Bugzilla #42109:

Fixed the rendering of zero-width spaces for certain fonts by not generating them into the area tree.
Submitted by: Paul Vinkenoog <paul.at.vinkenoog.nl>

git-svn-id: https://svn.apache.org/repos/asf/xmlgraphics/fop/trunk@540049 13f79535-47bb-0310-9956-ffa450edef68

src/java/org/apache/fop/layoutmgr/inline/CharacterLayoutManager.java

@@ -75,11 +75,14 @@ public class CharacterLayoutManager extends LeafNodeLayoutManager {
     }
 
     private org.apache.fop.area.inline.TextArea getCharacterInlineArea(Character node) {
-        org.apache.fop.area.inline.TextArea text 
+        org.apache.fop.area.inline.TextArea text
             = new org.apache.fop.area.inline.TextArea();
         char ch = node.getCharacter();
         if (CharUtilities.isAnySpace(ch)) {
-            text.addSpace(ch, 0, CharUtilities.isAdjustableSpace(ch));
+            // add space unless it's zero-width:
+            if (!CharUtilities.isZeroWidthSpace(ch)) {
+                text.addSpace(ch, 0, CharUtilities.isAdjustableSpace(ch));
+            }
         } else {
             text.addWord(String.valueOf(ch), 0);
         }

src/java/org/apache/fop/layoutmgr/inline/TextLayoutManager.java

@@ -23,7 +23,6 @@ import java.util.ArrayList;
 import java.util.List;
 import java.util.LinkedList;
 import java.util.ListIterator;
-import java.util.NoSuchElementException;
 
 import org.apache.commons.logging.Log;
 import org.apache.commons.logging.LogFactory;
@@ -378,7 +377,7 @@ public class TextLayoutManager extends LeafNodeLayoutManager {
         iTotalAdjust += (iWordSpaceDim - wordSpaceIPD.opt) * iWScount;
         if (iTotalAdjust != iDifference) {
             // the applied adjustment is greater or smaller than the needed one
-            log.trace("TextLM.addAreas: error in word / letter space adjustment = " 
+            log.trace("TextLM.addAreas: error in word / letter space adjustment = "
                     + (iTotalAdjust - iDifference));
             // set iTotalAdjust = iDifference, so that the width of the TextArea
             // will counterbalance the error and the other inline areas will be
@@ -458,11 +457,13 @@ public class TextLayoutManager extends LeafNodeLayoutManager {
             areaInfo = (AreaInfo) vecAreaInfo.get(i);
             if (areaInfo.isSpace) {
                 // areaInfo stores information about spaces
-                // add the spaces to the TextArea
+                // add the spaces - except zero-width spaces - to the TextArea
                 for (int j = areaInfo.iStartIndex; j < areaInfo.iBreakIndex; j++) {
                     char spaceChar = textArray[j];
-                    textArea.addSpace(spaceChar, 0, 
-                            CharUtilities.isAdjustableSpace(spaceChar));
+                    if (!CharUtilities.isZeroWidthSpace(spaceChar)) {
+                        textArea.addSpace(spaceChar, 0,
+                                CharUtilities.isAdjustableSpace(spaceChar));
+                    }
                 }
             } else {
                 // areaInfo stores information about a word fragment

src/java/org/apache/fop/util/CharUtilities.java

@@ -64,8 +64,8 @@ public class CharUtilities {
     public static final char ZERO_WIDTH_NOBREAK_SPACE = '\uFEFF';
     /** soft hyphen */
     public static final char SOFT_HYPHEN = '\u00AD';
-    
-    
+
+
     /**
      * Utility class: Constructor prevents instantiating when subclassed.
      */
@@ -98,7 +98,17 @@ public class CharUtilities {
     public static boolean isBreakableSpace(char c) {
         return (c == SPACE || isFixedWidthSpace(c));
     }
-    
+
+    /**
+     * Method to determine if the character is a zero-width space.
+     * @param c the character to check
+     * @return true if the character is a zero-width space
+     */
+    public static boolean isZeroWidthSpace(char c) {
+        return c == ZERO_WIDTH_SPACE           // 200Bh
+            || c == ZERO_WIDTH_NOBREAK_SPACE;  // FEFFh (also used as BOM)
+    }
+
     /**
      * Method to determine if the character is a (breakable) fixed-width space.
      * @param c the character to check
@@ -111,7 +121,7 @@ public class CharUtilities {
 //      c == '\u2002'                   // en space
 //      c == '\u2003'                   // em space
 //      c == '\u2004'                   // three-per-em space
-//      c == '\u2005'                   // four--per-em space
+//      c == '\u2005'                   // four-per-em space
 //      c == '\u2006'                   // six-per-em space
 //      c == '\u2007'                   // figure space
 //      c == '\u2008'                   // punctuation space
@@ -120,7 +130,7 @@ public class CharUtilities {
 //      c == '\u200B'                   // zero width space
 //      c == '\u3000'                   // ideographic space
     }
-    
+
     /**
      * Method to determine if the character is a nonbreaking
      * space.

status.xml

@@ -28,6 +28,10 @@
 
   <changes>
     <release version="FOP Trunk">
+      <action context="Code" dev="JM" type="fix" fixes-bug="42109" due-to="Paul Vinkenoog">
+        Fixed the rendering of zero-width spaces for certain fonts by not generating them into
+        the area tree.
+      </action>
       <action context="Code" dev="LF" type="fix">
         Fixed a problem with disappearing footnotes inside hyphenated inlines (and footnotes with hyphenated inline child).
       </action>

test/layoutengine/standard-testcases/block_white-space_4.xml

@@ -164,8 +164,8 @@
     <true xpath="//block[@prod-id='l-hair-space']/lineArea/text/*[6]/@adj = 'false'"/>
 
     <eval expected="55368" xpath="//block[@prod-id='l-zwsp']/lineArea/text/@ipd"/>
-    <true xpath="//block[@prod-id='l-zwsp']/lineArea/text/*[2]/@adj = 'false'"/>
-    <true xpath="//block[@prod-id='l-zwsp']/lineArea/text/*[6]/@adj = 'false'"/>
+    <!-- zwsp is not rendered to the area tree -->
+    <eval expected="1" xpath="count(//block[@prod-id='l-zwsp']/lineArea/text/space)"/>
 
     <eval expected="55368" xpath="//block[@prod-id='l-nosp']/lineArea/text/@ipd"/>
 
@@ -232,9 +232,9 @@
     <true xpath="//block[@prod-id='c-hair-space']/lineArea/text/*[6]/@adj = 'false'"/>
 
     <eval expected="55368" xpath="//block[@prod-id='c-zwsp']/lineArea/text/@ipd"/>
-    <true xpath="//block[@prod-id='c-zwsp']/lineArea/text/*[2]/@adj = 'false'"/>
-    <true xpath="//block[@prod-id='c-zwsp']/lineArea/text/*[6]/@adj = 'false'"/>
-
+    <!-- zwsp is not rendered to the area tree -->
+    <eval expected="1" xpath="count(//block[@prod-id='c-zwsp']/lineArea/text/space)"/>
+    
     <eval expected="55368" xpath="//block[@prod-id='c-nosp']/lineArea/text/@ipd"/>
 
     <eval expected="190804" xpath="//block[@prod-id='j-sp']/lineArea/text/@ipd"/>
@@ -313,9 +313,9 @@
 
     <eval expected="84280" xpath="//block[@prod-id='j-zwsp']/lineArea/text/@ipd"/>
     <eval expected="28912" xpath="//block[@prod-id='j-zwsp']/lineArea/text/@twsadjust"/>
-    <true xpath="//block[@prod-id='j-zwsp']/lineArea/text/*[2]/@adj = 'false'"/>
-    <true xpath="//block[@prod-id='j-zwsp']/lineArea/text/*[6]/@adj = 'false'"/>
-
+    <!-- zwsp is not rendered to the area tree -->
+    <eval expected="1" xpath="count(//block[@prod-id='j-zwsp']/lineArea/text/space)"/>
+    
     <eval expected="134904" xpath="//block[@prod-id='j-nosp']/lineArea/text/@ipd"/>
     <eval expected="79536" xpath="//block[@prod-id='j-nosp']/lineArea/text/@twsadjust"/>
 
@@ -408,9 +408,9 @@
     <eval expected="85480" xpath="//block[@prod-id='lsj-zwsp']/lineArea/text/@ipd"/>
     <eval expected="14112" xpath="//block[@prod-id='lsj-zwsp']/lineArea/text/@twsadjust"/>
     <eval expected="2000" xpath="//block[@prod-id='lsj-zwsp']/lineArea/text/@tlsadjust"/>
-    <true xpath="//block[@prod-id='lsj-zwsp']/lineArea/text/*[2]/@adj = 'false'"/>
-    <true xpath="//block[@prod-id='lsj-zwsp']/lineArea/text/*[6]/@adj = 'false'"/>
-
+    <!-- zwsp is not rendered to the area tree -->
+    <eval expected="1" xpath="count(//block[@prod-id='lsj-zwsp']/lineArea/text/space)"/>
+    
     <eval expected="140237" xpath="//block[@prod-id='lsj-nosp']/lineArea/text/@ipd"/>
     <eval expected="64869" xpath="//block[@prod-id='lsj-nosp']/lineArea/text/@twsadjust"/>
     <eval expected="2000" xpath="//block[@prod-id='lsj-nosp']/lineArea/text/@tlsadjust"/>