Required changes
From Apache OpenOffice Wiki
GlyphProcessor is quite complex and makes PDFIProcessor untidy. The main reason of moving glyph processing to separate class was to avoid mess and make the code more scalable. Moreover some additional changes in function like optimizeText from drawprocessing had to be changes.
Solution
GlyphProcessor and associated class have been moved to files glyphprocessor.hxx and glyphprocessor.cxx. Moreover function optimixeText was prepared to work with ParagraphLineElement instead of ParagraphElement.