[
https://issues.apache.org/jira/browse/HIVE-26320?focusedWorklogId=813740&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-813740
]
ASF GitHub Bot logged work on HIVE-26320:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 30/Sep/22 15:31
Start Date: 30/Sep/22 15:31
Worklog Time Spent: 10m
Work Description: jfsii commented on code in PR #3628:
URL: https://github.com/apache/hive/pull/3628#discussion_r984714606
##########
ql/src/java/org/apache/hadoop/hive/ql/io/parquet/convert/ETypeConverter.java:
##########
@@ -481,12 +485,36 @@ protected BytesWritable convert(Binary binary) {
},
ESTRING_CONVERTER(String.class) {
@Override
- PrimitiveConverter getConverter(final PrimitiveType type, final int index,
final ConverterParent parent, TypeInfo hiveTypeInfo) {
+ PrimitiveConverter getConverter(final PrimitiveType type, final int index,
final ConverterParent parent,
+ TypeInfo hiveTypeInfo) {
+ // If we have type information, we should return properly typed strings.
However, there are a variety
+ // of code paths that do not provide the typeInfo in those cases we
default to Text. This idiom is also
+ // followed by for example the BigDecimal converter in which if there is
no type information,
+ // it defaults to the widest representation
+ if (hiveTypeInfo != null) {
+ String typeName = hiveTypeInfo.getTypeName().toLowerCase();
+ if (typeName.startsWith(serdeConstants.CHAR_TYPE_NAME)) {
+ return new BinaryConverter<HiveCharWritable>(type, parent, index) {
+ @Override
+ protected HiveCharWritable convert(Binary binary) {
+ return new HiveCharWritable(binary.getBytes(), ((CharTypeInfo)
hiveTypeInfo).getLength());
+ }
+ };
+ } else if (typeName.startsWith(serdeConstants.VARCHAR_TYPE_NAME)) {
+ return new BinaryConverter<HiveVarcharWritable>(type, parent, index)
{
+ @Override
+ protected HiveVarcharWritable convert(Binary binary) {
+ return new HiveVarcharWritable(binary.getBytes(),
((VarcharTypeInfo) hiveTypeInfo).getLength());
+ }
+ };
+ }
+ }
Review Comment:
Yeah, I did follow the file on its convention. However, I changed it to your
suggestion because it looks cleaner and might as well encourage this still in
case someone copies me
Issue Time Tracking
-------------------
Worklog Id: (was: 813740)
Time Spent: 3.5h (was: 3h 20m)
> Incorrect case evaluation for Parquet based table
> -------------------------------------------------
>
> Key: HIVE-26320
> URL: https://issues.apache.org/jira/browse/HIVE-26320
> Project: Hive
> Issue Type: Bug
> Components: HiveServer2, Query Planning
> Affects Versions: 4.0.0-alpha-1
> Reporter: Chiran Ravani
> Assignee: John Sherman
> Priority: Major
> Labels: pull-request-available
> Time Spent: 3.5h
> Remaining Estimate: 0h
>
> Query involving case statement with two or more conditions leads to incorrect
> result for tables with parquet format, The problem is not observed with ORC
> or TextFile.
> *Steps to reproduce*:
> {code:java}
> create external table case_test_parquet(kob varchar(2),enhanced_type_code
> int) stored as parquet;
> insert into case_test_parquet values('BB',18),('BC',18),('AB',18);
> select case when (
> (kob='BB' and enhanced_type_code='18')
> or (kob='BC' and enhanced_type_code='18')
> )
> then 1
> else 0
> end as logic_check
> from case_test_parquet;
> {code}
> Result:
> {code}
> 0
> 0
> 0
> {code}
> Expected result:
> {code}
> 1
> 1
> 0
> {code}
> The problem does not appear when setting hive.optimize.point.lookup=false.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
