Updated details on getAscender bug

git-svn-id: https://svn.apache.org/repos/asf/xmlgraphics/fop/trunk@197461 13f79535-47bb-0310-9956-ffa450edef68

diff --git a/src/java/org/apache/fop/render/awt/AWTFontMetrics.java b/src/java/org/apache/fop/render/awt/AWTFontMetrics.java
index 4d0ec8d08635ec469c827ddedba3336e79ab5e45..fdfd8b826bdcbeb6daf8a5d903e3b6ff5a495cf0 100644 (file)
--- a/src/java/org/apache/fop/render/awt/AWTFontMetrics.java
+++ b/src/java/org/apache/fop/render/awt/AWTFontMetrics.java
@@ -115,6 +115,23 @@ public class AWTFontMetrics {
 
         // workaround for sun bug on FontMetrics.getAscent()
         // http://developer.java.sun.com/developer/bugParade/bugs/4399887.html
+        /*
+         * Bug 4399887 has status Closed, not a bug.  The comments on the bug
+         * are:
+         * The submitter is incorrectly assuming that the string he has used
+         * is displaying characters which represent those with the maximum
+         * ascent in the font. If (for example) the unicode character
+         * \u00c1 which is the A-acute character used in many European
+         * languages is placed in the bodu of the "Wrong" string it can be
+         * seen that the JDK is exactly correct in its determination of the
+         * ascent of the font.
+         * If the bounds of a particular string are interesting then the
+         * Rectangle FontMetrics.getStringBounds(..) method can be called.
+         * The y value of the rectangle is the offset from the origin
+         * (baseline) apparently needed by the sample test program
+         * 
+         * xxxxx@xxxxx 2001-05-15
+         */
         int realAscent = fmt.getAscent()
                          - (fmt.getDescent() + fmt.getLeading());
         return FONT_FACTOR * realAscent;