[
https://issues.apache.org/jira/browse/DRILL-8393?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17700907#comment-17700907
]
ASF GitHub Bot commented on DRILL-8393:
---------------------------------------
cgivre commented on PR #2747:
URL: https://github.com/apache/drill/pull/2747#issuecomment-1471002379
@LYCJeff Thanks for making these changes. I have a few questions:
1. Are you certain that these filters are in fact being pushed down as
intended?
2. I'm really concerned about what would happen if a user aliased a data
source as `header` or `tail`.
IE:
```sql
SELECT ...
FROM api.foo
INNER JOIN dfs.`tail.csv` AS tail
ON tail.id = foo.id
WHERE tail.name = 'something'
```
Do we know how this would be interpreted?
> Allow parameters to be passed to headers through SQL in WHERE clause
> --------------------------------------------------------------------
>
> Key: DRILL-8393
> URL: https://issues.apache.org/jira/browse/DRILL-8393
> Project: Apache Drill
> Issue Type: Improvement
> Components: Storage - HTTP
> Affects Versions: 1.20.0
> Reporter: Yuchen Liang
> Priority: Major
>
> Some APIs require parameters (e.g. digital signature) in the headers to be
> generated at access time.So I'm wondering if we can pass it in through filter
> statement.
> Perhaps we could design it like the params field in connections parameter.
> For example:
>
> Config:
> { "url": "https://api.sunrise-sunset.org/json";, "requireTail": false,
> "params": ["body.lat", "body.lng", "body.date", "header.header1"],
> "parameterLocation": "json_body" }
>
> SQL Query:
> SELECT * FROM api.sunrise
> WHERE `body.lat` = 36.7201600
> AND `body.lng` = -4.4203400
> AND `body.date` = '2019-10-02'
> AND `header.header1` = 'value1';
>
> Post body:
> { "lat": 36.7201600, "lng": -4.4203400, "date": "2019-10-02"}
>
> Headers:
> { "header1": "value1", ……}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
