@seneca/repl
TypeScript icon, indicating that this package has built-in type declarations

9.1.0 • Public • Published

Seneca REPL

Seneca REPL is a plugin for Seneca

A plugin that provides a REPL (Read-Execute-Print-Loop) for Seneca instances, very much like the standard Node.js command line REPL, with a some extra Seneca convenience commands.

npm version build Coverage Status Known Vulnerabilities DeepScan grade Maintainability

Voxgig This open source module is sponsored and supported by Voxgig.

Seneca compatibility

Supports Seneca versions 3.x and higher.

Recent Articles

Install

This is Seneca plugin, so you'll also need the Seneca framework installed to use the REPL.

$ npm install seneca
$ npm install @seneca/repl

To use the REPL client on the command line, you should install globally:

$ npm i -g seneca @seneca/repl
$ seneca-repl # now works!

Installing optional components

This plugin can provide a REPL for AWS Lambda functions (via invoke). You will need to install the AWS SDK so the REPL client can use it to connect to your lambda function.

$ npm i -g @aws-sdk/client-lambda

Usage

Add the REPL as a plugin to your Seneca instance. By default the plugin will listen on localhost port 30303.

var Seneca = require('seneca')

var seneca = Seneca()
  // open repl on default port 30303
  .use('repl') 

  // open another repl on port 10001
  .use('repl', {port: 10001})

  // open yet another repl on a free port chosen by your OS
  // look at the INFO level logs for the host and port
  // or get them from seneca.export('repl/address')
  .use('repl', {port: 0})

To access the REPL, run the seneca-repl command provided by this plugin.

$ seneca-repl

You can specify the target Seneca server using a URI

$ seneca-repl telnet://localhost:30303 # same as default

NOTE: If the connection drops, the seneca-repl client will attempt to reconnect at regular intervals. This means you can stop and start your development server without needing to restart the REPL.

Skip ahead to the Commands section if this is all you need.

REPL over Seneca Message

You can submit REPL commands using the message sys:repl,send:cmd. This message requires an id property to indicate the REPL instance to use:

const Seneca = require('seneca')

const seneca = Seneca()
  .use('promisify')  // npm install @seneca/promisify
  .use('repl')
  .act('sys:repl,use:repl,id:foo')
  
await seneca.ready()

let res = await seneca.post('sys:repl,send:cmd,id:foo', {
  cmd: '1+1'
})

// Prints { ok: true, out:'4\n' }
console.log(res)

You can use this to expose a REPL connector in custom environments. This plugin provides a REPL over HTTP, and over AWS Lambda invocations. Review the implementation code for these if you want to write your own REPL connector.

REPL over HTTP(S)

Opening a local port is usually only possible for local development, so you can also expose the REPL via a HTTP endpoint. This can be useful to debug build or staging systems. This is NOT recommended for production.

WARNING This is a security risk. Your app will need to apply additional constraints to prevent arbitrary message submission via the REPL.

On the server, use the sys:repl,use:repl message to start a new REPL inside the Seneca instance. Do this on startup (without a REPL instance, a REPL connection will not operate). You will need to special an identifier for this REPL.

seneca.act('sys:repl,use:repl,id:web')

Next you will need to call the sys:repl,send:cmd message when your chosen HTTP endpoint for the REPL is called. For express, this might look like:

const Express = require('express')
const BodyParser = require('body-parser')
const Seneca = require('seneca')

const app = express()
const seneca = Seneca()

seneca
  .use('repl')
  .act('sys:repl,use:repl,id:web')
    
app.use(bodyParser.json())

// Accepts body = {cmd:'...repl cmd goes here...'}
app.post('/seneca-repl', (req, res) => {
  const body = req.body

  seneca.act(
    { sys: 'repl', send: 'cmd', id: 'web', cmd: body.cmd }, 
    function (err, result) {
      if (err) {
        return res.status(500).json({ ok: false, error: err.message })
      }

      return res.json(result.out)
    })
})

app.listen(8080)

On the command line, access the REPL using a HTTP URL:

$ seneca-repl http://localhost:8888/seneca-repl?id=web

By default, HTTP URLs with use web as the identifier.

REPL over AWS Lambda Invoke

To expose a REPL from an AWS Lambda function using Seneca, use the set up code for the HTTP example in the Lambda itself.

For the REPL client, you will need to install the @aws-sdk/client-lambda package, and correctly configure your AWS access using the $AWS_PROFILE environment variable.

Connect to your Lambda function REPL using:

seneca-repl "aws://lambda/FUNCTION?region=REGION&id=invoke"

where FUNCTION is the name of the Lambda function,REGION is the AWS region, such as us-east-1 (the default). The default REPL id value is invoke. NOTE: make sure to quote or escape the &.

This mechanism uses AWS Lambda invocation mechanism and thus relies on AWS for security. Special care should be taken with Lambdas that are externally exposed to prevent external requests from calling the Seneca REPL messages.

WARNING This is a security risk. Your app will need to apply additional constraints to prevent arbitrary message submission via the REPL.

This can be useful to debug build or staging systems, but is NOT recommended for production, unless used with specifically access controlled Lambda functions.

Interactive Interface

The seneca-repl command provides a convenient REPL interface including line editing and history. In remote settings you'll want to create an SSH tunnel or similar for this purpose.

Alternatively you can telnet to the port:

$ telnet localhost 30303

Replace localhost if remote with the address of the remote system.

For more comfortable experience with working cursor keys, use rlwrap

$ rlwrap telnet localhost 30303

Commands

The repl evaluates JavaScript directly:

> 1+1
2

You also have a seneca instance available:

> seneca.id
'SENECA-ID'

You can submit messages directly using jsonic format (JSON, but not strict!):

> role:seneca,cmd:stats
{
  start: '2023-08-01T17:37:39.880Z',
  act: { calls: 122, done: 121, fails: 8, cache: 0 },
  actmap: undefined,
  now: '2023-08-01T17:49:17.316Z',
  uptime: 697436
}

This is very useful for local debugging.

To access entity data, use the list$, load$, save$ and remove$ commands:

> list$ foo
[
  { entity$: '-/-/foo', ...},
  { entity$: '-/-/foo', ...},
  ...
]

These all accept the parameters:

  • entity canon (required): zone/base/name
  • query (optional): {field:value,...}

NOTE: this is a Node.js REPL, so you also get some of the features of a Node.js REPL:

  • The value of the last response is placed into the _ variable
  • You can use standard movement shortcuts like Ctrl-A, Ctrl-E, etc
  • Command history

Available Commands

  • list <pin>|plugin:
    • <pin>: list local message patterns, optionally narrowed by pin (e.g. foo:1)
    • plugin: list all plugins by full name
  • find <pin>|<plugin-name>:
    • <pin: find an exact matching message pattern definition (e.g. sys:entity,cmd:load)
    • <plugin-name>: find a plugin definition
  • list$ canon <query>: list entity data (like seneca.entity(canon).list$(query))
  • load$ canon <query>: load entity data (like seneca.entity(canon).load$(query))
  • save$ canon <data>: save entity data (like seneca.entity(canon).save$(data))
  • remove$ canon <query>: remove entity data (like seneca.entity(canon).remove$(query))
  • entity$ canon: describe an entity (like seneca.entity(canon))
  • stats: print local statistics
  • stats full: print full local statistics
  • exit or quit: exit the repl session
  • last: run last command again
  • set <path> <value>: set a seneca option, e.g: set debug.deprecation true
  • get <path>: get a seneca option
  • alias <name> <cmd>: define a new alias
  • log: toggle printing of remote log entries in test format (NOTE: these are unfiltered)
  • log match <literal>: when logging is enabled, only print lines matching the provided literal string
  • depth <number>: set depth of Util.inspect printing

History

The command history is saved to text files in a .seneca in your home folder. History is unique to each target server. You can also add additional URL parameters to isolate a server history:

$ seneca-repl localhost?project=foo # separate history for foo server
$ seneca-repl localhost?project=bar # separate history for bar server

Options

  • test : boolean false

Set plugin options when loading with:

seneca.use('repl', { name: value, ... })

Note: foo.bar in the list above means { foo: { bar: ... } }

Action Patterns

Action Descriptions

« add:cmd,sys:repl »

Add a REPL command dynamically


« echo:true,sys:repl »

No description provided.


« send:cmd,sys:repl »

No description provided.


« sys:repl,use:repl »

No description provided.


Contributing

The Senecajs org encourages open participation. If you feel you can help in any way, be it with documentation, examples, extra testing, or new features please get in touch.

Test

To run tests, simply use npm:

> npm run test

License

Copyright (c) 2015-2020, Richard Rodger and other contributors. Licensed under MIT.

Readme

Keywords

Package Sidebar

Install

npm i @seneca/repl

Weekly Downloads

762

Version

9.1.0

License

MIT

Unpacked Size

96.7 kB

Total Files

17

Last publish

Collaborators

  • lmscunha
  • alex01
  • lilsweetcaligula
  • stokesriona
  • rjrodger