I'm using Tess4j to prototype an OCR tool. If I do OCR with traineddata set to kor, there will be a space between the characters. I know it can be controlled by using '-c preserve_interword_spaces' on the command. How can I apply this option in tess4j?