Github user gvramana commented on a diff in the pull request:
https://github.com/apache/carbondata/pull/2131#discussion_r178511046
--- Diff:
hadoop/src/main/java/org/apache/carbondata/hadoop/api/CarbonTableInputFormat.java
---
@@ -224,22 +198,28 @@ protected CarbonTable
getOrCreateCarbonTable(Configuration configuration) throws
// do block filtering and get split
List<InputSplit> splits =
- getSplits(job, filterInterface, filteredSegmentToAccess,
matchedPartitions, partitionInfo,
- null, updateStatusManager);
+ getSplits(job, filterInterface, filteredSegmentToAccess,
matchedPartitions, partitionInfo,
+ null, updateStatusManager, readCommitted);
+
// pass the invalid segment to task side in order to remove index
entry in task side
- if (invalidSegments.size() > 0) {
- for (InputSplit split : splits) {
- ((org.apache.carbondata.hadoop.CarbonInputSplit)
split).setInvalidSegments(invalidSegments);
- ((org.apache.carbondata.hadoop.CarbonInputSplit) split)
- .setInvalidTimestampRange(invalidTimestampsList);
+ if (readCommitted instanceof TableStatusReadCommitted) {
--- End diff --
One inputformat will read table according to one readCommitted. So make it
member variable of inputformat
---
