[HIVE-22061] Selecting from JsonSerDe backed table fails for array type

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.1.0
• Assignee: Unassigned
• Reporter: Elkhan Eminov
• Created: 30/Jul/19 14:09
• Updated: 15/Jul/20 20:17

Description

Table ddl:
CREATE EXTERNAL TABLE tmp.la_ocs_edr
(
SPEND_STRUCT map<string,array<map<string,string>>>
)
ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'
LOCATION '/tmp/vm';
Json file structure:
{"SPEND_STRUCT": {"def":[

{"COSTS":"0","BALANCE_TYPES":"128","BALANCES":"22"}

]}}
{"SPEND_STRUCT": {"def":[

{"COSTS":"0","BALANCE_TYPES":"478","BALANCES":"25"}

]}}
A select query fails with:
hive> select * from json_example;
OK
Failed with exception java.io.IOException:org.apache.hadoop.hive.serde2.SerDeException: java.io.IOException: Start of Array expected
Same result from beeline.
Strangely, inserting into the table as follows works and the JSON file is populated properly:
hive> insert into json_example select map("def",array(map("COSTS","0","BALANCE_TYPES","128","BALANCES","22"))) from t1;
then
hive> !sh hdfs dfs -cat /user/hive/warehouse/test.db/json_example/*;
{"spend_struct":{"def":[

{"BALANCE_TYPES":"128","BALANCES":"22","COSTS":"0"}

]}}

Issue Links

• supercedes: HIVE-18785 Make JSON SerDe First-Class SerDe