@ldo/jsonld-dataset-proxy
v1.0.0-alpha.32TypeScript
Edit RDFJS Dataset just like regular JavaScript Object Literals.
0/weekUpdated 5 months agoMITUnpacked: 427.4 KB
Published by Jackson Morgan
npm install @ldo/jsonld-dataset-proxyJSONLD Dataset Proxy
Edit RDFJS Dataset just like regular JavaScript Object Literals.
Just a few lines of familiar code:
``typescript`
const personNode = namedNode("http://example.com/Person1");
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject
person.age = 23;
person.name.add("John");
are equivalent to:
`typescript`
dataset.deleteMatches(
namedNode("http://example.com/Person1"),
namedNode("http://xmlns.com/foaf/0.1/age")
);
dataset.add(
quad(
namedNode("http://example.com/Person1"),
namedNode("http://xmlns.com/foaf/0.1/age"),
literal("23", "http://www.w3.org/2001/XMLSchema#integer")
)
);
dataset.add(
quad(
namedNode("http://example.com/Person1"),
namedNode("http://xmlns.com/foaf/0.1/name"),
literal("John", "http://www.w3.org/2001/XMLSchema#string")
)
);
Plus, you get IntelliSense typescript suggestions to help you write your code!
Installation
`
bash
npm install @ldo/jsonld-dataset-proxy
`Simple Example
typescript
import jsonldDatasetProxy, { write, LdSet } from "jsonld-dataset-proxy";
import { ContextDefinition } from "jsonld";
import { serializedToDataset } from "@ldo/dataset";
import { namedNode } from "@ldo/rdf-utils";async function start() {
// Define initial data
const initialData =
@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:name "Johnathan"^^xsd:string;
foaf:age "22"^^xsd:integer.
;
// Create a dataset loaded with initial data
const dataset = await serializedToDataset(initialData);
// Make a JSONLD Dataset Proxy
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
// Make Modifications
person.age = 23;
person.name.add("John");
write(namedNode("http://example.com/otherGraph")).using(person);
person.name.add("Smith"); console.log(dataset.toString());
// Logs:
// "Johnathan" .
// "John" .
// "23"^^ .
// "Smith" .
}
// Person Typescript Typing
interface IPerson {
name: LdSet;
age: number;
}
// Person JSONLD Context
const PersonContext: ContextDefinition = {
name: {
"@id": "http://xmlns.com/foaf/0.1/name",
"@type": "http://www.w3.org/2001/XMLSchema#string",
"@container": "@set",
},
age: {
"@id": "http://xmlns.com/foaf/0.1/age",
"@type": "http://www.w3.org/2001/XMLSchema#integer",
},
};
start();
`Full Usage
- Defining a Context and Type
- Getting a Jsonld DatasetProxy
-
-
-
-
- Getting Field Values and Traversing
- Setting a Primitive
- Setting an Object
- Set Methods
- Overwriting an Object
- Changing an Object's Id
- Removing an Object Connection
- Deleting an Entire Object
- Using Blank Nodes
- Writing Information to a Specific Graph
-
-
-
- Detecting a the graph of specific informationFor the most part, you can think of a JSONLD Dataset Proxy as a JavaScript Object Literal with "Sets" instead of arrays.
$3
The first step to getting a JSONLD Dataset Proxy is defining the JSONLD Context and TypeScript Typings. This can either be done through a generator or defining them manually.In this example typescript typing
is an interface that represents a person. Notice the @id and @context fields. Be sure to include them in your interfaces if you wish to use those properties.`typescript
import { LdoJsonldContext, LdSet } from "@ldo/jsonld-dataset-proxy";interface IPerson {
"@id"?: string;
"@context"?: LdoJsonldContext;
name?: LdSet;
age?: number;
bestFriend?: IPerson;
knows?: LdSet;
}
`We can make a JSONLD context to match this type:
typescript
import { ContextDefinition } from "jsonld";const PersonContext: ContextDefinition = {
name: {
"@id": "http://xmlns.com/foaf/0.1/name",
"@type": "http://www.w3.org/2001/XMLSchema#string",
"@container": "@set",
},
age: {
"@id": "http://xmlns.com/foaf/0.1/age",
"@type": "http://www.w3.org/2001/XMLSchema#integer",
},
bestFriend: {
"@id": "http://xmlns.com/foaf/0.1/bestFriend",
"@type": "@id",
},
knows: {
"@id": "http://xmlns.com/foaf/0.1/knows",
"@type": "@id",
"@container": "@set",
},
};
`To do this, create an object that has corresponding fields to your type. Each field is an object that contains the following properties:
-
: indicates the URI of the corresponding predicate
- @type: If the corresponding type is a pimitive (Like a number or string), use this field to list the RDF Literal type (Most often this is an XMLSchema type). If the corresponding type is an object, list @id here.
- @container: If the corresponding type is an array of items, set @container to @set, if not, do not include the @container property.Note that only the features described here work with JSONLD Dataset Proxy. Other features of JSONLD Contexts are not yet supported.
$3
Once the Typescript Typings and Context have been defined, we can get the JSONLD Dataset Proxy for a specific dataset.`typescript
import jsonldDatasetProxy from "jsonld-dataset-proxy";
import { createDataset } from "o-dataset-pack";const dataset = await createDataset();
// Make a JSONLD Dataset Proxy
const person = jsonldDatasetProxy(
dataset,
PersonContext
) // ...
`The functon
takes in three parameters:
- dataset: The dataset you wish to traverse and manipulate. This can be any dataset that follows the RDFJS Dataset Interface. Note that this is not to be confused with the RDFJS Dataset Core interface. This example uses the "o-dataset-pack", but any implementation of the RDFJS Dataset Interface is acceptable.
- context: The JSONLD context. After defining the
dataset and context there are a few ways to get Jsonld Dataset Proxies:####
.fromSubject
lets you define a an entryNode, the place of entry for the graph. The object returned by jsonldDatasetProxy will represent the given node. This parameter accepts both namedNodes and blankNodes. fromSubject takes a generic type representing the typescript type of the given subject.`typescript
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
`####
returns a Jsonld Dataset Proxy representing all subjects in the dataset matching the given predicate, object, and graph.`typescript
const people = jsonldDatasetProxy(
dataset,
PersonContext
).matchSubject(
namedNode("http://www.w3.org/1999/02/22-rdf-syntax-ns#type"),
namedNode("http://xmlns.com/foaf/0.1/Person")
);
people.forEach((person) => {
console.log(person.name);
});
`####
returns a Jsonld Dataset Proxy representing all objects in the dataset matching the given subject, predicate, and graph.`typescript
const friendsOfPerson1 = jsonldDatasetProxy(
dataset,
PersonContext
).matchSubject(
namedNode("http://example.com/Person1"),
namedNode("http://xmlns.com/foaf/0.1/knows")
);
friendsOfPerson1.forEach((person) => {
console.log(person.name);
});
`####
will take any regular Json, add the information to the dataset, and return a Jsonld Dataset Proxy representing the given data.`typescript
import { jsonldDatasetProxy, set } from "@ldo/jsonld-dataset-proxy";const person2 = jsonldDatasetProxy(
dataset,
PersonContext
).fromJson({
"@id": "http://example.com/Person2",
name: set("Jane", "Doe"),
birthdate: "1990/11/03",
age: 33,
});
`$3
Getting a field and traversing the object is just as easy as getting data out of a standard JavaScript Object Literal.In all the following example, we will use a dataset loaded with the following data:
typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:name "Johnathan"^^xsd:string, "John"^^xsd:string;
foaf:age "22"^^xsd:integer;
foaf:bestFriend example:Person2;
foaf:friends example:Person2, example:Person3.
example:Person2
foaf:name "Alice"^^xsd:string;
foaf:age "28"^^xsd:integer.
example:Person3
foaf:name "Dave"^^xsd:string;
foaf:age "33"^^xsd:integer.
);
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
// Get primitives
console.log(person.age); // 22
// Get nested primitives
console.log(person?.bestFriend?.age); // 28
// All array methods work
console.log(person.name?.reduce((agg, cur) => agg + cur, "")); // JonathanJohn
// You can also access array items via their index
// But this isn't recommened. The library will do its best to maintain the
// ordering in the array, but as datasets have no concept of order, this is
// not always accurate.
console.log(person.name?.toArray()[0]); // John
// Get the id of the object
// (If the node is a blankNode the @id will be undefined)
console.log(person.bestFriend?.["@id"]); // "http://example.com/Person2"
// Finally, you can retrieve the context
console.log(person["@context"]); // { "name": { "@id": ... }}
`$3
Setting a non-array primitive will remove the existing triple from the dataset and add a new triple.typescript
const dataset = createDataset();
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"))
person.age = 23;
console.log(dataset.toString());
// "23"^^ .
`$3
Setting a field to a JavaScript object literal will recursively add all parts of the object literal to the dataset.typescript
import { set } from "@ldo/jsonld-dataset-pack";const dataset = createDataset();
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
person.bestFriend = {
"@id": "http://example.com/Person2",
name: set("Alice"),
bestFriend: {
"@id": "http://example.com/Person3",
name: set("Bob"),
},
};
console.log(dataset.toString());
// .
// .
// "Alice" .
// "Bob" .
`$3
Any methods that on a JavaScript Set are available.
typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:name "Garrett"^^xsd:string, "Bobby"^^xsd:string.
);
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
person.name?.add("Ferguson");
console.log(dataset.toString());
// "Garrett" .
// "Bobby" .
// "Ferguson" .
`For convenience, some methods commonly used on arrays have been added. The methods available on Sets are defined as follows:
typescript
LdSet {
add(value: T): this;
clear(): void;
delete(value: T): boolean;
has(value: T): boolean;
readonly size: number;
[Symbol.iterator](): IterableIterator;
entries(): IterableIterator<[T, T]>;
keys(): IterableIterator;
values(): IterableIterator;
every(predicate: (value: T, set: LdSet) => value is S, thisArg?: any): this is LdSet;
every(predicate: (value: T, set: LdSet) => unknown, thisArg?: any): boolean;
some(predicate: (value: T, set: LdSet) => unknown, thisArg?: any): boolean;
forEach(callbackfn: (value: T, value2: T, set: LdSet) => void, thisArg?: any): void;
map(callbackfn: (value: T, set: LdSet) => U, thisArg?: any): U[];
filter(predicate: (value: T, set: LdSet) => value is S, thisArg?: any): LdSet;
filter(predicate: (value: T, set: LdSet) => unknown, thisArg?: any): LdSet;
reduce(callbackfn: (previousValue: T, currentValue: T, set: LdSet) => T): T;
reduce(callbackfn: (previousValue: T, currentValue: T, set: LdSet) => T, initialValue: T): T;
reduce(callbackfn: (previousValue: U, currentValue: T, array: LdSet) => U, initialValue: U): U;
toArray(): T[];
toJSON(): T[];
difference(other: Set): LdSet;
intersection(other: Set): LdSet;
isDisjointFrom(other: Set): boolean;
isSubsetOf(other: Set): boolean;
isSupersetOf(other: Set): boolean;
symmetricDifference(other: Set): LdSet;
union(other: Set): LdSet;
}
`$3
If an object literal is set and the id is equivalent to an existing id, that node will be overwritten. All triples from the previous object are removed and replaced with triples from the new object.typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person2
foaf:name "Alice"^^xsd:string;
foaf:age "28"^^xsd:integer.
);
const person = jsonldDatasetProxy(
dataset,
PersonContext,
).fromSubject(namedNode("http://example.com/Person1"));
person.bestFriend = {
"@id": "http://example.com/Person2",
name: set("Jane"),
};
console.log(dataset.toString());
// "Jane" .
// .
`$3
You can rename an object by setting its field. This will update all triples that reference the id to the new id. Setting the @id field to undefined will turn it into a blank node.`typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:name "Alice"^^xsd:string;
foaf:bestFriend example:Person2.
example:Person2
foaf:bestFriend example:Person1.);`
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject
person["@id"] = "http://example.com/NewPersonId";
console.log(dataset.toString());
//
//
//
$3
Removing a connection between nodes can be done by setting a field to undefined
or using the delete operator. Values can be removed from a set by using the delete method or the clear method to remove all connections.`typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:name "Alice"^^xsd:string;
foaf:bestFriend example:Person2.
example:Person2
foaf:name "Bob"^^xsd:string;
foaf:bestFriend example:Person1.);`
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject
perons.bestFriend.name.delete("Bob");
person.bestFriend = undefined;
console.log(dataset.toString());
//
//
$3
If you want to delete all triples represented by an object you can call delete
on the @id property.`typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:name "Alice"^^xsd:string;
foaf:bestFriend example:Person2.
example:Person2
foaf:name "Bob"^^xsd:string;
foaf:bestFriend example:Person1.);`
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject
delete person["@id"];
console.log(dataset.toString());
//
$3
If you want to create an object with a blankNode subject, simply omit the @id
field when you're making the object.
`typescript
const dataset = await createDataset();
const person = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
person.bestFriend = {
name: set("Charlie"),
};
console.log(dataset.toString());
// _:b1 .
// _:b1 "Charlie" .
`If your dataset has blank nodes and you want to assign that blank node as a triple's object, you can retrieve it from the JSONLD Dataset Proxy and assign it.
typescript
const dataset = await serializedToDataset(@prefix example:
@prefix foaf:
@prefix xsd:
example:Person1
foaf:knows [
foaf:name "Alice"^^xsd:string;
].
);
const person = jsonldDatasetProxy(
dataset,
PersonContext,
).fromSubject(namedNode("http://example.com/Person1"));const alice = person.knows?.toArray()[0];
person.bestFriend = alice;
console.log(dataset.toString());
// _:n3-0 "Alice" .
// _:n3-0 .
// _:n3-0 .
`$3
By default, all new quads are added to the default graph, but you can change the graph to which new quads are added in a few different ways:NOTE: These operations only dictate the graph for new triples. Any operations that delete triples will delete triples regardless of their graph.
####
The write graph can be set upon creating a jsonld dataset proxy by using the method. This method takes in any number of graphs.`typescript
const person1 = jsonldDatasetProxy(dataset, PersonContext)
.write(namedNode("http://example.com/ExampleGraph"))
.fromSubject(namedNode("http://example.com/Person1"));
person1.name.add("Jack");
console.log(dataset.toString());
// Logs:
// "Jack" .
`####
The function lets you define the graphs you wish to write to using specific jsonldDatasetProxies.`typescript
import jsonldDatasetProxy, { write } from "jsonld-dataset-proxy";const person1 = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
// Now all additions with person1 will be on ExampleGraph1
write(namedNode("http://example.com/ExampleGraph1")).using(person1);
person1.name.add("Jack");
// Now all additions with person1 will be on ExampleGraph2
write(namedNode("http://example.com/ExampleGraph2")).using(person1);
person1.name.add("Spicer");
console.log(dataset.toString());
// Logs:
// "Jack" .
// "Spicer" .
`The function also returns an
function that will reset the graph to what it was before. This is useful for nesting graph modifications.`typescript
const person1 = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
person1.name.add("default");
const end1 = write(namedNode("http://example.com/Graph1")).using(person1);
person1.name.add("1");
const end2 = write(namedNode("http://example.com/Graph2")).using(person1);
person1.name.add("2");
const end3 = write(namedNode("http://example.com/Graph3")).using(person1);
person1.name.add("3");
end3();
person1.name.add("2 again");
end2();
person1.name.add("1 again");
end1();
person1.name.add("default again");
console.log(dataset.toString());
// Logs:
// "default" .
// "default again" .
// "1" .
// "1 again" .
// "2" .
// "2 again" .
// "3" .
`####
If you would like a new variable to write to without modifying the original Jsonld Dataset Proxy, you can use .`typescript
const person1 = jsonldDatasetProxy(
dataset,
PersonContext
).fromSubject(namedNode("http://example.com/Person1"));
const [person1WritingToNewGraph] = write(
namedNode("http://example.com/NewGraph")
).usingCopy(person1);
person1WritingToNewGraph.name.add("Brandon");
person1.name.add("Sanderson");
console.log(dataset.toString());
// Logs:
// "Brandon" .
// "Sanderson" .
`$3
The graph of specific information can be detected using the
function. The graphOf function takes in two to three arguments. -
subject: A Jsonld Dataset Proxy that represents the subject of a quad.
- predicate: A string key
- object?: A representation of the direct object of the triple in question. This could be a jsonld-dataset-proxy or simply and object with an @id field.`typescript
graphOf(person, "name", "Bob"); // returns defaultGraph()
graphOf(person, "age"); // returns defaultGraph()
`Managing Language Tags
RDF includes a special attribute for string literals called a language tag. Lanugage tags let developers provide string representations for many different translations and JSON-LD Dataset Proxy helps you manage them.
To use language tags, RDF requires the datatype of a literal to be
, and LDO's functions will only work on literals of type that type.For the following examples, we'll use this context and dataset, typescript typing and JSON-LD Context. Notice that there is a field called "label" with translations for French and Korean and one language string that doesn't have a language tag. There's also a field called "description" that holds multiple strings per language.
`typescript
// Define initial data
const initialData = @prefix example:
@prefix rdfs:
@prefix ns:
example:Hospital
rdfs:label "Hospital"^^ns:langString;
rdfs:label "Hôpital"@fr;
rdfs:label "병원"@ko;
rdfs:description "Heals patients"^^ns:langString;
rdfs:description "Has doctors"^^ns:langString;
rdfs:description "Guérit les malades"@fr;
rdfs:description "A des médecins"@fr;
rdfs:description "환자를 치료하다"@ko;
rdfs:description "의사 있음"@ko.
;// Typescript Typing
interface IThing {
label: string;
description: string[];
}
// Define JSON-LD Context
const PersonContext: ContextDefinition = {
label: {
"@id": "http://www.w3.org/2000/01/rdf-schema#label",
"@type": "http://www.w3.org/1999/02/22-rdf-syntax-ns#langString",
},
description: {
"@id": "http://www.w3.org/2000/01/rdf-schema#description",
"@type": "http://www.w3.org/1999/02/22-rdf-syntax-ns#langString",
"@container": "@set",
},
};
`$3
A language preference is an ordered list telling the JSON-LD Dataset Proxy the language you prefer as well as callbacks. Valid values for the language preferences includes any IETF Language Tag as well as the special tags
and @other. @none represents any language literal that doesn't have a language tag. @other represents any language literal that isn't listed among the language preferences.For read operations, the JSON-LD Dataset Proxy will search for values in order of the preference. Write operations will choose the first language in the language preference, unless that language is
@other, in which case it will choose the next language.`typescript
// Read Spansih first, then Korean, then language strings with no language
// New writes are in Spanish
["es", "ko", "@none"]// Read any language other than french, then french
// New writes are in French
["@other", "fr"]
`Language preferences can be set when a JSON-LD Dataset Proxy is created using the
method.`typescript
// Create a dataset loaded with initial data
const dataset = await serializedToDataset(initialData);
// Make a JSONLD Dataset Proxy
const hospitalInfo = jsonldDatasetProxy(dataset, PersonContext)
.setLanguagePreferences("es", "ko", "@none")
.fromSubject(namedNode("http://example.com/Hospital"));console.log(hospitalInfo.label); // Logs "병원"
console.log(hospitalInfo.description.size); // Logs "2" for the 2 korean entries
console.log(hospitalInfo.description.toArray()[0]); // Logs "환자를 치료하다"
console.log(hospitalInfo.description.toArray()[1]); // Logs "의사 있음"
// Adds a string to the description in spanish, because spanish if the first
// language in the language preference
hospitalInfo.description.add("Cura a las pacientes");
// Now that a spanish entry exists, JSON-LD dataset proxy focuses on that
console.log(hospitalInfo.description.length); // Logs "1" for the 1 spanish entry
console.log(hospitalInfo.description.toArray()[0]); // Logs "Cura a las pacientes"
`$3
The function sets the language preferences for a set of JSON-LD Dataset Proxies. It follows roughly the same paridigms as the write(...).using(...) function.`typescript
import { setLanguagePreferences } from "jsonld-dataset-proxy";setLanguagePreferences("fr", "ko").using(hospitalInfo);
console.log(hospitalInfo.label); // Logs "Hôpital"
setLanguagePreferences("@none").using(hospitalInfo);
console.log(hospitalInfo.label); // Logs "Hospital"
`$3
The function returns a copy of the provided JSON-LD Dataset Proxies with the given language preferences. It follows roughly the same paridigms as the write(...).usingCopy(...) function.`typescript
import { setLanguagePreferences } from "jsonld-dataset-proxy";// ...
const [frenchPreference] = setLanguagePreferences("fr").usingCopy(hospitalInfo);
const [koreanPreference] = setLanguagePreferences("ko").usingCopy(hospitalInfo);
console.log(frenchPreference.label); // Logs "Hôpital"
console.log(koreanPreference.label); // Logs "병원"
`$3
The function lets you view and modify the languages more directly. languageOf takes two properties: -
jsonldDatasetProxy: A JSON-LD dataset proxy
- key: A key on the JSON-LD dataset proxy pointing to a language string.It returns a mapping of languages to strings or sets of strings depending on the cardinality of the JSON-LD context.
`typescript
const labelLanguages = languagesOf(hospitalInfo, "label");
// labelLanguages: { '@none': 'Hospital', fr: 'Hôpital', ko: '병원' }
const descriptionLanguages = languagesOf(hospitalInfo, "description");
// descriptionLanguages:
// {
// '@none': Set(2) { 'Heals patients', 'Has doctors' },
// fr: Set(2) { 'Guérit les malades', 'A des médecins' },
// ko: Set(2) { '환자를 치료하다', '의사 있음' }
// }
`You can also modify languauages by changing the mappings. Mappings with sets of strings follow the JavaScript
interface.`typescript
// Adds a Chinese label
labelLanguages.zh = "医院";
// Changes the no-language label from to "Super Hospital"
labelLanguages["@none"] = "Super Hospital";
// Removes the French label
delete labelLanguages.fr;
// Adds a Hindi description
descriptionLanguages.hi?.add("रोगियों को ठीक करता है");
// Checks to see if the korean label contains "의사 있음"
descriptionLanguages.ko?.has("의사 있음"); // returns true
// Removes "Has Doctors" from the no-language description
descriptionLanguages["@none"]?.delete("Has Doctors");
``Limitations
- Currently this library only supports the following features of JSON-LD context:
- "@id",
- "@type",
- "@container": "@set"
Sponsorship
This project was made possible by a grant from NGI Zero Entrust via nlnet. Learn more on the NLnet project page.
Liscense
MIT