server/scripts/updateSheetUrls.ts (5)

19-19: Consider making the service account connector ID configurable

The serviceAccountConnectorId is imported from a hard-coded configuration file. This reduces flexibility and requires code changes for different environments.

Consider accepting the connector ID as an environment variable or command-line argument:

-import { serviceAccountConnectorId } from "./googleConfig"
+const serviceAccountConnectorId = process.env.SERVICE_ACCOUNT_CONNECTOR_ID 
+  ? parseInt(process.env.SERVICE_ACCOUNT_CONNECTOR_ID) 
+  : (() => { throw new Error("SERVICE_ACCOUNT_CONNECTOR_ID environment variable is required") })()

---

737-741: Consider making the script timeout configurable

The hard-coded 30-minute timeout might be insufficient for large organizations with many users and spreadsheets.

-const SCRIPT_TIMEOUT = 30 * 60 * 1000 // 30 minutes
+const SCRIPT_TIMEOUT = parseInt(process.env.SCRIPT_TIMEOUT_MINUTES || '30') * 60 * 1000
+Logger.info(`Script timeout set to ${SCRIPT_TIMEOUT / 60000} minutes`)
 const timeoutId = setTimeout(() => {
   Logger.error("Script timed out after 30 minutes, forcing exit...")
   process.exit(1)
 }, SCRIPT_TIMEOUT)

---

431-431: Make the spreadsheet processing limit configurable

The hard-coded limit of 100 spreadsheets per user might be too restrictive for users with extensive Google Drive usage.

-    const sheetPromises = googleSheets.slice(0, 100).map(spreadsheet => // Limit to 100 spreadsheets
+    const maxSpreadsheetsPerUser = parseInt(process.env.MAX_SPREADSHEETS_PER_USER || '100')
+    const sheetPromises = googleSheets.slice(0, maxSpreadsheetsPerUser).map(spreadsheet =>

---

718-721: Improve database connection cleanup

The current cleanup logic has defensive checks but could be more robust.

-    if (db && db.$client && typeof db.$client.end === 'function') {
-      await db.$client.end()
-      Logger.info("Database connection closed")
-    }
+    try {
+      // Drizzle ORM typically uses $client for the underlying connection
+      if (db?.$client) {
+        if (typeof db.$client.end === 'function') {
+          await db.$client.end()
+        } else if (typeof db.$client.close === 'function') {
+          await db.$client.close()
+        } else if (typeof db.$client.destroy === 'function') {
+          await db.$client.destroy()
+        }
+        Logger.info("Database connection closed")
+      }
+    } catch (dbError) {
+      Logger.warn({ error: dbError }, "Failed to close database connection gracefully")
+    }

---

494-494: Monitor memory usage for large-scale migrations

The processedDocIds Map stores all processed document IDs in memory. For organizations with hundreds of thousands of sheets, this could consume significant memory.

Consider implementing a size limit or using a more memory-efficient deduplication strategy:

// Add after line 494
const MAX_PROCESSED_DOCS = parseInt(process.env.MAX_PROCESSED_DOCS || '1000000')
if (processedDocIds.size > MAX_PROCESSED_DOCS) {
  Logger.warn(`Processed documents map size (${processedDocIds.size}) exceeds limit. Consider running in smaller batches.`)
}

Would you like me to implement a disk-based or database-backed deduplication strategy for handling very large datasets?

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro