drill-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Prabhakar Bhosaale <bhosale....@gmail.com>
Subject Re: Querying json files from multiple subdirectories
Date Sun, 19 Jan 2020 06:51:53 GMT
Thanks charles. Will try few options and get back to you.

Regards
Prabhakar

On Sun, Jan 19, 2020, 04:45 Charles Givre <cgivre@gmail.com> wrote:

> Hi Prabhakar,
> You'll need to find some common identifier for the files you want to
> query.
> It could be something like:
>
> SELECT
> FROM dfs.`<path>/Year*/`
>
> Alternatively, you could have multiple SELECT queries and join them
> together via a UNION statement.  IE:
>
> SELECT * FROM dfs.`Year2013/trans.json`
> UNION
> SELECT * FROM dfs.`Year2014/trans.json`
>
>
>
> -- C
>
> > On Jan 17, 2020, at 11:07 PM, Prabhakar Bhosaale <bhosale.p.v@gmail.com>
> wrote:
> >
> > Hi Charls,
> > Thanks for your suggestion. Actually the transactions folder will have
> more
> > yearwise folder. But i want to query only few folders at a time. The
> >
> > Regards
> > Prabhakar
> >
> > On Fri, Jan 17, 2020, 20:01 Charles Givre <cgivre@gmail.com> wrote:
> >
> >> Hi there,
> >> If you have that directory structure, the following query should work:
> >>
> >> SELECT *
> >> FROM dfs.<workspace>.`transactions/` as t1
> >>
> >> Obviously replacing <workspace> with your workspace.  You can then join
> >> that with anything that Drill can query.
> >> Best,
> >> -- C
> >>
> >>
> >>
> >>> On Jan 17, 2020, at 1:27 AM, Prabhakar Bhosaale <bhosale.p.v@gmail.com
> >
> >> wrote:
> >>>
> >>> Hi All,
> >>>
> >>> I am new to apache drill and trying to retrieve data from json files by
> >>> querying the directories.
> >>>
> >>> The directory structure is
> >>>
> >>>                       |------>Year2012--->trans.json
> >>>                       |
> >>>                       |
> >>> transactions-->|
> >>>                       |
> >>>                       |------>Year2013--->trans.json
> >>>
> >>> I would like to query trans.json from both the sub-directories as one
> >> table
> >>> and then join the resultant table with another table in a single query.
> >>> Please help with possible options. thx
> >>>
> >>> Regards
> >>
> >>
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message