Semantic Annotations for Linked Avro Data (SALAD) §
Author:
- Peter Amstutz, Curoverse (now peter.amstutz@curii.com)
Contributors:
- The developers of Apache Avro
- The developers of JSON-LD
- Nebojša Tijanić nebojsa.tijanic@sbgenomics.com, Seven Bridges Genomics
Abstract §
Salad is a schema language for describing structured linked data documents in JSON or YAML documents. A Salad schema provides rules for preprocessing, structural validation, and link checking for documents described by a Salad schema. Salad builds on JSON-LD and the Apache Avro data serialization system, and extends Avro with features for rich data modeling such as inheritance, template specialization, object identifiers, and object references. Salad was developed to provide a bridge between the record oriented data modeling supported by Apache Avro and the Semantic Web.
Status of This Document §
This document is the product of the Common Workflow Language working group. The latest version of this document is available in the "schema_salad" directory at
https://github.com/common-workflow-language/schema_salad
The products of the CWL working group (including this document) are made available under the terms of the Apache License, version 2.0.
Table of contents
1. Introduction §
The JSON data model is an extremely popular way to represent structured data. It is attractive because of it's relative simplicity and is a natural fit with the standard types of many programming languages. However, this simplicity means that basic JSON lacks expressive features useful for working with complex data structures and document formats, such as schemas, object references, and namespaces.
JSON-LD is a W3C standard providing a way to describe how to interpret a JSON document as Linked Data by means of a "context". JSON-LD provides a powerful solution for representing object references and namespaces in JSON based on standard web URIs, but is not itself a schema language. Without a schema providing a well defined structure, it is difficult to process an arbitrary JSON-LD document as idiomatic JSON because there are many ways to express the same data that are logically equivalent but structurally distinct.
Several schema languages exist for describing and validating JSON data, such as the Apache Avro data serialization system, however none understand linked data. As a result, to fully take advantage of JSON-LD to build the next generation of linked data applications, one must maintain separate JSON schema, JSON-LD context, RDF schema, and human documentation, despite significant overlap of content and obvious need for these documents to stay synchronized.
Schema Salad is designed to address this gap. It provides a schema language and processing rules for describing structured JSON content permitting URI resolution and strict document validation. The schema language supports linked data through annotations that describe the linked data interpretation of the content, enables generation of JSON-LD context and RDF schema, and production of RDF triples by applying the JSON-LD context. The schema language also provides for robust support of inline documentation.
1.1 Introduction to draft 1 §
This is the first version of Schema Salad. It is developed concurrently with draft 3 of the Common Workflow Language for use in specifying the Common Workflow Language, however Schema Salad is intended to be useful to a broader audience.
1.2 References to Other Specifications §
Javascript Object Notation (JSON): http://json.org
JSON Linked Data (JSON-LD): http://json-ld.org
YAML: http://yaml.org
Avro: https://avro.apache.org/docs/current/spec.html
Uniform Resource Identifier (URI) Generic Syntax: https://tools.ietf.org/html/rfc3986)
Resource Description Framework (RDF): http://www.w3.org/RDF/
UTF-8: https://www.ietf.org/rfc/rfc2279.txt)
1.3 Scope §
This document describes the syntax, data model, algorithms, and schema language for working with Salad documents. It is not intended to document a specific implementation of Salad, however it may serve as a reference for the behavior of conforming implementations.
1.4 Terminology §
The terminology used to describe Salad documents is defined in the Concepts section of the specification. The terms defined in the following list are used in building those definitions and in describing the actions of an Salad implementation:
may: Conforming Salad documents and Salad implementations are permitted but not required to be interpreted as described.
must: Conforming Salad documents and Salad implementations are required to be interpreted as described; otherwise they are in error.
error: A violation of the rules of this specification; results are undefined. Conforming implementations may detect and report an error and may recover from it.
fatal error: A violation of the rules of this specification; results are undefined. Conforming implementations must not continue to process the document and may report an error.
at user option: Conforming software may or must (depending on the modal verb in the sentence) behave as described; if it does, it must provide users a means to enable or disable the behavior described.
2. Document model §
2.1 Data concepts §
An object is a data structure equivalent to the "object" type in JSON, consisting of a unordered set of name/value pairs (referred to here as fields) and where the name is a string and the value is a string, number, boolean, array, or object.
A document is a file containing a serialized object, or an array of objects.
A document type is a class of files that share a common structure and semantics.
A document schema is a formal description of the grammar of a document type.
A base URI is a context-dependent URI used to resolve relative references.
An identifier is a URI that designates a single document or single object within a document.
A vocabulary is the set of symbolic field names and enumerated symbols defined by a document schema, where each term maps to absolute URI.
2.2 Syntax §
Conforming Salad documents are serialized and loaded using YAML syntax and UTF-8 text encoding. Salad documents are written using the JSON-compatible subset of YAML. Features of YAML such as headers and type tags that are not found in the standard JSON data model must not be used in conforming Salad documents. It is a fatal error if the document is not valid YAML.
A Salad document must consist only of either a single root object or an array of objects.
2.3 Document context §
2.3.1 Implied context §
The implicit context consists of the vocabulary defined by the schema and the base URI. By default, the base URI must be the URI that was used to load the document. It may be overridden by an explicit context.
2.3.2 Explicit context §
If a document consists of a root object, this object may contain the
fields $base, $namespaces, $schemas, and $graph:
$base: Must be a string. Set the base URI for the document used to resolve relative references.$namespaces: Must be an object with strings as values. The keys of the object are namespace prefixes used in the document; the values of the object are the prefix expansions.$schemas: Must be an array of strings. This field may list URI references to documents in RDF-XML format which will be queried for RDF schema data. The subjects and predicates described by the RDF schema may provide additional semantic context for the document, and may be used for validation of prefixed extension fields found in the document.
Other directives beginning with $ must be ignored.
2.4 Document graph §
If a document consists of a single root object, this object may contain the
field $graph. This field must be an array of objects. If present, this
field holds the primary content of the document. A document that consists
of array of objects at the root is an implicit graph.
2.5 Document metadata §
If a document consists of a single root object, metadata about the document, such as authorship, may be declared in the root object.
2.6 Document schema §
Document preprocessing, link validation and schema validation require a document schema. A schema may consist of:
At least one record definition object which defines valid fields that make up a record type. Record field definitions include the valid types that may be assigned to each field and annotations to indicate fields that represent identifiers and links, described below in "Semantic Annotations".
Any number of enumerated type objects which define a set of finite set of symbols that are valid value of the type.
Any number of documentation objects which allow in-line documentation of the schema.
The schema for defining a salad schema (the metaschema) is described in detail in "Schema validation".
2.6.1 Record field annotations §
In a document schema, record field definitions may include the field
jsonldPredicate, which may be either a string or object. Implementations
must use the following document preprocessing of fields by the following
rules:
If the value of
jsonldPredicateis@id, the field is an identifier field.If the value of
jsonldPredicateis an object, and contains that object contains the field_typewith the value@id, the field is a link field.If the value of
jsonldPredicateis an object, and contains that object contains the field_typewith the value@vocab, the field is a vocabulary field, which is a subtype of link field.
2.7 Document traversal §
To perform document document preprocessing, link validation and schema validation, the document must be traversed starting from the fields or array items of the root object or array and recursively visiting each child item which contains an object or arrays.
3. Document preprocessing §
After processing the explicit context (if any), document preprocessing
begins. Starting from the document root, object fields values or array
items which contain objects or arrays are recursively traversed
depth-first. For each visited object, field names, identifier fields, link
fields, vocabulary fields, and $import and $include directives must be
processed as described in this section. The order of traversal of child
nodes within a parent node is undefined.
3.1 Field name resolution §
The document schema declares the vocabulary of known field names. During preprocessing traversal, field name in the document which are not part of the schema vocabulary must be resolved to absolute URIs. Under "strict" validation, it is an error for a document to include fields which are not part of the vocabulary and not resolvable to absolute URIs. Fields names which are not part of the vocabulary are resolved using the following rules:
- If an field name URI begins with a namespace prefix declared in the
document context (
@context) followed by a colon:, the prefix and
If there is a vocabulary term which maps to the URI of a resolved field, the field name must be replace with the vocabulary term.
If a field name URI is an absolute URI consisting of a scheme and path and is not part of the vocabulary, no processing occurs.
Field name resolution is not relative. It must not be affected by the base URI.
3.1.1 Field name resolution example §
Given the following schema:
{
"$namespaces": {
"acid": "http://example.com/acid#"
},
"$graph": [{
"name": "ExampleType",
"type": "record",
"fields": [{
"name": "base",
"type": "string",
"jsonldPredicate": "http://example.com/base"
}]
}]
}
Process the following example:
{
"base": "one",
"form": {
"http://example.com/base": "two",
"http://example.com/three": "three",
},
"acid:four": "four"
}
This becomes:
{
"base": "one",
"form": {
"base": "two",
"http://example.com/three": "three",
},
"http://example.com/acid#four": "four"
}
3.2 Identifier resolution §
The schema may designate one or more fields as identifier fields to identify specific objects. Processing must resolve relative identifiers to absolute identifiers using the following rules:
If an identifier URI is prefixed with
#it is a URI relative fragment identifier. It is resolved relative to the base URI by setting or replacing the fragment portion of the base URI.If an identifier URI does not contain a scheme and is not prefixed
#it is a parent relative fragment identifier. It is resolved relative to the base URI by the following rule: if the base URI does not contain a document fragment, set the fragment portion of the base URI. If the base URI does contain a document fragment, append a slash/followed by the identifier field to the fragment portion of the base URI.If an identifier URI begins with a namespace prefix declared in
$namespacesfollowed by a colon:, the prefix and colon must be replaced by the namespace declared in$namespaces.If an identifier URI is an absolute URI consisting of a scheme and path, no processing occurs.
When preprocessing visits a node containing an identifier, that identifier must be used as the base URI to process child nodes.
It is an error for more than one object in a document to have the same absolute URI.
3.2.1 Identifier resolution example §
Given the following schema:
{
"$namespaces": {
"acid": "http://example.com/acid#"
},
"$graph": [{
"name": "ExampleType",
"type": "record",
"fields": [{
"name": "id",
"type": "string",
"jsonldPredicate": "@id"
}]
}]
}
Process the following example:
{
"id": "http://example.com/base",
"form": {
"id": "one",
"things": [
{
"id": "two"
},
{
"id": "#three",
},
{
"id": "four#five",
},
{
"id": "acid:six",
}
]
}
}
This becomes:
{
"id": "http://example.com/base",
"form": {
"id": "http://example.com/base#one",
"things": [
{
"id": "http://example.com/base#one/two"
},
{
"id": "http://example.com/base#three"
},
{
"id": "http://example.com/four#five",
},
{
"id": "http://example.com/acid#six",
}
]
}
}
3.3 Link resolution §
The schema may designate one or more fields as link fields reference other objects. Processing must resolve links to either absolute URIs using the following rules:
If a reference URI is prefixed with
#it is a relative fragment identifier. It is resolved relative to the base URI by setting or replacing the fragment portion of the base URI.If a reference URI does not contain a scheme and is not prefixed with
#it is a path relative reference. If the reference URI contains#in any position other than the first character, the reference URI must be divided into a path portion and a fragment portion split on the first instance of#. The path portion is resolved relative to the base URI by the following rule: if the path portion of the base URI ends in a slash/, append the path portion of the reference URI to the path portion of the base URI. If the path portion of the base URI does not end in a slash, replace the final path segment with the path portion of the reference URI. Replace the fragment portion of the base URI with the fragment portion of the reference URI.If a reference URI begins with a namespace prefix declared in
$namespacesfollowed by a colon:, the prefix and colon must be replaced by the namespace declared in$namespaces.If a reference URI is an absolute URI consisting of a scheme and path, no processing occurs.
Link resolution must not affect the base URI used to resolve identifiers and other links.
3.3.1 Link resolution example §
Given the following schema:
{
"$namespaces": {
"acid": "http://example.com/acid#"
},
"$graph": [{
"name": "ExampleType",
"type": "record",
"fields": [{
"name": "link",
"type": "string",
"jsonldPredicate": {
"_type": "@id"
}
}]
}]
}
Process the following example:
{
"$base": "http://example.com/base",
"link": "http://example.com/base/zero",
"form": {
"link": "one",
"things": [
{
"link": "two"
},
{
"link": "#three",
},
{
"link": "four#five",
},
{
"link": "acid:six",
}
]
}
}
This becomes:
{
"$base": "http://example.com/base",
"link": "http://example.com/base/zero",
"form": {
"link": "http://example.com/one",
"things": [
{
"link": "http://example.com/two"
},
{
"link": "http://example.com/base#three"
},
{
"link": "http://example.com/four#five",
},
{
"link": "http://example.com/acid#six",
}
]
}
}
3.4 Vocabulary resolution §
The schema may designate one or more vocabulary fields which use terms defined in the vocabulary. Processing must resolve vocabulary fields to either vocabulary terms or absolute URIs by first applying the link resolution rules defined above, then applying the following additional rule:
* If a reference URI is a vocabulary field, and there is a vocabulary
term which maps to the resolved URI, the reference must be replace with
the vocabulary term.
3.4.1 Vocabulary resolution example §
Given the following schema:
{
"$namespaces": {
"acid": "http://example.com/acid#"
},
"$graph": [{
"name": "Colors",
"type": "enum",
"symbols": ["acid:red"]
},
{
"name": "ExampleType",
"type": "record",
"fields": [{
"name": "voc",
"type": "string",
"jsonldPredicate": {
"_type": "@vocab"
}
}]
}]
}
Process the following example:
{
"form": {
"things": [
{
"voc": "red",
},
{
"voc": "http://example.com/acid#red",
},
{
"voc": "http://example.com/acid#blue",
}
]
}
}
This becomes:
{
"form": {
"things": [
{
"voc": "red",
},
{
"voc": "red",
},
{
"voc": "http://example.com/acid#blue",
}
]
}
}
3.5 Import §
During preprocessing traversal, an implementation must resolve $import
directives. An $import directive is an object consisting of exactly one
field $import specifying resource by URI string. It is an error if there
are additional fields in the $import object, such additional fields must
be ignored.
The URI string must be resolved to an absolute URI using the link
resolution rules described previously. Implementations must support
loading from file, http and https resources. The URI referenced by
$import must be loaded and recursively preprocessed as a Salad document.
The external imported document does not inherit the context of the
importing document, and the default base URI for processing the imported
document must be the URI used to retrieve the imported document. If the
$import URI includes a document fragment, the fragment must be excluded
from the base URI used to preprocess the imported document.
Once loaded and processed, the $import node is replaced in the document
structure by the object or array yielded from the import operation.
URIs may reference document fragments which refer to specific an object in
the target document. This indicates that the $import node must be
replaced by only the object with the appropriate fragment identifier.
It is a fatal error if an import directive refers to an external resource or resource fragment which does not exist or is not accessible.
3.5.1 Import example §
import.yml:
{
"hello": "world"
}
parent.yml:
{
"form": {
"bar": {
"$import": "import.yml"
}
}
}
This becomes:
{
"form": {
"bar": {
"hello": "world"
}
}
}
3.6 Include §
During preprocessing traversal, an implementation must resolve $include
directives. An $include directive is an object consisting of exactly one
field $include specifying a URI string. It is an error if there are
additional fields in the $include object, such additional fields must be
ignored.
The URI string must be resolved to an absolute URI using the link
resolution rules described previously. The URI referenced by $include must
be loaded as a text data. Implementations must support loading from
file, http and https resources. Implementations may transcode the
character encoding of the text data to match that of the parent document,
but must not interpret or parse the text document in any other way.
Once loaded, the $include node is replaced in the document structure by a
string containing the text data loaded from the resource.
It is a fatal error if an import directive refers to an external resource which does not exist or is not accessible.
3.6.1 Include example §
parent.yml:
{
"form": {
"bar": {
"$include": "include.txt"
}
}
}
include.txt:
hello world
This becomes:
{
"form": {
"bar": "hello world"
}
}
4. SaladRecordSchema §
Fields
typerecordMust be record
docA documentation string for this type, or an array of strings which should be concatenated.
docParentHint to indicate that during documentation generation, documentation
for this type should appear in a subsection under docParent.
docChildHint to indicate that during documentation generation, documentation
for docChild should appear in a subsection under this type.
docAfterHint to indicate that during documentation generation, documentation
for this type should appear after the docAfter section at the same
level.
documentRootIf true, indicates that the type is a valid at the document root. At
least one type in a schema must be tagged with documentRoot: true.
abstractIf true, this record is abstract and may be used as a base for other records, but is not valid on its own.
extendsIndicates that this record inherits fields from one or more base records.
specializeOnly applies if extends is declared. Apply type specialization using the
base record as a template. For each field inherited from the base
record, replace any instance of the type specializeFrom with
specializeTo.
4.1 SaladRecordField §
A field of a record.
Fields
typeThe field type
4.1.1 PrimitiveType §
Salad data types are based on Avro schema declarations. Refer to the Avro schema declaration documentation for detailed information.
Symbols
| symbol | description |
|---|---|
null | no value |
boolean | a binary value |
int | 32-bit signed integer |
long | 64-bit signed integer |
float | single precision (32-bit) IEEE 754 floating-point number |
double | double precision (64-bit) IEEE 754 floating-point number |
string | Unicode character sequence |
4.1.2 RecordSchema §
Fields
typerecordMust be record
4.1.3 RecordField §
A field of a record.
Fields
typeThe field type
4.1.3.1 EnumSchema §
Define an enumerated type.
Fields
typeenumMust be enum
4.1.3.2 ArraySchema §
Fields
typearrayMust be array
itemsDefines the type of the array elements.
4.1.4 JsonldPredicate §
Attached to a record field to define how the parent record field is handled for URI resolution and JSON-LD context generation.
Fields
_idThe predicate URI that this field corresponds to.
Corresponds to JSON-LD @id directive.
_typeThe context type hint, corresponds to JSON-LD @type directive.
If the value of this field is
@idandidentityis false or unspecified, the parent field must be resolved using the link resolution rules. Ifidentityis true, the parent field must be resolved using the identifier expansion rules.If the value of this field is
@vocab, the parent field must be resolved using the vocabulary resolution rules.
identityIf true and _type is @id this indicates that the parent field must
be resolved according to identity resolution rules instead of link
resolution rules. In addition, the field value is considered an
assertion that the linked value exists; absence of an object in the loaded document
with the URI is not an error.
noLinkCheckIf true, this indicates that link validation traversal must stop at this field. This field (it is is a URI) or any fields under it (if it is an object or array) are not subject to link checking.
4.2 SpecializeDef §
Fields
5. SaladEnumSchema §
Define an enumerated type.
Fields
typeenumMust be enum
docA documentation string for this type, or an array of strings which should be concatenated.
docParentHint to indicate that during documentation generation, documentation
for this type should appear in a subsection under docParent.
docChildHint to indicate that during documentation generation, documentation
for docChild should appear in a subsection under this type.
docAfterHint to indicate that during documentation generation, documentation
for this type should appear after the docAfter section at the same
level.
documentRootIf true, indicates that the type is a valid at the document root. At
least one type in a schema must be tagged with documentRoot: true.
6. Documentation §
A documentation section. This type exists to facilitate self-documenting schemas but has no role in formal validation.
Fields
typedocumentationMust be documentation
docA documentation string for this type, or an array of strings which should be concatenated.
docParentHint to indicate that during documentation generation, documentation
for this type should appear in a subsection under docParent.
docChildHint to indicate that during documentation generation, documentation
for docChild should appear in a subsection under this type.
docAfterHint to indicate that during documentation generation, documentation
for this type should appear after the docAfter section at the same
level.