[carbondata] branch master updated: [CARBONDATA-3998] Fixed file not found exception for hive during insert

[ak...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

akashrn5 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/carbondata.git


The following commit(s) were added to refs/heads/master by this push:
     new 7399c90  [CARBONDATA-3998] Fixed file not found exception for hive during insert
7399c90 is described below

commit 7399c908b04861585155573be223cccfdeba3438
Author: kunal642 <ku...@gmail.com>
AuthorDate: Mon Sep 21 17:04:20 2020 +0530

    [CARBONDATA-3998] Fixed file not found exception for hive during insert
    
    Why is this PR needed?
    FileNotFoundException being thrown in hive during insert.
    
    What changes were proposed in this PR?
    During insert the path in the parameters for write is for staging directory,
    because carbon is non-native table therefore this directory does not exist.
    When getAbsolutePath is called on this directory, it throws FileNotFound
    exception. Removed the getAbsolutePath call to avoid this.
    
    Does this PR introduce any user interface change?
    No
    
    Is any new testcase added?
    No
    
    This closes #3941
---
 .../java/org/apache/carbondata/hive/MapredCarbonInputFormat.java  | 4 +++-
 .../java/org/apache/carbondata/hive/MapredCarbonOutputFormat.java | 8 ++++----
 2 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonInputFormat.java b/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonInputFormat.java
index 8ef2529..27fa209 100644
--- a/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonInputFormat.java
+++ b/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonInputFormat.java
@@ -158,7 +158,9 @@ public class MapredCarbonInputFormat extends CarbonTableInputFormat<ArrayWritabl
     try {
       splitList = carbonInputFormat.getSplits(jobContext);
     } catch (IOException ex) {
-      if (ex.getMessage().contains("No Index files are present in the table location :")) {
+      LOGGER.error("Unable to get splits: ", ex);
+      if (ex.getMessage().contains("No Index files are present in the table location :") ||
+          ex.getMessage().contains("CarbonData file is not present in the table location")) {
         splitList = new ArrayList<>();
       } else {
         throw ex;
diff --git a/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonOutputFormat.java b/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonOutputFormat.java
index a486fc0..7e1da4d 100644
--- a/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonOutputFormat.java
+++ b/integration/hive/src/main/java/org/apache/carbondata/hive/MapredCarbonOutputFormat.java
@@ -100,10 +100,9 @@ public class MapredCarbonOutputFormat<T> extends CarbonTableOutputFormat
         carbonLoadModel.getCarbonDataLoadSchema().getCarbonTable().getPartitionInfo();
     final int partitionColumn =
         partitionInfo != null ? partitionInfo.getColumnSchemaList().size() : 0;
-    String finalOutputPath = FileFactory.getCarbonFile(finalOutPath.toString()).getAbsolutePath();
     if (carbonLoadModel.getCarbonDataLoadSchema().getCarbonTable().isHivePartitionTable()) {
-      carbonLoadModel.getMetrics().addToPartitionPath(finalOutputPath);
-      context.getConfiguration().set("carbon.outputformat.writepath", finalOutputPath);
+      carbonLoadModel.getMetrics().addToPartitionPath(finalOutPath.toString());
+      context.getConfiguration().set("carbon.outputformat.writepath", finalOutPath.toString());
     }
     CarbonTableOutputFormat.setLoadModel(jc, carbonLoadModel);
     org.apache.hadoop.mapreduce.RecordWriter<NullWritable, ObjectArrayWritable> re =
@@ -116,7 +115,8 @@ public class MapredCarbonOutputFormat<T> extends CarbonTableOutputFormat
           if (isHivePartitionedTable) {
             Object[] actualRow = ((CarbonHiveRow) writable).getData();
             Object[] newData = Arrays.copyOf(actualRow, actualRow.length + partitionColumn);
-            String[] partitionValues = finalOutputPath.substring(tablePath.length()).split("/");
+            String[] partitionValues = finalOutPath.toString().substring(tablePath.length())
+                .split("/");
             for (int j = 0, i = actualRow.length; j < partitionValues.length; j++) {
               if (partitionValues[j].contains("=")) {
                 newData[i++] = partitionValues[j].split("=")[1];