"})";
expect(not glz::read_json(a, buffer));
expect(a.x == 999.2);
}
```
## Custom Type Parsing
Custom wrappers allow you to define specialized serialization/deserialization behavior for specific types. This is useful when working with formats that differ from standard JSON representations.
### Example: Python-style Booleans
Some systems (like Python) represent booleans as `True`/`False` instead of JSON's standard `true`/`false`. Here's how to create a custom wrapper to handle this format:
#### Step 1: Define the Wrapper Helper
```c++
template <class T>
struct python_bool_helper
{
T& val; // Reference to the actual boolean value
};
```
#### Step 2: Specialize the Parser (`from`)
```c++
namespace glz
{
template <class T>
struct from<JSON, python_bool_helper<T>>
{
template <auto Opts>
static void op(auto&& value, auto&& ctx, auto&& it, auto&& end)
{
skip_ws<Opts>(ctx, it, end);
if (it >= end) {
ctx.error = error_code::unexpected_end;
return;
}
// Check for 'T' (True) or 'F' (False)
if (*it == 'T') {
if (it + 4 <= end && std::string_view(it, 4) == "True") {
value.val = true;
it += 4;
} else {
ctx.error = error_code::expected_true_or_false;
}
} else if (*it == 'F') {
if (it + 5 <= end && std::string_view(it, 5) == "False") {
value.val = false;
it += 5;
} else {
ctx.error = error_code::expected_true_or_false;
}
} else {
ctx.error = error_code::expected_true_or_false;
}
}
};
```
#### Step 3: Specialize the Serializer (`to`)
```c++
template <class T>
struct to<JSON, python_bool_helper<T>>
{
template <auto Opts, class B>
static void op(auto&& value, is_context auto&&, B&& b, auto&& ix) noexcept
{
if (value.val) {
std::memcpy(&b[ix], "True", 4);
ix += 4;
} else {
std::memcpy(&b[ix], "False", 5);
ix += 5;
}
}
};
}
```
#### Step 4: Create a Convenience Function
```c++
template <auto MemPtr>
constexpr decltype(auto) python_bool()
{
return [](auto&& val) {
return python_bool_helper<std::decay_t<decltype(val.*MemPtr)>>{val.*MemPtr};
};
}
```
#### Step 5: Use in Your Structs
```c++
struct ConfigData
{
std::string name;
int port;
bool debug_mode;
bool enable_logging;
};
template <>
struct glz::meta<ConfigData>
{
using T = ConfigData;
static constexpr auto value = object(
&T::name,
&T::port,
"debug_mode", python_bool<&T::debug_mode>(), // Custom wrapper for this field
"enable_logging", python_bool<&T::enable_logging>() // Custom wrapper for this field
);
};
```
#### Usage Example
```c++
ConfigData config{.name = "MyApp", .port = 8080, .debug_mode = true, .enable_logging = false};
// Writing produces: {"name":"MyApp","port":8080,"debug_mode":True,"enable_logging":False}
std::string json_output;
glz::write_json(config, json_output);
// Reading accepts Python-style booleans
std::string python_json = R"({"name":"TestApp","port":3000,"debug_mode":False,"enable_logging":True})";
ConfigData parsed_config;
glz::read_json(parsed_config, python_json);
```
### Alternative Approaches
1. **Custom Type Wrapper**: Create your own boolean type with `glz::to`/`glz::from` specializations, allowing pure reflection usage
2. **Global Override**: Override boolean parsing globally (affects all boolean fields)
The custom wrapper approach shown above provides the most flexibility and performance, allowing field-by-field control over serialization behavior.
|
