Fixed #33

Author: cboehme

src/main/java/org/culturegraph/mf/stream/converter/RegexDecoder.java

@@ -76,7 +76,7 @@
  * </blockquote>
  * </p>
  * 
- * @author Thomas Seidel
+ * @author Thomas Seidel, Christoph Böhme
  * 
  */
 @Description("Decodes an incoming string based on a regular expression using named-capturing groups")
@@ -122,25 +122,32 @@ public void setDefaultLiteralName(final String defaultLiteralName) {
 	}
 
 	@Override
-	public void process(final String string) {
+	public void process(final String str) {
 		assert !isClosed();
-		matcher.reset(string);
+		
+		matcher.reset(str);
+		if (!matcher.find()) {
+			LOG.info("Ignoring non-matching input: {}", str);
+			return;
+		}
 
+		// Extract the record id:
 		final String id;
 		final int groupIndex = captureGroupNames.indexOf(ID_CAPTURE_GROUP) + 1;
-		if (groupIndex > 0 && matcher.find()) {
+		if (groupIndex > 0) {
 			id = matcher.group(groupIndex);
 		} else {
 			id = "";
 		}
 		getReceiver().startRecord(id);
 
+		// Add a literal containing the unmodified input string:
 		if (defaultLiteralName != null) {
-			getReceiver().literal(defaultLiteralName, string);
+			getReceiver().literal(defaultLiteralName, str);
 		}
 
-		matcher.reset();
-		while (matcher.find()) {
+		// Emit literals:
+		do {
 			final int groupCount = matcher.groupCount();
 			LOG.debug("groupCount() is: {}", Integer.valueOf(groupCount));
 			for (int group = 1; group <= groupCount; ++group) {
@@ -150,7 +157,7 @@ public void process(final String string) {
 						Integer.valueOf(group), literalName, literalValue);
 				getReceiver().literal(literalName, literalValue);
 			}
-		}
+		} while (matcher.find());
 
 		getReceiver().endRecord();
 	}

src/test/java/org/culturegraph/mf/stream/converter/RegexDecoderTest.java

@@ -15,18 +15,18 @@
  */
 package org.culturegraph.mf.stream.converter;
 
-import org.culturegraph.mf.exceptions.FormatException;
-import org.culturegraph.mf.stream.converter.RegexDecoder;
-import org.culturegraph.mf.stream.sink.EventList;
-import org.culturegraph.mf.stream.sink.StreamValidator;
-import org.junit.Assert;
-import org.junit.Test;
+import static org.junit.Assert.fail;
+
+import org.culturegraph.mf.exceptions.FormatException;
+import org.culturegraph.mf.stream.sink.EventList;
+import org.culturegraph.mf.stream.sink.StreamValidator;
+import org.junit.Test;
 
 
 /**
  * Test {@link RegexDecoderTest}.
  * 
- * @author Thomas Seidel
+ * @author Thomas Seidel, Christoph Böhme
  * 
  */
 public final class RegexDecoderTest {
@@ -62,8 +62,8 @@ public void testRegex() {
 		try {
 			regexDecoder.process(INPUT);
 			regexDecoder.closeStream();
-		} catch(FormatException e) {
-			Assert.fail(e.toString());
+		} catch (FormatException e) {
+			fail(e.toString());
 		}
 	}
 
@@ -85,10 +85,37 @@ public void testRecordId() {
 		try {
 			regexDecoder.process("RECORD-ID:28,DATA:test");
 			regexDecoder.closeStream();
-		} catch(FormatException e) {
-			Assert.fail(e.toString());
+		} catch (FormatException e) {
+			fail(e.toString());
 		}
 
+	}
+	
+	@Test
+	public void testIgnoreNonMatching() {
+		final EventList expected = new EventList();
+		
+		expected.startRecord("");
+		expected.literal("l", "v1");
+		expected.endRecord();
+		expected.startRecord("");
+		expected.literal("l", "v2");
+		expected.endRecord();
+		expected.closeStream();
+		
+		final RegexDecoder regexDecoder = new RegexDecoder("^l:(?<l>.*?)$");
+		final StreamValidator validator = new StreamValidator(expected.getEvents());
+		
+		regexDecoder.setReceiver(validator);
+		
+		try {
+			regexDecoder.process("l:v1");
+			regexDecoder.process("garbage should be ignored");
+			regexDecoder.process("l:v2");
+			regexDecoder.closeStream();
+		} catch (FormatException e) {
+			fail(e.toString());
+		}
 	}
 
 }