Huge News!Announcing our $40M Series B led by Abstract Ventures.Learn More
Socket
Sign inDemoInstall
Socket

@epam/ai-dial-chat-visualizer-connector

Package Overview
Dependencies
Maintainers
0
Versions
626
Alerts
File Explorer

Advanced tools

Socket logo

Install Socket

Detect and block malicious and high-risk dependencies

Install

@epam/ai-dial-chat-visualizer-connector

Package for custom visualizer to connect with Dial Chat

  • 0.22.0
  • latest
  • Source
  • npm
  • Socket score

Version published
Maintainers
0
Created
Source

DIAL Chat Visualizer Connector

DIAL Chat Visualizer Connector is a library for connecting custom visualizers - applications which could visualize some special type data (for example plot data for the Plotly).

Public classes to use

ChatVisualizerConnector - class which provides needed methods for the Visualizer(rendered in the iframe) to interact with DIAL Chat (receive data to visualize).

Prerequisites

For security reason your Dial Chat application should configure sources where your custom visualizers hosted:

  • ALLOWED_IFRAME_SOURCES - list of allowed iframe sources in <source> <source> format.

Note: For development purposes you can set *

ALLOWED_IFRAME_SOURCES=http://localhost:8000

Moreover, it needs to be configured some Visualizer properties:

  • CUSTOM_VISUALIZERS - list of the objects with custom visualizers properties. This properties are : { title, description, icon, contentType, url }.
interface CustomVisualizer {
  title: string;
  description: string;
  icon: string;
  contentType: string;
  url: string;
}
CUSTOM_VISUALIZERS=[
                    {
                      "title":"CUSTOM_VISUALIZER", // Visualizer title
                      "description": "CUSTOM VISUALIZER to render images", // Short description for the Visualizer
                      "icon":"data:image/svg+xml;base64,some-base64-image", // Icon for the Visualizer
                      "contentType":"image/png,image/jpg", // List of MIME types that Visualizer could render separated by ","
                      "url":"http://localhost:8000" // Visualizer host
                    },
                    {
                      //Other Visualizer
                    }

                  ]

Futhermore, model or application should send data in json-like format which should include layout property. This layout object should have width and height properties. All other properties could be set to anything you need for your Visualizer.

export interface CustomVisualizerDataLayout {
  width: number;
  height: number;
}

export interface CustomVisualizerData {
  layout: CustomVisualizerDataLayout;
}

To integrate Visualizer with DIAL CHAT.

  1. Install library
npm i @epam/ai-dial-chat-visualizer-connector
  1. Add file to serving folder in your application or just import it in code
import { AttachmentData, ChatVisualizerConnector, CustomVisualizerDataLayout } from '@epam/ai-dial-chat-visualizer-connector';

ChatVisualizerConnector - class which provides needed methods for the Visualizer(rendered in the iframe) to interact with DIAL Chat (receive data to visualize). Expects following required arguments:

/**
 * Params for a ChatVisualizerConnector
 * @param dialHost {string} DIAL CHAT host
 * @param appName {string} name of the Visualizer same as in config
 * @param dataCallback {(visualizerData: AttachmentData) => void} callback to get data that will be used in the Visualizer
 */

//instance example
new ChatVisualizerConnector(dialHost, appName, setData);

AttachmentData - interface for the payload you will get from the DIAL CHAT

export interface AttachmentData {
  mimeType: string;
  visualizerData: CustomVisualizerData;
}

CustomVisualizerDataLayout - interface for the layout you will get from the DIAL CHAT. Properties width and height is needed for the proper rendering in the DIAL CHAT.

export interface CustomVisualizerDataLayout {
  width: number;
  height: number;
}
  1. Set dialHost to the DIAL CHAT host you want to connect:
const dialHost = 'https://hosted-dial-chat-domain.com';
  1. Set appName same as title in the DIAL CHAT configuration in the CUSTOM_VISUALIZERS environmental variable:
const appName = 'CUSTOM_VISUALIZER';
  1. Create an instance of ChatVisualizerConnector in your code.

ChatVisualizerConnector:

//Here you store your data which you get from the DIAL CHAT
const [data, setData] = useState<AttachmentData>();

const chatVisualizerConnector = useRef<ChatVisualizerConnector | null>(null);

useEffect(() => {
  if (!chatVisualizerConnector.current && dialHost && appName) {
    chatVisualizerConnector.current = new ChatVisualizerConnector(dialHost, appName, setData);

    return () => {
      chatVisualizerConnector.current?.destroy();
      chatVisualizerConnector.current = null;
    };
  }
}, [appName, dialHost]);
  1. Send 'READY' event via sendReady() to the DIAL CHAT to inform that your Visualizer is ready (this action will hide loader). Then you could do some preparation (login, etc.) and, after that, send 'READY TO INTERACT' event via sendReadyToInteract() to inform DIAL CHAT that Visualizer is ready to receive data.
useEffect(() => {
  if (appName && dialHost) {
    chatVisualizerConnector.current?.sendReady();
    //Make some actions if needed
    chatVisualizerConnector.current?.sendReadyToInteract();
  }
}, [dialHost, appName]);
  1. Make needed type assertion for the data from the DIAL CHAT

Note: Data send by model/application from DIAL CHAT should be the same type as you expect.

//layout should include width and height properties
interface YourVisualizerLayout extends CustomVisualizerDataLayout {
  //any other layout properties expected from the model/application output
}
data.visualizerData as { dataToRender: string; layout: YourVisualizerLayout };
  1. Render data in your Visualizer;
<div>{typedVisualizerData.dataToRender}</div>

Full React code example to connect your custom visualizer:

Module.tsx


import { AttachmentData, ChatVisualizerConnector, CustomVisualizerDataLayout } from '@epam/ai-dial-chat-visualizer-connector';


export const Module: FC = () => {
  const [data, setData] = useState<AttachmentData>();

  const chatVisualizerConnector = useRef<ChatVisualizerConnector | null>(
    null
  );

  //DIAL CHAT host
  const dialHost = 'https://hosted-dial-chat-domain.com';

  //Visualizer title. Should be same as in the DIAL CHAT configuration in CUSTOM_VISUALIZERS
  const appName = 'CUSTOM_VISUALIZER';

  useEffect(() => {
    if (!chatVisualizerConnector.current && dialHost && appName) {
      chatVisualizerConnector.current = new ChatVisualizerConnector(
        dialHost,
        appName,
        setData
      );

      return () => {
        chatVisualizerConnector.current?.destroy();
        chatVisualizerConnector.current = null;
      };
    }
  }, [appName, dialHost]);

  useEffect(() => {
    if (appName && dialHost) {
      chatVisualizerConnector.current?.sendReady();
      chatVisualizerConnector.current?.sendReadyToInteract();
    }
  }, [dialHost, appName]);

  const typedVisualizerData = useMemo(() => {
    return (
      data?.visualizerData && (data.visualizerData as unknown as { dataToRender: string; layout: YourVisualizerLayout })
    );
  }, [data?.visualizerData]);

  return (
    <div>
      {!!typedVisualizerData?.dataToRender && (
          <div>
            {typedVisualizerData.dataToRender}
          </div>
      )}
    </div>
  );
};

index.ts

//...other imports
import { Module } from "./Module.tsx";

const root = createRoot(document.getElementById("root"));
root.render(<Module />);

FAQs

Package last updated on 18 Dec 2024

Did you know?

Socket

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install

Related posts

SocketSocket SOC 2 Logo

Product

  • Package Alerts
  • Integrations
  • Docs
  • Pricing
  • FAQ
  • Roadmap
  • Changelog

Packages

npm

Stay in touch

Get open source security insights delivered straight into your inbox.


  • Terms
  • Privacy
  • Security

Made with ⚡️ by Socket Inc