___ _ __ _____ ___
(_ | \ ___) | \ ) ____)
| | | (__ | ) / / __
_ | | | __) | / ( ( ( \
( |_| | | (___ | |\ \ \ \__) )
_\ /__/ )_| |_\ \___) (__
jerg generates this:
% Monetary price
% Intended to be used embedded in other schemas
-record(price,
{
currency_code :: binary(),
amount :: number()
}).
% Product (Short View)
% Intended to be used in search results
-record(product_short,
{
% Numeric object identity, set to -1 at creation time
id = -1 :: integer(),
name :: binary(),
list_price :: #price{},
discount_price :: #price{}
}).
% Product Category
-record(product_category,
{
% Numeric object identity, set to -1 at creation time
id = -1 :: integer(),
name :: binary()
}).
% Product (Full View)
% Intended to be used when full description is needed
-record(product_full,
{
% Numeric object identity, set to -1 at creation time
id = -1 :: integer(),
name :: binary(),
list_price :: #price{},
discount_price :: #price{},
description :: binary(),
categories :: list(#product_category{})
}).
out of that:
| category.json |
|---|
{
"title": "Product Category",
"recordName": "product_category",
"type": "object",
"extends" : {
"$ref": "identified_object.json"
},
"properties": {
"name": {
"type": "string"
}
}
}
|
| identified_object.json |
{
"title": "Identity field for all objects",
"description" : "Intended to be extended by identified objects",
"type": "object",
"abstract" : "true",
"properties": {
"id": {
"description" : "Numeric object identity, set to -1 at creation time",
"type": "integer",
"default" : -1
}
}
}
|
| price.json |
{
"title": "Monetary price",
"description" : "Intended to be used embedded in other schemas",
"type": "object",
"properties": {
"currency_code": {
"type": "string"
},
"amount": {
"type": "number"
}
}
}
|
| product_full.json |
{
"title": "Product (Full View)",
"description" : "Intended to be used when full description is needed",
"type": "object",
"extends" : {
"$ref": "product_short.json"
},
"properties": {
"description": {
"type": "string"
},
"categories": {
"type" : "array",
"items" : {
"$ref": "category.json"
}
}
}
}
|
| product_short.json |
{
"title": "Product (Short View)",
"description" : "Intended to be used in search results",
"type": "object",
"extends" : {
"$ref": "identified_object.json"
},
"properties": {
"name": {
"type": "string"
},
"list_price": {
"$ref": "price.json"
},
"discount_price": {
"$ref": "price.json"
}
}
}
|
jerg supports:
- cross-references for properties and collection items (ie the
$refproperty), - default values,
- integer enumerations (other types can not be enumerated per limitation of the Erlang type specification).
jerg also supports a few extensions to JSON schema:
extends: to allow a schema to extend another one in order to inherit all the properties of its parent (and any ancestors above it),abstract: to mark schemas that actually do not need to be output as records because they are just used through references and extension,recordName: to customize the name of the record generated from the concerned schema definition.
jerg is built with rebar.
Run:
rebar get-deps compile
rebar escriptize skip_deps=true
$ jerg
Usage: jerg [-s <source>] [-t [<target>]] [-p <prefix>] [-v]
-s, --source Source directory or file.
-t, --target Generation target. Use - to output to stdout. [default: include/json_records.hrl]
-p, --prefix Records name prefix.
-v, --version Display version information and stop.
jerg is intended to be used as part of your build chain in order to produce Erlang records right before compilation starts.
If you use rebar, this is done by adding the following to your rebar.config:
{pre_hooks, [{compile, "jerg -s priv/json_schemas"}]}.
This assumes that your JSON schemas are located in priv/json_schemas and you're OK with outputting the generated records to include/json_records.hrl.
- Use json_rec to serialize and deserialize JSON to and from Erlang records.
- Use jesse to validate JSON data against JSON schema.
jerg doesn't support the following:
- Embedded object definitions: for now all objects must be defined as top level schema files,
- Hierarchical organization of schema files,
- URL and self references to schema files,
- Cyclic schema dependencies (to support it, a potential approach would be to introduce an
any()type in a dependency cycle in order to break it), - Additional properties.