arina-ielchiieva commented on a change in pull request #1387: DRILL-6603: Set
num_nulls for parquet statistics to -1 when actual number is not defined.
URL: https://github.com/apache/drill/pull/1387#discussion_r203665408
##########
File path:
exec/java-exec/src/test/java/org/apache/drill/exec/planner/logical/TestConvertCountToDirectScan.java
##########
@@ -159,4 +159,30 @@ public void ensureConvertForSeveralColumns() throws
Exception {
}
}
+ @Test
+ public void ensureCorrectCountWithMissingStatistics() throws Exception {
+ dirTestWatcher.copyResourceToRoot(Paths.get("parquet",
"wide_string.parquet"));
+
+ test("use dfs.tmp");
+ String tableName = "wide_str_table";
+ try {
+ // table will contain two partitions: one - with null value, second -
with non null value
+ test("create table %s partition by (col_str) as select * from
dfs.`parquet/wide_string.parquet`", tableName);
+
+ String query = String.format("select count(col_str) as cnt_str, count(*)
as cnt_total from %s", tableName);
+
+ // direct scan should not be applied since we don't have statistics
+ testPlanMatchingPatterns(query, new String[]{}, new
String[]{"DynamicPojoRecordReader"});
Review comment:
Done, thanks.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
