TextSegments Issue

I am currently facing an issue with PDF files.

The scenario is as such:

we are currently evaluating ASPOSE Pdf  to do a post process on PDF files and we have tagged some words such as |myword^, where | is symbolyzing the start of the word and the ^ a tag that will be replaced by a calculated supertscript number. The problem is sometimes the fragment is splitted into 3 textSegments (|,myword and ^) sometimes not and consider |myword^ as a single TextSegment. 

So how can I be sure to have always 3 segments?

Thanks a lot for the help.