fix NPE in XML writing, restore MAB hack for subfield 'a' in control field

gradle.properties

@@ -1,5 +1,5 @@
 group = org.xbib
 name = marc
-version = 2.9.11
+version = 2.9.13
 
 org.gradle.warning.mode = ALL

src/main/java/org/xbib/marc/MarcField.java

@@ -714,6 +714,16 @@ public class MarcField implements Comparable<MarcField> {
         }
     }
 
+    /**
+     * Helper method to create a subfield.
+     * @param id the subfield ID
+     * @param value the subfeild value
+     * @return the subfield
+     */
+    public static Subfield subfield(String id, String value) {
+        return new Subfield(id, value);
+    }
+
     /**
      * MARC subfield. A subfield consists of an ID and a value.
      */

src/main/java/org/xbib/marc/MarcRecord.java

@@ -159,6 +159,12 @@ public class MarcRecord implements Map<String, Object> {
         return recordLabel;
     }
 
+    public void filterFields(Predicate<? super MarcField> predicate) {
+        if (marcFields != null) {
+            marcFields = marcFields.stream().filter(predicate).toList();
+        }
+    }
+
     /**
      * Return the MARC fields of this record.
      *
@@ -376,7 +382,7 @@ public class MarcRecord implements Map<String, Object> {
         this.delegate = createMapFromMarcFields(Comparator.naturalOrder());
     }
 
-    public void rebuildFields(Comparator<MarcField> comparator) {
+    public void sortFields(Comparator<MarcField> comparator) {
         this.marcFields.sort(comparator);
     }
 

src/main/java/org/xbib/marc/xml/MarcXchangeWriter.java

@@ -346,21 +346,22 @@ public class MarcXchangeWriter extends MarcContentHandler implements Flushable,
                 return;
             }
             if (field.isControl()) {
-                Iterator<Attribute> attrs =
-                        Collections.singletonList(eventFactory.createAttribute(TAG_ATTRIBUTE,
-                                transform(field.getTag()))).iterator();
-                xmlEventConsumer.add(eventFactory.createStartElement(getControlfieldElement(), attrs, namespaces));
                 String value = field.getValue();
-                if (value != null && !value.isEmpty()) {
+                if (value == null || value.isEmpty()) {
-                    xmlEventConsumer.add(eventFactory.createCharacters(transform(value)));
-                } else {
                     // the control field is disguised as a data field, try lookup value in first subfield of " "
                     value = field.getFirstSubfieldValue(" ");
-                    if (value != null && !value.isEmpty()) {
+                    // still no value? Then it is some exotic like MAB with subfield "a"?
-                        xmlEventConsumer.add(eventFactory.createCharacters(transform(value)));
+                    if (value == null || value.isEmpty()) {
+                        value = field.getFirstSubfieldValue("a");
                     }
                 }
-                xmlEventConsumer.add(eventFactory.createEndElement(getControlfieldElement(), namespaces));
+                if (value != null && !value.isEmpty()) {
+                    Iterator<Attribute> attrs = Collections.singletonList(eventFactory.createAttribute(TAG_ATTRIBUTE,
+                            transform(field.getTag()))).iterator();
+                    xmlEventConsumer.add(eventFactory.createStartElement(getControlfieldElement(), attrs, namespaces));
+                    xmlEventConsumer.add(eventFactory.createCharacters(transform(value)));
+                    xmlEventConsumer.add(eventFactory.createEndElement(getControlfieldElement(), namespaces));
+                }
             } else if (!field.isEmpty()) {
                 String tag = field.getTag();
                 String indicator = field.getIndicator();
@@ -372,6 +373,12 @@ public class MarcXchangeWriter extends MarcContentHandler implements Flushable,
                 attrs.add(eventFactory.createAttribute(IND_ATTRIBUTE + "2", transform(ind2)));
                 xmlEventConsumer.add(eventFactory.createStartElement(getDatafieldElement(), attrs.iterator(), namespaces));
                 for (MarcField.Subfield subfield : field.getSubfields()) {
+                    String value = subfield.getValue();
+                    // we skip null values because XML event consumer will fail on length() with NPE.
+                    // we do not skip empty values because of subfield ID transport in MAB or UNIMARC.
+                    if (value == null) {
+                        continue;
+                    }
                     String code = subfield.getId();
                     // From https://www.loc.gov/standards/iso25577/ISO_DIS_25577_2(E)070727.doc
                     // "There is one restriction. A special mode (identifier length = 0) of ISO 2709 operates with
@@ -385,7 +392,7 @@ public class MarcXchangeWriter extends MarcContentHandler implements Flushable,
                     subfieldattrs.add(eventFactory.createAttribute(CODE_ATTRIBUTE, transform(code)));
                     xmlEventConsumer.add(eventFactory.createStartElement(getSubfieldElement(),
                             subfieldattrs.iterator(), namespaces));
-                    xmlEventConsumer.add(eventFactory.createCharacters(transform(subfield.getValue())));
+                    xmlEventConsumer.add(eventFactory.createCharacters(transform(value)));
                     xmlEventConsumer.add(eventFactory.createEndElement(getSubfieldElement(), namespaces));
                 }
                 String value = field.getValue();

src/test/java/org/xbib/marc/dialects/mab/MabTest.java

@@ -26,6 +26,9 @@ import java.io.ByteArrayOutputStream;
 import java.io.IOException;
 import java.io.OutputStream;
 import java.io.OutputStreamWriter;
+import java.io.Writer;
+import java.nio.file.Files;
+import java.nio.file.Paths;
 import org.junit.jupiter.api.Test;
 import org.xbib.marc.Marc;
 import org.xbib.marc.MarcField;

src/test/resources/org/xbib/marc/dialects/mab/DE-605-aleph-publish.xml.mapped.xml

@@ -7,8 +7,8 @@
     <datafield ind2=" " ind1=" " tag="FMT">
       <subfield code=" ">MH</subfield>
     </datafield>
-    <controlfield tag="001"></controlfield>
+    <controlfield tag="001">HT004330420</controlfield>
-    <controlfield tag="002"></controlfield>
+    <controlfield tag="002">19921110</controlfield>
     <datafield ind2=" " ind1=" " tag="030">
       <subfield code=" ">a|1uc||||||17</subfield>
     </datafield>