[
https://issues.apache.org/jira/browse/DRILL-4102?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15010276#comment-15010276
]
Sudheesh Katkam commented on DRILL-4102:
----------------------------------------
Drill supports json files of a [certain
format|https://drill.apache.org/docs/json-data-model/#reading-json].
A simple change to the file allows for queries that you might be interested in:
{code}
{
"keys" :
{
"Key1":
{
"htmltags": "<htmltag attr1='bravo' /><htmltag attr2='delta' /><htmltag
attr3='charlie' />"
},
"Key2":
{
"htmltags": "<htmltag attr1='kilo' /><htmltag attr2='lima' /><htmltag
attr3='mike' />"
},
"Key3":
{
"htmltags": "<htmltag attr1='november' /><htmltag attr2='foxtrot'
/><htmltag attr3='sierra' />"
}
}
}
{code}
Queries:
{code}
> select kvgen(keys) from dfs.`/root/data.json`;
+--------+
| EXPR$0 |
+--------+
| [{"key":"Key1","value":{"htmltags":"<htmltag attr1='bravo' /><htmltag
attr2='delta' /><htmltag attr3='charlie'
/>"}},{"key":"Key2","value":{"htmltags":"<htmltag attr1='kilo' /><htmltag
attr2='lima' /><htmltag attr3='mike'
/>"}},{"key":"Key3","value":{"htmltags":"<htmltag attr1='november' /><htmltag
attr2='foxtrot' /><htmltag attr3='sierra' />"}}] |
+--------+
> select flatten(kvgen(keys)) from dfs.`/root/data.json`;
+--------------------------------------------------------------------------------------------------------------------------+
| EXPR$0
|
+--------------------------------------------------------------------------------------------------------------------------+
| {"key":"Key1","value":{"htmltags":"<htmltag attr1='bravo' /><htmltag
attr2='delta' /><htmltag attr3='charlie' />"}} |
| {"key":"Key2","value":{"htmltags":"<htmltag attr1='kilo' /><htmltag
attr2='lima' /><htmltag attr3='mike' />"}} |
| {"key":"Key3","value":{"htmltags":"<htmltag attr1='november' /><htmltag
attr2='foxtrot' /><htmltag attr3='sierra' />"}} |
+--------------------------------------------------------------------------------------------------------------------------+
> select t.r.key, t.r.`value` from (select flatten(kvgen(keys)) as r from
> dfs.`/root/data.json`) t;
+---------+---------------------------------------------------------------------------------------------------+
| EXPR$0 | EXPR$1
|
+---------+---------------------------------------------------------------------------------------------------+
| Key1 | {"htmltags":"<htmltag attr1='bravo' /><htmltag attr2='delta'
/><htmltag attr3='charlie' />"} |
| Key2 | {"htmltags":"<htmltag attr1='kilo' /><htmltag attr2='lima'
/><htmltag attr3='mike' />"} |
| Key3 | {"htmltags":"<htmltag attr1='november' /><htmltag attr2='foxtrot'
/><htmltag attr3='sierra' />"} |
+---------+---------------------------------------------------------------------------------------------------+
{code}
> Only one row found in a JSON document that contains multiple items.
> -------------------------------------------------------------------
>
> Key: DRILL-4102
> URL: https://issues.apache.org/jira/browse/DRILL-4102
> Project: Apache Drill
> Issue Type: Bug
> Environment: OS X, Drill embedded, v1.1.0 installed via HomeBrew
> Reporter: aditya menon
>
> I tried to analyse a JSON file that had the following (sample) structure:
> {code:json}
> {
> "Key1": {
> "htmltags": "<htmltag attr1='bravo' /><htmltag attr2='delta' /><htmltag
> attr3='charlie' />"
> },
> "Key2": {
> "htmltags": "<htmltag attr1='kilo' /><htmltag attr2='lima' /><htmltag
> attr3='mike' />"
> },
> "Key3": {
> "htmltags": "<htmltag attr1='november' /><htmltag attr2='foxtrot'
> /><htmltag attr3='sierra' />"
> }
> }
> {code}
> (Apologies for the obfuscation, I am unable to publish the original dataset.
> But the structure is exactly the same. Note especially how the keys and other
> data points *differ* in some places, and remain identical in others.)
> When I run a {code:sql}SELECT * FROM DataFile.json{code} what I get is a
> single row listed under three columns: {code:html}"<htmltag attr1='bravo'
> /><htmltag attr2='delta' /><htmltag attr3='charlie' />"{code} [i.e., only the
> entry `Key1.htmltags`] .
> Ideally, I should see three rows, each with entries from Key1..Key3, listed
> under the correct respective column.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
