import JupyterContext from 'stencila-node/lib/contexts/JupyterContext.js'

public class | source

JupyterContext

Extends:

Context → JupyterContext

An execution context using Jupyter kernels

This class of execution context acts as a bridge between Stencila and Jupyter kernels. It exposes methods of the Stencila Context API e.g. executeEval which delegate execution to a Jupyter kernel. This is done via the Jupyter Messageing Protocol (JMP) over ZeroMQ sockets.

The discover static method should be called initially to find all Jupyter kernels currently installed on the machine and update JupyterContext.spec.kernels:

JupyterContext.discover()

New Jupyter execution contexts can be constructed using the language option which will search for a kernel with a matching lowercased language property:

new JupyterContext({language:'r'})

Alternively, you can specify a kernel directly:

new JupyterContext({kernel:'ir'})

See https://github.com/jupyter/jupyter/wiki/Jupyter-kernels for a list of available Jupyter kernels.

Many thanks to the nteract community for kernelspecs and spawnteract, and to Nicolas Riesco for (jmp)[https://github.com/n-riesco/jmp], all of which made this implementation far easier!

Static Method Summary

Static Public Methods
public static	discover(): Promise Discover Jupyter kernels on the current machine

Constructor Summary

Public Constructor
public	constructor(options: Object) Construct a Jupyter execution context

Member Summary

Public Members
public	debug: *
public	kernel: *
public	timeout: *

Private Members
private	_config: *
private	_connectionFile: *
private	_ioSocket: *
private	_kernelInfo: *
private	_process: *
private	_requests: {}
private	_sessionId: *
private	_shellSocket: *
private	_spec: *

Method Summary

Public Methods
public	async execute(cell: ): Execute a cell
public	finalize(): Promise Finalize the context
public	initialize(): Promise Initialize the context

Private Methods
private	_request(requestType: String, content: Object, responseTypes: String): Promise Send a request message to the kernal
private	_response(response: Message) Receive a response message from the kernel
private	_unbundle(bundle: Object): Promise Convert a "MIME bundle" within a JMP message (e.g.

Inherited Summary

From class Context
public get	host: *
public get	id: *
public get	location: {"context": , "host": , "machine": *}
public get	name: *
private	_host: *
private	_id: string
private	_name: *
private	_variables: {}
public	async compile(cell: ): {"source": , "expr": , "global": , "options": , "inputs": , "outputs": , "messages": }
public	async evaluate(node: ):
public	async execute(cell: ):
public	async libraries(): {}
public	async pack(value: ):
public	async packPackage(value: ): {"type": , "data": *}
public	async packPointer(objectPattern: {"type": , "name": , "preview": }): {"type": , "path": , "preview": }
public	async provide(what: ):
public	async resolve(what: ):
public	async unpack(node: Object): [type] Unpack a data node into a native data value
public	async unpackPackage(pkg: ):
public	async unpackPointer(pointer: ):

Static Public Methods

public static discover(): Promise source

Discover Jupyter kernels on the current machine

Looks for Jupyter kernels that have been installed on the system and puts that list in JupyterContext.spec.kernels so that peers know the capabilities of this "meta-context".

Return:

Promise

A promise

Public Constructors

public constructor(options: Object) source

Construct a Jupyter execution context

Override:

Context#constructor

Params:

Name	Type	Attribute	Description
options	Object		Options for specifying which kernel to use

Public Members

public debug: * source

public kernel: * source

public timeout: * source

Private Members

private _config: * source

private _connectionFile: * source

private _ioSocket: * source

private _kernelInfo: * source

private _process: * source

private _requests: {} source

private _sessionId: * source

private _shellSocket: * source

private _spec: * source

Public Methods

public async execute(cell: ): source

Execute a cell

For cells with expr: true utilises user_expressions property of an execute_request to evaluate expression side-effect free.

Override:

Context#execute

Params:

Name	Type	Attribute	Description
cell	*

Return:

public finalize(): Promise source

Finalize the context

Return:

Promise

A resolved promise

public initialize(): Promise source

Initialize the context

Return:

Promise

A promise

Private Methods

private _request(requestType: String, content: Object, responseTypes: String): Promise source

Send a request message to the kernal

Params:

Name	Type	Attribute	Description
requestType	String		Type of request e.g. 'execute'
content	Object		Content of message
responseTypes	String		Types of response message to resolve

Return:

Promise

Promise resolving to the {request, response} messages

private _response(response: Message) source

Receive a response message from the kernel

Params:

Name	Type	Attribute	Description
response	Message		Response message

private _unbundle(bundle: Object): Promise source

Convert a "MIME bundle" within a JMP message (e.g. a execute_result or display data message) into a data node e.g. {'text/plain': 'Hello'} to {type: 'string', data: 'Hello'}

Params:

Name	Type	Attribute	Description
bundle	Object		A JMP MIME bundle

Return:

Promise

Promise resolving to a data node

JupyterContext

Extends:

Static Method Summary

Constructor Summary

Member Summary

Method Summary

Inherited Summary

Static Public Methods

public static discover(): Promise source

Return:

Public Constructors

public constructor(options: Object) source

Override:

Params:

Public Members

public debug: * source

public kernel: * source

public timeout: * source

Private Members

private _config: * source

private _connectionFile: * source

private _ioSocket: * source

private _kernelInfo: * source

private _process: * source

private _requests: {} source

private _sessionId: * source

private _shellSocket: * source

private _spec: * source

Public Methods

public async execute(cell: *): * source

Override:

Params:

Return:

public finalize(): Promise source

Return:

public initialize(): Promise source

Return:

Private Methods

private _request(requestType: String, content: Object, responseTypes: String): Promise source

Params:

Return:

private _response(response: Message) source

Params:

private _unbundle(bundle: Object): Promise source

Params:

Return:

public async execute(cell: ): source