Generate JSON Schema documents from Rust code
If you don't really care about the specifics, the easiest way to generate a JSON schema for your types is to #[derive(JsonSchema)]
and use the schema_for!
macro. All fields of the type must also implement JsonSchema
- Schemars implements this for many standard library types.
use schemars::{schema_for, JsonSchema};
#[derive(JsonSchema)]
pub struct MyStruct {
pub my_int: i32,
pub my_bool: bool,
pub my_nullable_enum: Option<MyEnum>,
}
#[derive(JsonSchema)]
pub enum MyEnum {
StringNewType(String),
StructVariant { floats: Vec<f32> },
}
let schema = schema_for!(MyStruct);
println!("{}", serde_json::to_string_pretty(&schema).unwrap());
Click to see the output JSON schema...
{
"$schema": "http://json-schema.org/draft-07/schema#",
"title": "MyStruct",
"type": "object",
"required": [
"my_bool",
"my_int"
],
"properties": {
"my_bool": {
"type": "boolean"
},
"my_int": {
"type": "integer",
"format": "int32"
},
"my_nullable_enum": {
"anyOf": [
{
"$ref": "#/definitions/MyEnum"
},
{
"type": "null"
}
]
}
},
"definitions": {
"MyEnum": {
"anyOf": [
{
"type": "object",
"required": [
"StringNewType"
],
"properties": {
"StringNewType": {
"type": "string"
}
},
"additionalProperties": false
},
{
"type": "object",
"required": [
"StructVariant"
],
"properties": {
"StructVariant": {
"type": "object",
"required": [
"floats"
],
"properties": {
"floats": {
"type": "array",
"items": {
"type": "number",
"format": "float"
}
}
}
}
},
"additionalProperties": false
}
]
}
}
}
One of the main aims of this library is compatibility with Serde. Any generated schema should match how serde_json would serialize/deserialize to/from JSON. To support this, Schemars will check for any #[serde(...)]
attributes on types that derive JsonSchema
, and adjust the generated schema accordingly.
use schemars::{schema_for, JsonSchema};
use serde::{Deserialize, Serialize};
#[derive(Deserialize, Serialize, JsonSchema)]
#[serde(rename_all = "camelCase", deny_unknown_fields)]
pub struct MyStruct {
#[serde(rename = "myNumber")]
pub my_int: i32,
pub my_bool: bool,
#[serde(default)]
pub my_nullable_enum: Option<MyEnum>,
}
#[derive(Deserialize, Serialize, JsonSchema)]
#[serde(untagged)]
pub enum MyEnum {
StringNewType(String),
StructVariant { floats: Vec<f32> },
}
let schema = schema_for!(MyStruct);
println!("{}", serde_json::to_string_pretty(&schema).unwrap());
Click to see the output JSON schema...
{
"$schema": "http://json-schema.org/draft-07/schema#",
"title": "MyStruct",
"type": "object",
"required": [
"myBool",
"myNumber"
],
"properties": {
"myBool": {
"type": "boolean"
},
"myNullableEnum": {
"default": null,
"anyOf": [
{
"$ref": "#/definitions/MyEnum"
},
{
"type": "null"
}
]
},
"myNumber": {
"type": "integer",
"format": "int32"
}
},
"additionalProperties": false,
"definitions": {
"MyEnum": {
"anyOf": [
{
"type": "string"
},
{
"type": "object",
"required": [
"floats"
],
"properties": {
"floats": {
"type": "array",
"items": {
"type": "number",
"format": "float"
}
}
}
}
]
}
}
}
#[serde(...)]
attributes can be overriden using #[schemars(...)]
attributes, which behave identically (e.g. #[schemars(rename_all = "camelCase")]
). You may find this useful if you want to change the generated schema without affecting Serde's behaviour, or if you're just not using Serde.
If you want a schema for a type that can't/doesn't implement JsonSchema
, but does implement serde::Serialize
, then you can generate a JSON schema from a value of that type. However, this schema will generally be less precise than if the type implemented JsonSchema
- particularly when it involves enums, since schemars will not make any assumptions about the structure of an enum based on a single variant.
use schemars::schema_for_value;
use serde::Serialize;
#[derive(Serialize)]
pub struct MyStruct {
pub my_int: i32,
pub my_bool: bool,
pub my_nullable_enum: Option<MyEnum>,
}
#[derive(Serialize)]
pub enum MyEnum {
StringNewType(String),
StructVariant { floats: Vec<f32> },
}
let schema = schema_for_value!(MyStruct {
my_int: 123,
my_bool: true,
my_nullable_enum: Some(MyEnum::StringNewType("foo".to_string()))
});
println!("{}", serde_json::to_string_pretty(&schema).unwrap());
Click to see the output JSON schema...
{
"$schema": "http://json-schema.org/draft-07/schema#",
"title": "MyStruct",
"examples": [
{
"my_bool": true,
"my_int": 123,
"my_nullable_enum": {
"StringNewType": "foo"
}
}
],
"type": "object",
"properties": {
"my_bool": {
"type": "boolean"
},
"my_int": {
"type": "integer"
},
"my_nullable_enum": true
}
}
derive
(enabled by default) - provides#[derive(JsonSchema)]
macroimpl_json_schema
- implementsJsonSchema
for Schemars types themselvespreserve_order
- keep the order of struct fields inSchema
andSchemaObject
Schemars can implement JsonSchema
on types from several popular crates, enabled via feature flags (dependency versions are shown in brackets):
chrono
- chrono (^0.4)indexmap1
- indexmap (^1.2)either
- either (^1.3)uuid08
- uuid (^0.8)uuid1
- uuid (^1.0)smallvec
- smallvec (^1.0)arrayvec05
- arrayvec (^0.5)arrayvec07
- arrayvec (^0.7)url
- url (^2.0)bytes
- bytes (^1.0)enumset
- enumset (^1.0)rust_decimal
- rust_decimal (^1.0)bigdecimal
- bigdecimal (^0.3)smol_str
- smol_str (^0.1.17)
For example, to implement JsonSchema
on types from chrono
, enable it as a feature in the schemars
dependency in your Cargo.toml
like so:
[dependencies]
schemars = { version = "0.8", features = ["chrono"] }