I look at the arail.ttf.xml file which org.apache.fop.fonts.apps.TTFReader makes it I found that some charachter's unicode does'nt exist in it for example arabic lowercase charachres this leads to showing lowercase chars same as uppercase
Characters without a glyph in the font would show as a sharp sign ("#"). The PDF serializer doesn't substitute characters on its own.
J.Pietschmann