I was able to reproduce the behavior you're seeing.  The complication is that, when using Retrieve Text, some whitespace characters (probably \r\n) are prepended or appended to each page.  Remove Empty Items checks the text, sees that it has characters, and doesn't remove it.  The workaround is to use a Trim function before the Remove Empty Items function to remove prepended/appended whitespace.

Is the starting token actually a multi-value token or just text with multiple lines? The Retrieve Document Text activity can generate a multi-value token split by page, but good to confirm.

The Remove Empty Items function doesn't remove blank spaces from a string of text, it only removes empty values from a multi-value.

You can use Token Calculator with the Substitute function to remove blank spaces, but if it is just text with line breaks you might need to use Pattern Matching to split it into a multi-value token, then Remove Empty Items with "all values separated by..." will do what you want.

Token Function Split on New Line - Laserfiche Answers

Shan,

You need to verify that it truly is a blank and not a line with spaces in it. Copy the line and paste it into Notepad to see if the cursor stays where it is at or if it moves.

Wesley Funderberg