Github user geetikagupta16 commented on a diff in the pull request:
https://github.com/apache/carbondata/pull/2014#discussion_r171464754
--- Diff:
hadoop/src/main/java/org/apache/carbondata/hadoop/streaming/CarbonStreamRecordWriter.java
---
@@ -181,75 +181,81 @@ private void initializeAtFirstRow() throws
IOException, InterruptedException {
initializeAtFirstRow();
}
- // parse and convert row
- currentRow.setData(rowParser.parseRow((Object[]) value));
- converter.convert(currentRow);
-
// null bit set
nullBitSet.clear();
- for (int i = 0; i < dataFields.length; i++) {
- if (null == currentRow.getObject(i)) {
- nullBitSet.set(i);
+ Object[] rowData = (Object[]) value;
+ currentRow.setRawData(rowData);
+ // parse and convert row
+ currentRow.setData(rowParser.parseRow(rowData));
+ CarbonRow updatedCarbonRow = converter.convert(currentRow);
+ if (updatedCarbonRow == null) {
+ output.skipRow();
+ currentRow.clearData();
+ } else {
--- End diff --
@mohammadshahidkhan Yes, else part remains the same
---
