Hi Tsung Han,

For each function, we will add a new sql and a session interface at the same 
time. Session is for developers and sql for data analysts.

Therefore, for issue[1], we will add both. The antlr is upgrading now (all sql 
grammars are changing). Therefore, you can add an interface in Session first. 

PR[2] only finishes some part of deleting storage group of new sql, about 30% 
of all work :) This PR will be closed and you can work on a new one.

Besides, the naming of deleteTimeseries should be delete.

[1]https://issues.apache.org/jira/projects/IOTDB/issues/IOTDB-188?filter=allopenissues
[2] https://github.com/apache/incubator-iotdb/pull/376

Best,
--
Jialin Qiao
School of Software, Tsinghua University

乔嘉林
清华大学 软件学院

> -----原始邮件-----
> 发件人: "Jack Tsai" <jack870...@outlook.com>
> 发送时间: 2019-09-14 00:17:50 (星期六)
> 收件人: "dev@iotdb.apache.org" <dev@iotdb.apache.org>
> 抄送: 
> 主题: 回覆: Enrich Session interfaces
> 
> Hi Jialin,
> 
> A minor question is that, regarding to issue[1], is it for Session interface? 
> Or it is just the new sql?
> 
> And I don't really understand what is the meaning of the "30%" in the commit 
> message of PR[2]? If PR[2] is already for deleting the storage group, what 
> else do I need to do for it?
> 
> 
> [1]https://issues.apache.org/jira/projects/IOTDB/issues/IOTDB-188?filter=allopenissues
> [2] https://github.com/apache/incubator-iotdb/pull/376
> 
> Best,
> Tsung Han Tsai
> ________________________________
> 寄件者: Jialin Qiao <qj...@mails.tsinghua.edu.cn>
> 寄件日期: 2019年9月6日 下午 01:35
> 收件者: dev@iotdb.apache.org <dev@iotdb.apache.org>
> 主旨: Re: Enrich Session interfaces
> 
> Hi Tsung Han,
> 
> I have checked your PR and give some comments. No big problem, only some 
> optimizations.
> 
> Besides, before adding the query interface, could you please add the deleting 
> storage group interface first? This is what this issue[1] is for. You could 
> refer to this PR [2].
> 
> To clarify these delete interfaces for others. I list them in the following:
> 
> (1) Deleting data in one timeseries: This is what "delete from root.sg1.d1.s1 
> where time<=100" statement and DeletePlan do, also the "delete interface" 
> means.
> (2) Deleting timeseries, including its data and schema. This is what "delete 
> timeseries root.sg1.d1.s1" and MetadataPlan do.
> (3) Deleting storage group: deleting all data and timeseries in one storage 
> group, including the corresponding storage group processor and system info in 
> data/system folder.
> 
> 
> [1]https://issues.apache.org/jira/projects/IOTDB/issues/IOTDB-188?filter=allopenissues
> [2] https://github.com/apache/incubator-iotdb/pull/376
> 
> Best,
> --
> Jialin Qiao
> School of Software, Tsinghua University
> 
> 乔嘉林
> 清华大学 软件学院
> 
> > -----原始邮件-----
> > 发件人: "Jack Tsai" <jack870...@outlook.com>
> > 发送时间: 2019-09-05 23:25:39 (星期四)
> > 收件人: "dev@iotdb.apache.org" <dev@iotdb.apache.org>
> > 抄送:
> > 主题: 回覆: Enrich Session interfaces
> >
> > Hi,
> >
> > I have added the delete interface for the Session API.  I followed the 
> > pattern which is similar to the insert interface.
> >
> > Please have a check whether it is conform to the requirement. I would 
> > implement the query interface for Session API if I have more time.
> >
> > Best,
> > Tsung-Han Tsai
> > ________________________________
> > 寄件者: Jialin Qiao <qj...@mails.tsinghua.edu.cn>
> > 寄件日期: 2019年9月2日 下午 06:48
> > 收件者: dev@iotdb.apache.org <dev@iotdb.apache.org>
> > 主旨: Re: Enrich Session interfaces
> >
> > Hi Tsung Han,
> >
> > The new insert interface only inserts one row data, not in batch, for 
> > example : (deviceId, time, s1, s2,s3, s1_value, s2_value, s3_value).
> >
> > Besides, the Session interfaces are only for IoTDB, not TsFile.
> >
> > Best,
> > --
> > Jialin Qiao
> > School of Software, Tsinghua University
> >
> > 乔嘉林
> > 清华大学 软件学院
> >
> > > -----原始邮件-----
> > > 发件人: "Jack Tsai" <jack870...@outlook.com>
> > > 发送时间: 2019-09-02 18:44:34 (星期一)
> > > 收件人: "dev@iotdb.apache.org" <dev@iotdb.apache.org>
> > > 抄送:
> > > 主题: Re: Enrich Session interfaces
> > >
> > > Hi,
> > >
> > > I wonder is the performance of this new kind of insert interface 
> > > different from the insert RowBatch interface? Or it is just a more 
> > > convenient way to insert batched data?
> > >
> > > Best regards,
> > > Tsung Han Tsai
> > >
> > > 取得 iOS 版 Outlook<https://aka.ms/o0ukef>
> > > ________________________________
> > > 寄件者: Jialin Qiao <qj...@mails.tsinghua.edu.cn>
> > > 寄件日期: Monday, September 2, 2019 4:35:55 PM
> > > 收件者: dev@iotdb.apache.org <dev@iotdb.apache.org>
> > > 主旨: Enrich Session interfaces
> > >
> > > Hi,
> > >
> > >
> > > I got some feedback from our users in Hunan province. Except for 
> > > insertBatch, they want an easier interface to insert data. Therefore, I 
> > > submit a PR [1] to add an insert interface, which is easier to understand 
> > > than the insert RowBatch interface.
> > >
> > >
> > > Moreover, I have created an issue [2] to enrich the Session interfaces, 
> > > looking forward to more contributors for the session interfaces.
> > >
> > >
> > > [1] https://github.com/apache/incubator-iotdb/pull/367
> > > [2] https://issues.apache.org/jira/browse/IOTDB-191
> > >
> > > Best,
> > > --
> > > Jialin Qiao
> > > School of Software, Tsinghua University
> > >
> > > 乔嘉林
> > > 清华大学 软件学院

Reply via email to