Valico is a validation and coercion tool for JSON objects, written in Rust. It designed to be a support library for the various REST-like frameworks or other tools that need to validate and coerce JSON input from outside world.
Valico has two features:
- DSL — a set of simple validators and coercers inspired by [Grape]. It has built-in support for common coercers, validators and can return detailed error messages if something goes wrong.
- JSON Schema — An implementation of JSON Schema, based on IETF's draft v7.
References:
- http://json-schema.org
- http://json-schema.org/latest/json-schema-core.html
- http://json-schema.org/latest/json-schema-validation.html
# Cargo.toml
valico = "2"It passes the entire JSON-Schema-Test-Suite except for remoteRefs and maxLength/minLength when using unicode surrogate pairs. It also can validate your schema and give you an explanation about what is wrong in it.
extern crate serde_json;
extern crate valico;
use serde_json::Value;
use valico::json_schema;
use std::fs::File;
fn main() {
let json_schema: Value = serde_json::from_reader(File::open("tests/schema/schema.json").unwrap()).unwrap();
let mut scope = json_schema::Scope::new();
let schema = scope.compile_and_return(json_v4_schema.clone(), false).unwrap();
println!("Is valid: {}", schema.validate(&json_v4_schema).is_valid());
}Valico goes with valico::json_schema::schema(|scheme| { /* .. */ }) -> json::Json function that allows to use simple DSL to generate your schemes. It allows you not to use strings and raw JSON manipulation. It also prevent some kinds of spelling and type errors.
builder::schema(|s| {
s.properties(|properties| {
properties.insert("prop1", |prop1| {
prop1.maximum(10f64, false);
});
});
s.pattern_properties(|properties| {
properties.insert("prop.*", |prop| {
prop.maximum(1000f64, false);
});
});
s.additional_properties_schema(|additional| {
additional.maximum(5f64, false)
});
})TODO more docs about JSON Schema here
All Valico stuff is making by Builder instance. Below is a simple example showing how one can create and setup Builder:
let params = Builder::build(|params| {
params.req_nested("user", Builder::list(), |params| {
params.req_typed("name", json_dsl::string());
params.req_typed("friend_ids", json_dsl::array_of(json_dsl::u64()))
});
});Later params instance can be used to process one or more JSON objects with it's process method with signature fn process(&self, tree: &mut JsonObject) -> ValicoResult<()>.
Note that Valico will mutate borrowed JSON value if some coercion is needed.
Example:
extern crate valico;
extern crate serde_json;
use valico::json_dsl;
use serde_json::{from_str, to_string_pretty};
fn main() {
let params = json_dsl::Builder::build(|params| {
params.req_nested("user", json_dsl::array(), |params| {
params.req_typed("name", json_dsl::string());
params.req_typed("friend_ids", json_dsl::array_of(json_dsl::u64()))
});
});
let mut obj = from_str(r#"{"user": {"name": "Frodo", "friend_ids": ["1223"]}}"#).unwrap();
let state = params.process(&mut obj, &None);
if state.is_valid() {
println!("Result object is {}", to_string_pretty(&obj).unwrap());
} else {
panic!("Errors during process: {:?}", state);
}
}Also you can look to the specs for more details and examples.
You can define validations and coercion options for your parameters using a Builder::build block. Parameters can be optional and required. Requires parameters must be always present. Optional parameters can be omitted.
When parameter is present in JSON all validation and coercions will be applied and error fired if something goes wrong.
This functions are available in Builder to define parameters:
// Parameter is required, no coercion
fn req_defined(&mut self, name: &str);
// Parameter is required, with coercion
fn req_typed(&mut self, name: &str, coercer: Box<Coercer>);
// Parameter is required, with coercion and nested checks
fn req_nested(&mut self, name: &str, coercer: Box<Coercer>, nest_def: |&mut Builder|);
// Parameter is required, setup with Param DSL
fn req(&mut self, name: &str, param_builder: |&mut Param|);
// Parameter is optional, no coercion
fn opt_defined(&mut self, name: &str);
// Parameter is optional, with coercion
fn opt_typed(&mut self, name: &str, coercer: Box<Coercer>);
// Parameter is optional, with coercion and nested checks
fn opt_nested(&mut self, name: &str, coercer: Box<Coercer>, nest_def: |&mut Builder|);
// Parameter is required, setup with Param DSL
fn opt(&mut self, name: &str, param_builder: |&mut Param|);Available list of coercers:
- json_dsl::i64()
- json_dsl::u64()
- json_dsl::f64()
- json_dsl::string()
- json_dsl::boolean()
- json_dsl::null()
- json_dsl::array()
- json_dsl::array_of()
- json_dsl::encoded_array() — use it for string-encoded arrays e.g. "red,green,blue" -> ["red", "green", "blue"]
- json_dsl::encoded_array_of() — use it for string-encoded arrays of some type e.g. "1,2,3" -> [1, 2, 3]
- json_dsl::object()
Example of usage:
let params = Builder::build(|params| {
params.req_typed("id", json_dsl::u64());
params.req_typed("name", json_dsl::string());
params.opt_typed("is_active", json_dsl::boolean());
params.opt_typed("tags", json_dsl::array_of(json_dsl::strings()));
});You can specify rules to nesting processing for lists and objects:
let params = Builder::build(|params| {
params.req_nested("user", json_dsl::object(), |params| {
params.req_typed("name", json_dsl::string());
params.opt_typed("is_active", json_dsl::boolean());
params.opt_typed("tags", json_dsl::array_of(json_dsl::strings()));
});
});
let params = Builder::build(|params| {
params.req_nested("users", Builder::list(), |params| {
params.req_typed("name", json_dsl::string());
params.opt_typed("is_active", json_dsl::boolean());
params.opt_typed("tags", json_dsl::array_of(json_dsl::strings()));
});
});Nesting level is not limited in Valico.
DSL allows to use JSON Schema validations to validate objects at the Builder level and the Param level:
let params = json_dsl::Builder::build(|params| {
params.req("a", |a| {
a.schema(|schema| {
schema.integer();
schema.maximum(10f64, false);
})
});
});Note that JSON Schema validates object AFTER coerce pass:
let mut params = json_dsl::Builder::build(|params| {
params.req("a", |a| {
a.coerce(json_dsl::u64());
a.schema(|schema| {
schema.maximum(10f64, false);
})
});
});Don't forget to create a json_schema::Scope BEFORE processing:
let mut scope = json_schema::Scope::new();
params.build_schemes(&mut scope).unwrap();You can use DSL block to setup parameters with more flexible way:
let params = Builder::build(|params| {
params.req("user", |user| {
user.desc("Parameter is used to create new user");
user.coerce(json_dsl::object());
// this allows null to be a valid value
user.allow_null();
user.nest(|params| {
params.req_typed("name", json_dsl::string());
params.opt("kind", |kind| {
kind.coerce(json_dsl::string());
// optional parameters can have default values
kind.default("simeple_user".to_string())
});
});
});
});DSL supports several parameter validations. They considered outdated and likely to be removed in the future in favour of JSON Schema validation.
Parameters can be restricted to a specific set of values with allow_values:
let params = Builder::build(|params| {
params.req("kind", |kind| {
kind.coerce(json_dsl::string());
kind.allow_values(&["circle".to_string(), "square".to_string()]);
})
})Some values can be rejected with reject_values:
let params = Builder::build(|params| {
params.req("user_role", |kind| {
kind.coerce(json_dsl::string());
kind.reject_values(&["admin".to_string(), "manager".to_string()]);
})
})String values can be tested with Regex:
let params = Builder::build(|params| {
params.req("nickname", |a| {
a.coerce(json_dsl::string());
// force all nicknames to start with "Amazing"
a.regex(regex!("^Amazing"));
})
});Sometimes it's usefull to use some custom function as validator:
let params = Builder::build(|params| {
params.req("pushkin_birthday", |a| {
a.coerce(json_dsl::u64());
fn guess(val: &Json) -> Result<(), String> {
if *val == 1799u.to_json() {
Ok(())
} else {
Err("No!".to_string())
}
}
a.validate_with(guess);
});
});One can use custom validator. Docs in Progress.
Some validators can be specified in Builder DSL block to validate a set of parameters.
Parameters can be defined as mutually_exclusive, ensuring that they aren't present at the same time in a request.
let params = Builder::build(|params| {
params.opt_defined("vodka");
params.opt_defined("beer");
params.mutually_exclusive(&["vodka", "beer"]);
});Multiple sets can be defined:
let params = Builder::build(|params| {
params.opt_defined("vodka");
params.opt_defined("beer");
params.mutually_exclusive(&["vodka", "beer"]);
params.opt_defined("lard");
params.opt_defined("jamon");
params.mutually_exclusive(&["lard", "jamon"]);
});Warning: Never define mutually exclusive sets with any required params. Two mutually exclusive required params will mean params are never valid. One required param mutually exclusive with an optional param will mean the latter is never valid.
Parameters can be defined as 'exactly_one_of', ensuring that exactly one parameter gets selected.
let params = Builder::build(|params| {
params.opt_defined("vodka");
params.opt_defined("beer");
params.exactly_one_of(["vodka", "beer"]);
});Parameters can be defined as 'at_least_one_of', ensuring that at least one parameter gets selected.
let params = Builder::build(|params| {
params.opt_defined("vodka");
params.opt_defined("beer");
params.opt_defined("wine");
params.exactly_one_of(["vodka", "beer", "wine"]);
});Sometimes it's usefull to use some custom function as validator:
let params = Builder::build(|params| {
params.req_defined("monster_name");
fn validate_params(_: &JsonObject) -> Result<(),String> {
Err("YOU SHALL NOT PASS".to_string())
}
params.validate_with(validate_params);
});One can use custom validator. Docs in Progress.
For WebAssembly, enable the js feature:
# Cargo.toml
valico = { version = "2", features = ["js"] }