The forum poster noticed that the extracted text in a PDF document was missing some punctuation. I dog a little deeper and found that we where correctly apply the character encoding if a toUnicode cmap does not exist.

The following is a quick patch for the problem but will not be the final fix. I will update the interface code to make sure we handle this in a more generic maner.

Step One.
Create a new method in org.icepdf.core.pobjects.fonts.ofont.Ofont.java
public char toUnicode(char c1) {
char c = toUnicode==null?getCharDiff(c1):c1;
c = getCMapping(c);

if (!awtFont.canDisplay(c)) {
    c |= 0xF000;
}
if (!awtFont.canDisplay(c)) {
    c = findAlternateSymbol(c);
}
return c;
}

Step 2
Call the new toUnicode method from the content parser. Just after the 'charValue' is defined in drawString(..) ~ line 2160 add the following code.

if (textState.currentfont instanceof org.icepdf.core.pobjects.fonts.ofont.OFont){
charValue = ((org.icepdf.core.pobjects.fonts.ofont.OFont)
        textState.currentfont).toUnicode(unmodifiedDisplayText.charAt(i));
}

Once again this is not an official patch just a work around.