How to find string in file contents - sharepoint-online

Question

I'm trying to search for a string in a file that gets uploaded to SharePoint.
I'm using Send HTTP request to SharePoint to get the file content. But $content comes back in Base64 as an object. How would I search for a string in the returned $content ?
/_api/web/getfilebyserverrelativeurl('/sites/mySite/Shared%20Documents/myFileInWordOrPDFformat.docx')/$value?binaryStringResponseBody=true
headers:
{
"accept": "application/json;odata=nometadata",
"content-type": "application/json;odata=nometadata",
"odata-version": ""
}
Returns
{
"$content-type": "application/octet-stream",
"$content": "UEsDBBQAAAAIANZualN4vZl3IWQAAO2wAQARABwAd29-BlahBlahBlah-QQAAPCJAAAAAA=="
}
Posts like-
https://sharepoint.stackexchange.com/questions/273774/how-to-get-content-of-file-in-sharepoint-using-sharepoint-rest-api
https://debajmecrm.com/binary-to-base64-in-microsoft-power-automate-microsoft-flow/
https://linnzawwin.blogspot.com/2021/02/handle-base64-and-binary-file-content.html
have been helpful but I'm still not clear.
Do I need to convert the string I'm searching for to Base64 and search for it that way?
Any suggestions? Also, I don't have the option of utilizing premium connectors.
TIA!

Answer 1

This may not suit you but the way I would do it is to use the Cloudmersive document conversion actions to extract the text. It's a premium connector though so unless your user is enabled, it'll be a blocker.
It's the nicest way I found which can still be free if you don't have the need to call it more than 800 times a month and no more than once a second.
If your requirement exceeds that, you'll have to look at a paid subscription.
I looked at ways of using the SyncFusion PDF functionality in an Azure function which could be paired with the Adobe services to convert the DOCX file to a PDF but it started to become a bit too much effort for an answer in Stackoverflow. The Adobe task was easy, it was the reading of the PDF which became a little too much.
It's still an option though ... https://www.syncfusion.com/kb/7178/how-to-use-pdf-control-in-nodejs-environment
This is an example that allows you to search for text, you'd just need to adapt it using the previous link and those instructions. FYI, the reason for using Node is because C# Azure Functions only support .NET Framework in the V1 runtime which is very much obsolete ... https://learn.microsoft.com/en-us/azure/azure-functions/functions-versions ... and .NET Framework is what's required for the relevant Syncfusion functionality. I'm happy to be proven wrong by someone else in the community though, I could be wrong.
This is what the Cloudmersive connector will do for you ...
Sample Document
Flow
Result
Note: One thing I noticed is that the word "Microsoft" comes out wrong. It's a bit bizarre but you could always take that up with Cloudmersive.