Eclipse Community Forums: BIRT » Dataset joins uses long names as alias

Home

Home » Archived » BIRT » Dataset joins uses long names as alias

Dataset joins uses long names as alias [message #721882]

Sat, 03 September 2011 06:47

Eclipse User

Hello all,

I'm new to this forum. Hope to have some help here Smile

Before I was using BIRT 2.6 version, now I have moved to the 3.7 version and noticed one strange thing.
Lets say I'm using 9 data sets and later I want to make left outer join on all those datasets. My goal is to align all 9 data sets by date time field and make one big joint data set. So after I joined all sets I can access columns by using those blody names like "dataSetRow["Debitas_7::Debitas_6::Debitas_5::Debitas_4::Debitas_3::Debitas_2::Debitas_1::FT_10101_FLOW::FT_10101_FLOW_DIFF"]"

In version 2.6 all joint data sets were created with simple aliases and the final joint data set had simple names to access colums like "dataSetRow["FT_10101_FLOW_DIFF"]".

In version 3.7 if I create joints then aliases uses long names. Don't know why.

Re: Dataset joins uses long names as alias [message #722136 is a reply to message #721882]

Sun, 04 September 2011 11:53

Eclipse User

Ok, I try to make my question very easy to understand. I just upload screenshots from BIRT. Two of them are of version 2.6 and second two are 3.7. Just look at the pictures attached and you will understand what I'm trying to ask. Thanks.

Attachment: question.zip
(Size: 1.27MB, Downloaded 414 times)

Re: Dataset joins uses long names as alias [message #722682 is a reply to message #721882]

Tue, 06 September 2011 10:10

Eclipse User

If you open the joint data set in the data set editor you should be able
to choose output columns and change the alias to whatever you want it to be.

Jason

On 9/3/2011 6:47 AM, Darius wrote:
> Hello all,
>
> I'm new to this forum. Hope to have some help here :)
> Before I was using BIRT 2.6 version, now I have moved to the 3.7 version
> and noticed one strange thing.
> Lets say I'm using 9 data sets and later I want to make left outer join
> on all those datasets. My goal is to align all 9 data sets by date time
> field and make one big joint data set. So after I joined all sets I can
> access columns by using those blody names like
> "dataSetRow["Debitas_7::Debitas_6::Debitas_5::Debitas_4::Debitas_3::Debitas_2::Debitas_1::FT_10101_FLOW::FT_10101_FLOW_DIFF"]"
>
>
> In version 2.6 all joint data sets were created with simple aliases and
> the final joint data set had simple names to access colums like
> "dataSetRow["FT_10101_FLOW_DIFF"]".
>
> In version 3.7 if I create joints then aliases uses long names. Don't
> know why.

Re: Dataset joins uses long names as alias [message #722685 is a reply to message #721882]

Tue, 06 September 2011 10:10

Eclipse User

Re: Dataset joins uses long names as alias [message #722792 is a reply to message #722685]

Tue, 06 September 2011 13:53

Eclipse User

Hello,

Thanks for a reply. What you are saying is completely true, but in version 3.7 there is some strange things:
1. Simple named alias is not created automatically as in version 2.6
2. Even if I make nice simple aliases I can't use them in later designing process. I mean if I form column and try to select some datarow from data set then I don't see any aliases. Only those long names. Nothing more. Don't know why.

Re: Dataset joins uses long names as alias [message #722798 is a reply to message #722792]

Tue, 06 September 2011 14:09

Eclipse User

I am not certain why this was changed, but you should be able to
reference the alias like row["Myalias"] although it does not show up in
the expression builder. Also log an enhancement/bugzilla to get more
details on why it was changed.

Jason

On 9/6/2011 1:53 PM, Darius wrote:
> Hello,
>
> Thanks for a reply. What you are saying is completely true, but in
> version 3.7 there is some strange things:
> 1. Simple named alias is not created automatically as in version 2.6
> 2. Even if I make nice simple aliases I can't use them in later
> designing process. I mean if I form column and try to select some
> datarow from data set then I don't see any aliases. Only those long
> names. Nothing more. Don't know why.

Re: Dataset joins uses long names as alias [message #722893 is a reply to message #722798]

Wed, 07 September 2011 01:45

Eclipse User

Hello,

Yes, if I reference it like row["Myalias"] the its ok, but the problem is that I don't see those aliases, I need to know them by heart. Its difficult.
About bug. I'm not sure if its a bug Sad

Maybe it was just changed interface or I do something wrong, because if I open report on 3.7 which was made with 2.6 then its ok. I can use aliases in find. I've tried to look at XML file of the report, but it was too big to understand the differences Sad

Re: Dataset joins uses long names as alias [message #723529 is a reply to message #722893]

Thu, 08 September 2011 14:28

Eclipse User

I do not think it is a bug, but maybe should be classified as an
enhancement.

Jason

On 9/7/2011 1:45 AM, Darius wrote:
> Hello,
>
> Yes, if I reference it like row["Myalias"] the its ok, but the problem
> is that I don't see those aliases, I need to know them by heart. Its
> difficult.
> About bug. I'm not sure if its a bug :( Maybe it was just changed
> interface or I do something wrong, because if I open report on 3.7 which
> was made with 2.6 then its ok. I can use aliases in find. I've tried to
> look at XML file of the report, but it was too big to understand the
> differences :(

Re: Dataset joins uses long names as alias [message #723553 is a reply to message #723529]

Thu, 08 September 2011 15:14

Eclipse User

Hi,

I believe so, but I would be happy if any user report the same problem on 3.7. Then I would know that I'm not the only one and that I did everything correct.

Re: Dataset joins uses long names as alias [message #723580 is a reply to message #723553]

Thu, 08 September 2011 16:57

Eclipse User

I saw the same issue.

On 9/8/2011 3:14 PM, Darius wrote:
> Hi,
>
> I believe so, but I would be happy if any user report the same problem
> on 3.7. Then I would know that I'm not the only one and that I did
> everything correct.

Previous Topic:	Unable to convert a date string to a Date object in BIRT
Next Topic:	validate() function with customized message

Goto Forum:

-=] Back to Top [=-

Current Time: Wed Jul 23 09:57:34 EDT 2025

.:: Contact :: Home ::.

Breadcrumbs

Sign up to our Newsletter