JSON-API is a specification for building REST APIs for CRUD (create, read, update, and delete) operations.
Similar to GraphQL:
Unlike GraphQL, the JSON-API specification spells out exactly how to perform common CRUD operations including complex graph mutations.
JSON-API has no standardized schema introspection. However, Elide adds this capability to any service by exporting
an Open API Initiative document (formerly known as Swagger).
The json-api specification is the best reference for understanding JSON-API. The following sections describe commonly used JSON-API features as well as Elide additions for filtering, pagination, sorting, and swagger.
Elide generally follows the JSON-API recommendations for URL design.
There are a few caveats given that Elide allows developers control over how entities are exposed:
/articles/1/author/address./articles/1/author/34/addressElide supports three mechanisms by which a newly created entity is assigned an ID:
Elide looks for the JPA GeneratedValue annotation to disambiguate whether or not
the data store generates an ID for a given data model. If the client also generated
an ID during the object creation request, the data store ID overrides the client value.
When using the patch extension, Elide returns object entity bodies (containing newly assigned IDs) in the order in which they were created. The client can use this order to map the object created to its server assigned ID.
JSON-API allows the client to limit the attributes and relationships that should be included in the response payload for any given entity. The fields query parameter specifies the type (data model) and list of fields that should be included.
For example, to fetch the book collection but only include the book titles:
/book?fields[book]=title
{
"data": [
{
"attributes": {
"title": "The Old Man and the Sea"
},
"id": "1",
"type": "book"
},
{
"attributes": {
"title": "For Whom the Bell Tolls"
},
"id": "2",
"type": "book"
},
{
"attributes": {
"title": "Enders Game"
},
"id": "3",
"type": "book"
}
]
}
More information about sparse fields can be found here.
JSON-API allows the client to fetch a primary collection of elements but also include their relationships or their relationship’s relationships (arbitrarily nested) through compound documents. The include query parameter specifies what relationships should be expanded in the document.
The following example fetches the book collection but also includes all of the book authors. Sparse fields are used to limit the book and author fields in the response:
/book?include=authors&fields[book]=title,authors&fields[author]=name
{
"data": [
{
"attributes": {
"title": "The Old Man and the Sea"
},
"id": "1",
"relationships": {
"authors": {
"data": [
{
"id": "1",
"type": "author"
}
]
}
},
"type": "book"
},
{
"attributes": {
"title": "For Whom the Bell Tolls"
},
"id": "2",
"relationships": {
"authors": {
"data": [
{
"id": "1",
"type": "author"
}
]
}
},
"type": "book"
},
{
"attributes": {
"title": "Enders Game"
},
"id": "3",
"relationships": {
"authors": {
"data": [
{
"id": "2",
"type": "author"
}
]
}
},
"type": "book"
}
],
"included": [
{
"attributes": {
"name": "Ernest Hemingway"
},
"id": "1",
"type": "author"
},
{
"attributes": {
"name": "Orson Scott Card"
},
"id": "2",
"type": "author"
}
]
}
More information about compound documents can be found here.
JSON-API 1.0 is agnostic to filtering strategies. The only recommendation is that servers and clients should prefix filtering query parameters with the word ‘filter’.
Elide supports multiple filter dialects and the ability to add new ones to meet the needs of developers or to evolve the platform should JSON-API standardize them. Elide’s primary dialect is RSQL.
RSQL is a query language that allows conjunction (and), disjunction (or), and parenthetic grouping of Boolean expressions. It is a superset of the FIQL language.
Because RSQL is a superset of FIQL, FIQL queries should be properly parsed. RSQL primarily adds more friendly lexer tokens to FIQL for conjunction and disjunction: ‘and’ instead of ‘;’ and ‘or’ instead of ‘,’. RSQL also adds a richer set of operators. FIQL defines all String comparison operators to be case insensitive. Elide overrides this behavior making all operators case sensitive by default. For case insensitive queries, Elide introduces new operators.
Filter query parameters either look like:
filter[TYPE] where ‘TYPE’ is the name of the data model/entity. These are type specific filters and only apply to filtering collections of the given type.filter with no type specified. This is a global filter and can be used to filter across relationships (by performing joins in the persistence layer).Any number of typed filter parameters can be specified provided the ‘TYPE’ is different for each parameter. There can only be a single global filter for the entire query. Typed filters can be used for any collection returned by Elide. Global filters can only be used to filter root level collections.
The value of any query parameter is a RSQL expression composed of predicates. Each predicate contains an attribute of the data model or a related data model, an operator, and zero or more comparison values.
Filter attributes can be:
To join across relationships or drill into nested objects, the attribute name is prefixed by one or more relationship or field names separated by period (‘.’). For example, ‘author.books.price.total’ references all of the author’s books with a price having a particular total value.
Return all the books written by author ‘1’ with the genre exactly equal to ‘Science Fiction’:
/author/1/book?filter[book]=genre=='Science Fiction'
Return all the books written by author ‘1’ with the genre exactly equal to ‘Science Fiction’ and the title starts with ‘The’ and whose total price is greater than 100.00:
/author/1/book?filter[book]=genre=='Science Fiction';title==The*;price.total>100.00
Return all the books written by author ‘1’ with the publication date greater than a certain time or the genre not ‘Literary Fiction’ or ‘Science Fiction’:
/author/1/book?filter[book]=publishDate>1454638927411,genre=out=('Literary Fiction','Science Fiction')
Return all the books whose title contains ‘Foo’. Include all the authors of those books whose name does not equal ‘Orson Scott Card’:
/book?include=authors&filter[book]=title==*Foo*&filter[author]=name!='Orson Scott Card'
Return all the books with an author whose name is ‘Null Ned’ and whose title is ‘Life with Null Ned’:
/book?filter=authors.name=='Null Ned';title=='Life with Null Ned'
The following RSQL operators are supported:
=in= : Evaluates to true if the attribute exactly matches any of the values in the list. (Case Sensitive)=ini=: Evaluates to true if the attribute exactly matches any of the values in the list. (Case Insensitive)=out= : Evaluates to true if the attribute does not match any of the values in the list. (Case Sensitive)=outi= : Evaluates to true if the attribute does not match any of the values in the list. (Case Insensitive)==ABC* : Similar to SQL like 'ABC%. (Case Sensitive)==*ABC : Similar to SQL like '%ABC. (Case Sensitive)==*ABC* : Similar to SQL like '%ABC%. (Case Sensitive)=ini=ABC* : Similar to SQL like 'ABC%. (Case Insensitive)=ini=*ABC : Similar to SQL like '%ABC. (Case Insensitive)=ini=*ABC* : Similar to SQL like '%ABC%. (Case Insensitive)=isnull=true : Evaluates to true if the attribute is null=isnull=false : Evaluates to true if the attribute is not null=lt= : Evaluates to true if the attribute is less than the value.=gt= : Evaluates to true if the attribute is greater than the value.=le= : Evaluates to true if the attribute is less than or equal to the value.=ge= : Evaluates to true if the attribute is greater than or equal to the value.=isempty= : Determines if a collection is empty or not.=between= : Determines if a model attribute is >= and <= the two provided arguments.=notbetween= : Negates the between operator.=hasmember= : Determines if a collection contains a particular element.=hasnomember= : Determines if a collection does not contain a particular element.The operators ‘hasmember’ and ‘hasnomember’ can be applied to collections (book.awards) or across to-many relationships (book.authors.name).
By default, the FIQL operators =in=,=out=,== are case sensitive. This can be reverted to case insensitive by changing the case sensitive strategy:
@Bean
@ConditionalOnMissingBean
public Elide initializeElide(EntityDictionary dictionary,
DataStore dataStore, ElideConfigProperties settings) {
ElideSettingsBuilder builder = new ElideSettingsBuilder(dataStore)
.withEntityDictionary(dictionary)
.withDefaultMaxPageSize(settings.getMaxPageSize())
.withDefaultPageSize(settings.getPageSize())
.withJoinFilterDialect(new RSQLFilterDialect(dictionary), new CaseSensitivityStrategy.FIQLCompliant())
.withSubqueryFilterDialect(new RSQLFilterDialect(dictionary), new CaseSensitivityStrategy.FIQLCompliant())
.withAuditLogger(new Slf4jLogger())
.withISO8601Dates("yyyy-MM-dd'T'HH:mm'Z'", TimeZone.getTimeZone("UTC"));
return new Elide(builder.build());
}
Values are specified as URL encoded strings. Elide will type coerce them into the appropriate primitive data type for the attribute filter.
Some data stores like the Aggregation Store support parameterized model attributes. Parameters can be included in a filter predicate with the following syntax:
field[arg1:value1][arg2:value2]
Argument values must be URL encoded. There is no limit to the number of arguments provided in this manner.
Elide supports:
Elide allows pagination of the primary collection being returned in the response via the page query parameter.
The rough BNF syntax for the page query parameter is:
<QUERY> ::=
"page" "[" "size" "]" "=" <INTEGER>
| "page" "[" "number" "]" "=" <INTEGER>
| "page" "[" "limit" "]" "=" <INTEGER>
| "page" "[" "offset" "]" "=" <INTEGER>
| "page" "[" "totals" "]"
Legal combinations of the page query params include:
Whenever a page query parameter is specified, Elide will return a meta block in the JSON-API response that contains:
The values for totalPages and totalRecords are only returned if the page[totals] parameter was specified in the query.
Paginate the book collection starting at the 4th record. Include no more than 2 books per page. Include the total size of the collection in the meta block:
/book?page[offset]=3&page[limit]=2&page[totals]
{
"data": [
{
"attributes": {
"chapterCount": 0,
"editorName": null,
"genre": "Science Fiction",
"language": "English",
"publishDate": 1464638927412,
"title": "Enders Shadow"
},
"id": "4",
"relationships": {
"authors": {
"data": [
{
"id": "2",
"type": "author"
}
]
},
"chapters": {
"data": []
},
"publisher": {
"data": null
}
},
"type": "book"
},
{
"attributes": {
"chapterCount": 0,
"editorName": null,
"genre": "Science Fiction",
"language": "English",
"publishDate": 0,
"title": "Foundation"
},
"id": "5",
"relationships": {
"authors": {
"data": [
{
"id": "3",
"type": "author"
}
]
},
"chapters": {
"data": []
},
"publisher": {
"data": null
}
},
"type": "book"
}
],
"meta": {
"page": {
"limit": 2,
"number": 2,
"totalPages": 4,
"totalRecords": 8
}
}
}
Elide supports:
Elide allows sorting of the primary collection being returned in the response via the sort query parameter.
The rough BNF syntax for the sort query parameter is:
<QUERY> ::= "sort" "=" <LIST_OF_SORT_SPECS>
<LIST_OF_SORT_SPECS> = <SORT_SPEC> | <SORT_SPEC> "," <LIST_OF_SORT_SPECS>
<SORT_SPEC> ::= "+|-"? <PATH_TO_ATTRIBUTE>
<PATH_TO_ATTRIBUTE> ::= <RELATIONSHIP> <PATH_TO_ATTRIBUTE> | <ATTRIBUTE>
<RELATIONSHIP> ::= <TERM> "."
<ATTRIBUTE> ::= <TERM>
The keyword id can be used to sort by whatever field a given entity uses as its identifier.
Sort the collection of author 1’s books in descending order by the book’s publisher’s name:
/author/1/books?sort=-publisher.name
{
"data": [
{
"attributes": {
"chapterCount": 0,
"editorName": null,
"genre": "Literary Fiction",
"language": "English",
"publishDate": 0,
"title": "For Whom the Bell Tolls"
},
"id": "2",
"relationships": {
"authors": {
"data": [
{
"id": "1",
"type": "author"
}
]
},
"chapters": {
"data": []
},
"publisher": {
"data": {
"id": "2",
"type": "publisher"
}
}
},
"type": "book"
},
{
"attributes": {
"chapterCount": 0,
"editorName": null,
"genre": "Literary Fiction",
"language": "English",
"publishDate": 0,
"title": "The Old Man and the Sea"
},
"id": "1",
"relationships": {
"authors": {
"data": [
{
"id": "1",
"type": "author"
}
]
},
"chapters": {
"data": []
},
"publisher": {
"data": {
"id": "1",
"type": "publisher"
}
}
},
"type": "book"
}
]
}
JSON-API supported a now-deprecated mechanism for extensions.
The patch extension was a
JSON-API extension that allowed muliple mutation operations (create, delete, update) to be bundled together in as single request.
Elide supports the JSON-API patch extension because it allows complex & bulk edits to the data model in the context of a single transaction. For example, the following request creates an author (earnest hemingway), multiple of his books, and his book publisher in a single request:
[
{
"op": "add",
"path": "/author",
"value": {
"id": "12345678-1234-1234-1234-1234567890ab",
"type": "author",
"attributes": {
"name": "Ernest Hemingway"
},
"relationships": {
"books": {
"data": [
{
"type": "book",
"id": "12345678-1234-1234-1234-1234567890ac"
},
{
"type": "book",
"id": "12345678-1234-1234-1234-1234567890ad"
}
]
}
}
}
},
{
"op": "add",
"path": "/book",
"value": {
"type": "book",
"id": "12345678-1234-1234-1234-1234567890ac",
"attributes": {
"title": "The Old Man and the Sea",
"genre": "Literary Fiction",
"language": "English"
},
"relationships": {
"publisher": {
"data": {
"type": "publisher",
"id": "12345678-1234-1234-1234-1234567890ae"
}
}
}
}
},
{
"op": "add",
"path": "/book",
"value": {
"type": "book",
"id": "12345678-1234-1234-1234-1234567890ad",
"attributes": {
"title": "For Whom the Bell Tolls",
"genre": "Literary Fiction",
"language": "English"
}
}
},
{
"op": "add",
"path": "/book/12345678-1234-1234-1234-1234567890ac/publisher",
"value": {
"type": "publisher",
"id": "12345678-1234-1234-1234-1234567890ae",
"attributes": {
"name": "Default publisher"
}
}
}
]
[
{
"data": {
"attributes": {
"name": "Ernest Hemingway"
},
"id": "1",
"relationships": {
"books": {
"data": [
{
"id": "1",
"type": "book"
},
{
"id": "2",
"type": "book"
}
]
}
},
"type": "author"
}
},
{
"data": {
"attributes": {
"chapterCount": 0,
"editorName": null,
"genre": "Literary Fiction",
"language": "English",
"publishDate": 0,
"title": "The Old Man and the Sea"
},
"id": "1",
"relationships": {
"authors": {
"data": [
{
"id": "1",
"type": "author"
}
]
},
"chapters": {
"data": []
},
"publisher": {
"data": {
"id": "1",
"type": "publisher"
}
}
},
"type": "book"
}
},
{
"data": {
"attributes": {
"chapterCount": 0,
"editorName": null,
"genre": "Literary Fiction",
"language": "English",
"publishDate": 0,
"title": "For Whom the Bell Tolls"
},
"id": "2",
"relationships": {
"authors": {
"data": [
{
"id": "1",
"type": "author"
}
]
},
"chapters": {
"data": []
},
"publisher": {
"data": null
}
},
"type": "book"
}
},
{
"data": {
"attributes": {
"name": "Default publisher"
},
"id": "1",
"type": "publisher"
}
}
]
Important Caveats:
application/vnd.api+json; ext=jsonpatchJSON-API links are disabled by default. They can be enabled in application.yaml:
elide:
baseUrl: 'https://elide.io'
json-api:
path: /json
enabled: true
enableLinks: true
The enableLinks property switches the feature on. The baseUrl property provides the URL schema, host, and port your clients use to connect to your service. The path property provides the route where the JSON-API controller is rooted. All link URLs using the above configuration would be prefixed with ‘https://elide.io/json’.
If baseUrl is not provided, Elide will generate the link URL prefix using the client HTTP request.
For Elide standalone, you can enable links by overriding ElideStandaloneSettings and configure the settings:
@Override
public String getBaseUrl() {
return "https://elide.io";
}
@Override
public ElideSettings getElideSettings(final ServiceLocator injector) {
... //Removed other initialization for clarity
String jsonApiBaseUrl = getBaseUrl() + getJsonApiPathSpec().replaceAll("/\\*", "") + "/";
ElideSettingsBuilder builder = (new ElideSettingsBuilder(dataStore))
... // Removed other settings for clarity
.withJSONApiLinks(new DefaultJSONApiLinks(jsonApiBaseUrl));
return builder.build();
}
Enabling JSON-API links will result in payload responses that look like:
{
"data": [
{
"type": "group",
"id": "com.example.repository",
"attributes": {
"commonName": "Example Repository",
"description": "The code for this project"
},
"relationships": {
"products": {
"links": {
"self": "https://elide.io/api/v1/group/com.example.repository/relationships/products",
"related": "https://elide.io/api/v1/group/com.example.repository/products"
},
"data": [
]
}
},
"links": {
"self": "https://elide.io/api/v1/group/com.example.repository"
}
}
]
}
You can customize the links that are returned by registering your own implementation of JsonApiLinks with ElideSettings:
public interface JSONApiLinks {
Map<String, String> getResourceLevelLinks(PersistentResource var1);
Map<String, String> getRelationshipLinks(PersistentResource var1, String var2);
}
JSON-API supports returning non standard information in responses inside a meta block. Elide supports meta blocks in three scenarios:
To customize the document meta block, add fields to the RequestScope object inside a custom data store:
@Override
public <T> DataStoreIterable<T> loadObjects(EntityProjection projection, RequestScope scope) {
//Populates the JSON-API meta block with a new field, 'key':
scope.setMetadataField("key", 123);
This would produce a JSON response like:
{
"data": [
{
"type": "widget",
"id": "1"
}
],
"meta": {
"key": 123
}
}
To customize the resource meta block, the resource model class must implement the WithMetadata interface:
public interface WithMetadata {
/**
* Sets a metadata property for this request.
* @param property
* @param value
*/
default void setMetadataField(String property, Object value) { //NOOP }
/**
* Retrieves a metadata property from this request.
* @param property
* @return An optional metadata property.
*/
Optional<Object> getMetadataField(String property);
/**
* Return the set of metadata fields that have been set.
* @return metadata fields that have been set.
*/
Set<String> getMetadataFields();
}
For example, the following example model implements WithMetadata:
@Include
public class Widget implements WithMetadata {
static Map metadata = Map.of("key", 123);
@Id
private String id;
@Override
public Optional<Object> getMetadataField(String property) {
return Optional.ofNullable(Widget.metadata.get(property));
}
@Override
public Set<String> getMetadataFields() {
return Widget.metadata.keySet();
}
}
The models must be populated with at least one field for the meta block to be returned in the response. These fields can also be populated in a custom data store or lifecycle hook. This would produce a JSON response like:
{
"data": [
{
"type": "widget",
"id": "1",
"meta": {
"key": 123
}
}
]
}
Type coercion between the API and underlying data model has common support across JSON-API and GraphQL and is covered here.
Swagger documents can be highly customized. As a result, they are not enabled by default and instead must be initialized through code. The steps to do this are documented here.
Configuring custom error responses is documented here.