We're sending PDFs via a web app to the adobe API to be converted to word doc. These are PDFs with sensitive data and cannot be written to disk.
I was browsing through our web app logs today and STUNNED to see this:
Can someone expand on this? How can we turn this off?
These are temporary and should be deleted immediately afterwards - for Node for example we document this here, https://opensource.adobe.com/pdfservices-node-sdk-samples/apidocs/latest/FileRef.html:
"When a FileRef instance is created by this SDK while referring to a temporary file location, calling any of the methods to save the fileRef (For example, FileRef#writeToStream, FileRef#saveAsFile etc.) will delete the temporary file."
As far as I know, when a user uploads to a web app, there's always at least _some_ local copy of the file in temporary space before it's handled.
Our web app is an API and things are sent in encoded as base64. There's no record on disk until the adobe dll write it there.
What if on some server setups the web app server isn't given access to write to temporary storage? Does your dll fail?
If the process couldn't write to temp, then it would fail, so nothing would be stored. And again, it should be cleaned up automatically and not persist. It's not going to be web accessible so that too shouldn't be an issue.
You can use rest api instead of using sdk to avoid saving file to disk storage