[BACKLOG-18271] - CSV File Input: Columns with the same name in the c…

…sv are not read properly (#4274) * [BACKLOG-18271] - CSV File Input: Columns with the same name in the csv are not read properly * [BACKLOG-18271] - CSV File Input: Columns with the same name in the csv are not read properly - test
pentaho · Aug 14, 2017 · 04cd197 · 04cd197
1 parent 6a8fc33
commit 04cd197
Show file tree

Hide file tree

Showing 2 changed files with 28 additions and 2 deletions.
diff --git a/engine/src/main/java/org/pentaho/di/trans/steps/csvinput/NamedFieldsMapping.java b/engine/src/main/java/org/pentaho/di/trans/steps/csvinput/NamedFieldsMapping.java
@@ -22,9 +22,13 @@
 
 package org.pentaho.di.trans.steps.csvinput;
 
+import java.util.Collection;
 import java.util.HashMap;
 import java.util.Map;
 
+import org.apache.commons.collections4.MultiValuedMap;
+import org.apache.commons.collections4.multimap.ArrayListValuedHashMap;
+
 public class NamedFieldsMapping implements FieldsMapping {
 
   private final Map<Integer, Integer> actualToMetaFieldMapping;
@@ -45,13 +49,19 @@ public int size() {
   }
 
   public static NamedFieldsMapping mapping( String[] actualFieldNames, String[] metaFieldNames ) {
-    Map<String, Integer> metaNameToIndex = new HashMap<>();
+    MultiValuedMap<String, Integer> metaNameToIndex = new ArrayListValuedHashMap<String, Integer>();
     for ( int j = 0; j < metaFieldNames.length; j++ ) {
       metaNameToIndex.put( metaFieldNames[j], Integer.valueOf( j ) );
     }
     Map<Integer, Integer> actualToMetaFieldMapping = new HashMap<>();
     for ( int i = 0; i < actualFieldNames.length; i++ ) {
-      actualToMetaFieldMapping.put( i, metaNameToIndex.get( actualFieldNames[i] ) );
+      Collection<Integer> columnIndexes = metaNameToIndex.get( actualFieldNames[i] );
+      if ( columnIndexes.isEmpty() ) {
+        continue;
+      }
+      Integer columnIndex = columnIndexes.iterator().next();
+      metaNameToIndex.removeMapping( actualFieldNames[i], columnIndex );
+      actualToMetaFieldMapping.put( i, columnIndex );
     }
     return new NamedFieldsMapping( actualToMetaFieldMapping );
   }

diff --git a/engine/src/test/java/org/pentaho/di/trans/steps/csvinput/NamedFieldsMappingTest.java b/engine/src/test/java/org/pentaho/di/trans/steps/csvinput/NamedFieldsMappingTest.java
@@ -64,4 +64,20 @@ public void mapping() {
     assertEquals( 0, mapping.fieldMetaIndex( 1 ) );
   }
 
+  @Test
+  public void mappingWithNonUniqueColumnNames() {
+    NamedFieldsMapping mapping =
+        NamedFieldsMapping.mapping( new String[] { "Object", "Test", "Object" }, new String[] { "Object", "Test",
+          "Object" } );
+    assertEquals( 0, mapping.fieldMetaIndex( 0 ) );
+    assertEquals( 2, mapping.fieldMetaIndex( 2 ) );
+  }
+
+  @Test
+  public void fieldMetaIndexWithUnexistingField_nonUniqueColumnNames() {
+    NamedFieldsMapping mapping =
+        NamedFieldsMapping.mapping( new String[] { "Object", "Test", "Object" }, new String[] { "Object", "Test" } );
+    assertEquals( FieldsMapping.FIELD_DOES_NOT_EXIST, mapping.fieldMetaIndex( 2 ) );
+  }
+
 }