javadoc now uses a sentence BreakIterator to find the end of
the first sentence to use for the summary. Some common
constructs cause it to break too soon. For example:

CASE #1 --------------------------------------------------
import java.text.BreakIterator;

public class SentenceBug {
    public static void main(String[] argv) {
        BreakIterator bi = BreakIterator.getSentenceInstance();
        String test = "Test <code>Flags.Flag</code> class. Another test.";
        bi.setText(test);
        System.out.println(test.substring(bi.first(), bi.next()));
        System.exit(0);
    }
}

This prints "Test <code>Flags."
A period followed by a capital letter should not be a sentence boundary;
there should be whitespace between them.

CASE #2 --------------------------------------------------
import java.text.BreakIterator;

public class SentenceBug2 {
    public static void main(String[] argv) {
        BreakIterator bi = BreakIterator.getSentenceInstance();
        String test = "<P>Provides a set of &quot;lightweight&quot; (all-Java<FONT SIZE=\"-2\"><SUP>TM</SUP></FONT> language) components that, to the maximum degree possible, work the same on all platforms. Another test.";
        bi.setText(test);
        System.out.println(test.substring(bi.first(), bi.next()));
        System.exit(0);
    }
}

This prints:

   <P>Provides a set of &quot;lightweight&quot; (all-Java<FONT SIZE="-2"

Notice that it stops between the double quote (") and greater-than symbol (>).
There is no period, exclamation mark or question mark anywhere near.
                  
----
For sample files, see /java/web/docs/bugs/javadoc-bugs/bug4158381-breakiterator
doug.kramer@Eng 1998-09-17