[
https://issues.apache.org/jira/browse/HIVE-10592?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14529258#comment-14529258
]
Gopal V commented on HIVE-10592:
--------------------------------
LGTM - +1
But a follow-up usability JIRA advised for the multi-file output scenarios -
you need to produce an array of JSON objects instead of a JSON object of arrays.
Since there are no consumers for this output until there is some output, we
need to iterate on this after writing some analysis scripts once this makes it
into the build.
As an example of the difficulty in keeping JSON object walkers simple, try
running
{code}
./dist/hive/bin/hive --service orcfiledump -j -p
/apps/hive/warehouse/tpcds5_bin_partitioned_orc_200.db/customer_demographics/000000_0
/apps/hive/warehouse/tpcds5_bin_partitioned_orc_200.db/customer_demographics/000001_0
{
"fileName": [
"\/apps\/hive\/warehouse\/tpcds5_bin_partitioned_orc_200.db\/customer_demographics\/000000_0",
"\/apps\/hive\/warehouse\/tpcds5_bin_partitioned_orc_200.db\/customer_demographics\/000001_0"
],
"fileVersion": [
"0.12",
"0.12"
],
"writerVersion": [
"HIVE_8732",
"HIVE_8732"
],
...
{code}
> ORC file dump in JSON format
> ----------------------------
>
> Key: HIVE-10592
> URL: https://issues.apache.org/jira/browse/HIVE-10592
> Project: Hive
> Issue Type: New Feature
> Affects Versions: 1.3.0
> Reporter: Prasanth Jayachandran
> Assignee: Prasanth Jayachandran
> Attachments: HIVE-10592.1.patch, HIVE-10592.2.patch,
> HIVE-10592.3.patch
>
>
> ORC file dump uses custom format. Will be useful to dump ORC metadata in json
> format so that other tools can be built on top it.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
