Skip to content

MertenD/node-crawler

Repository files navigation

Project Name: Node-Crawler

Next.js ReactFlow

Overview

! Work in progress !

See https://mertend.github.io/node-crawler/ for a detailed documentation.

Node-Crawler is a highly customizable, Node-based web application for creating web crawlers and further processing and transforming the retrieved data. Users can build tailor-made web crawlers and manipulate and transform the collected data as needed. The output data format can be set to a wide range of formats including JSON, CSV, and various database formats.

The app is developed using Next.js and the React Flow library provides a framework for creating the node-based editor.

Overview Overview

Features

  • Node-based Editing: Users can create and edit their own crawler workflows by drag-and-dropping nodes.
  • Data Transformation: The application supports a variety of data manipulation and transformation operations for cleaning and restructuring the gathered data.
  • Data Export: The transformed data can be output in a variety of formats including JSON, CSV, and various database formats.

Installation

Make sure you have Node.js and npm installed on your system before you start.

  1. Clone the repository:
git clone https://github.com/MertenD/node-crawler.git
  1. Navigate into the directory and install the dependencies:
cd node-crawler
npm install
  1. Start the development server:
npm run dev

Now you should be able to see the web application on http://localhost:3000 in your browser.

Implementing a New Node

Follow these steps when you need to create a new Node:

Step 1: Add the Node to the NodeType Enum

Next, add your newly created Node to the config/NodeType enum.

Step 2: Create the New Node

First, create a new file in the components/editor/pages/canvas/nodes directory. In this file, you will define the following elements:

  1. Data Interface: Create a Data interface that stores all data the user can configure.

  2. Style Function: Create a Style function (using the createNodeShapeStyle() function), where you can customize the Node's appearance.

  3. Node Component: Create a Node component (using the createNodeComponent() function), which will be the Node on the canvas.

  4. Options Component: Create an Options component (using the createOptionsComponent() function), where the user can configure the Node's behavior.

You can use the following template to create a new Node:

// TODO: Replace [NAME] everywhere

// --- Data ---
export interface [NAME]NodeData extends NodeData {
    // TODO: Add data attributes here
}

// --- Style ---
export const [NAME]ShapeStyle = createNodeShapeStyle({
    // TODO: Add additional CSS for the node's shape here
})

// --- Node ---
export const [NAME]Node = createNodeComponent<[NAME]NodeData>(
    NodeType.[NAME]_NODE,
    [NAME]ShapeStyle,
    (id, selected, data) => {
        // TODO: Place the node content here
    }
)

// --- Options ---
export const [NAME]Options = createOptionsComponent<[NAME]NodeData>("Start", ({ id, data, onDataUpdated }) => {
    return // TODO: Place options here
})

Step 3: Add Metadata to NodesMetadata.tsx

Add all metadata of the new Node to the config/NodesMetadata.tsx file.

Step 4: Add Connection Rules

Define the connection rules for the new Node in the config/ConnectionRules.ts file.

Step 5: Create a New Node Class for Execution

You need to create a new class in the engine/nodes directory. This class should extend the BasicNode interface. Below is a basic template for your reference:

// TODO: Replace [NAME]
export class Engine[NAME]Node implements BasicNode {
    id: string;
    nodeType: NodeType
    data: // TODO

    constructor(id: string, data: /* TODO */) {
        this.id = id
        this.nodeType = // TODO
        this.data = data
    }

    async run() {
        // Optional: Get inputs from previous nodes
        const input = usePlayStore.getState().getInput(this.id, "input")

        if (input) {
            // TODO Put the logic of the node here
            
            // Optional: Add downloadable file
            usePlayStore.getState().addFile(/* TODO */)
            
            // Optional: Make outputs accessable for the next node 
            usePlayStore.getState().addOutgoingPipelines(this.id, /* TODO */)
            
            // Optional: Write to the log
            usePlayStore.getState().writeToLog(/* TODO */)
            
            // End with calling the next node
            usePlayStore.getState().nextNode()
        }
    }
}

Step 6: Add Node Transformation Logic

The final step involves adding the transformation logic for the node. This transformation will convert a React Flow Node into an instance of your newly created class from Step 5. To do this, navigate to the util/NodeMapTransformer.ts file and add a new case to the getNodeFromType() method where you create the instance.

About

Node-Crawler is a highly customizable, Node-based web application for creating web crawlers and further processing and transforming the retrieved data.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages