Develop

src/scripts/cleanUpEntityData.js

@@ -0,0 +1,76 @@
+const MongoClient = require('mongodb').MongoClient
+require('dotenv').config()
+
+async function deleteEntitiesInBatches(mongoUrl) {
+	const batchSize = 1000 // Configurable batch size
+	const entityTypes = ['state', 'district', 'block', 'cluster', 'school']
+
+	// Validate MongoDB URL
+	if (!mongoUrl) {
+		console.error('Error: MongoDB URL must be provided as a command-line argument or in .env as MONGODB_URL')
+		process.exit(1)
+	}
+
+	let client
+
+	try {
+		// Connect to MongoDB
+		client = await MongoClient.connect(mongoUrl, { useNewUrlParser: true, useUnifiedTopology: true })
+		console.log('Connected to MongoDB')
+
+		const db = client.db() // Use default database from URL
+		const collection = db.collection('entities')
+
+		// Count total matching documents
+		const totalDocs = await collection.countDocuments({ entityType: { $in: entityTypes } })
+		console.log(`Total documents to delete: ${totalDocs}`)
+
+		if (totalDocs === 0) {
+			console.log('No documents found matching the criteria. Exiting.')
+			return
+		}
+
+		// Delete in batches
+		let deletedCount = 0
+		while (deletedCount < totalDocs) {
+			// Find a batch of document IDs to delete
+			const batchDocs = await collection
+				.find({ entityType: { $in: entityTypes } })
+				.limit(batchSize)
+				.project({ _id: 1 })
+				.toArray()
+
+			if (batchDocs.length === 0) {
+				break // No more documents to delete
+			}
+
+			// Extract IDs for deletion
+			const batchIds = batchDocs.map((doc) => doc._id)
+
+			// Delete documents by IDs
+			const batchResult = await collection.deleteMany({ _id: { $in: batchIds } })
+			const batchDeleted = batchResult.deletedCount
+			deletedCount += batchDeleted
+			console.log(`Deleted ${batchDeleted} documents in this batch. Total deleted: ${deletedCount}`)
+		}
+
+		console.log(`Deletion complete. Total documents deleted: ${deletedCount}`)
+	} catch (error) {
+		console.error('Error during deletion:', error.message)
+		process.exit(1)
+	} finally {
+		if (client) {
+			await client.close()
+			console.log('MongoDB connection closed')
+		}
+	}
+}
+
+// Get MongoDB URL from command-line argument or environment variable
+const mongoUrl = process.argv[2] || process.env.MONGODB_URL
+
+// Run the script
+deleteEntitiesInBatches(mongoUrl).catch((error) => {
+	console.error('Script failed:', error.message)
+	process.exit(1)
+})

src/scripts/readme

@@ -0,0 +1,33 @@
+# MongoDB Entity Data Cleanup Script
+
+This Node.js script deletes all documents from the MongoDB `entities` collection where `entityType` is one of `state`, `district`, `block`, `cluster`, or `school`. It processes deletions in batches for efficiency and supports MongoDB 4.x with the MongoDB Node.js driver v3.x. The script accepts a MongoDB URL as a command-line argument or falls back to an environment variable.
+
+## Prerequisites
+
+-   **Node.js**: Version 14 or later.
+-   **MongoDB Server**: Version 4.x (e.g., 4.0 or 4.2).
+-   **NPM Packages**: `mongodb@3.6.12`, `dotenv`.
+-   **MongoDB URL**: A valid connection string (e.g., `mongodb://localhost:27017/elevate-entity`).
+
+## Usage
+
+The script can be executed **outside** or **inside** a Docker container. It deletes documents in batches of 1000 (configurable) and logs progress.
+
+### Option 1: Run Outside Docker
+
+1. Save the script as `cleanUpEntityData.js` (provided in the repository or separately).
+
+2. Run the script with a MongoDB URL as a command-line argument:
+    ```bash
+    node cleanUpEntityData.js mongodb://localhost:27017/prod-saas-elevate-entity
+    ```
+
+### Option 2: Run Inside Docker
+
+1. Save the script as `cleanUpEntityData.js` inside entity service docker container
+
+2. Run the container with the MongoDB URL as an argument:
+
+    ```bash
+    node cleanUpEntityData.js
+    ```