Discarding rows in dataset.

[Graham Blandford]

Hi all,

Would someone be able to tell me the most 'graceful' way of removing
unwanted rows from a dataset based on a condition prior to update? OR,
resetting the rows all to unchanged after they are initally added to the
recordset.

I create a dataset, which begins empty after the initial .Fill.

Then I create several rows with some default information, leaving one
critical piece of information value as 0;

dr = ds.Tables("myta ble").NewRow()
With dr
.Item("basicdat a1") = myvalue1
.Item("basicdat a2") = myvalue2
.Item("critical data") = 0
.Item("created" ) = Now()
End With
ds.Tables("myta ble").Rows.Add( dr)

Then, my update does something like;

da.Update(ds, "mytable")

All pretty basic stuff.... but what I would like is to ONLY insert new rows
into the database that have a "criticalda ta" value <> 0 and discard the
rest. I could use the GetChanges method if I knew how to mark the rows as
'unchanged' immediately after creation - but don't know how.

Any help would be greatly appreciated.
Thanks,

Graham

[William Ryan eMVP]

YOu can use GetChanges and just get the rows with RowState or added. Then
call update on the GetChanges table - then acceptChanges on the original
--
W.G. Ryan MVP Windows - Embedded

Have an opinion on the effectiveness of Microsoft Embedded newsgroups?
Let Microsoft know!
https://www.windowsembeddedeval.com/...ity/newsgroups
"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:tH******** *************@n ews20.bellgloba l.com...

Hi all,

Would someone be able to tell me the most 'graceful' way of removing
unwanted rows from a dataset based on a condition prior to update? OR,
resetting the rows all to unchanged after they are initally added to the
recordset.

I create a dataset, which begins empty after the initial .Fill.

Then I create several rows with some default information, leaving one
critical piece of information value as 0;

dr = ds.Tables("myta ble").NewRow()
With dr
.Item("basicdat a1") = myvalue1
.Item("basicdat a2") = myvalue2
.Item("critical data") = 0
.Item("created" ) = Now()
End With
ds.Tables("myta ble").Rows.Add( dr)

Then, my update does something like;

da.Update(ds, "mytable")

All pretty basic stuff.... but what I would like is to ONLY insert new rows into the database that have a "criticalda ta" value <> 0 and discard the
rest. I could use the GetChanges method if I knew how to mark the rows as
'unchanged' immediately after creation - but don't know how.

Any help would be greatly appreciated.
Thanks,

Graham

[Gerard]

I guess row.AcceptChang es will do.

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:tH******** *************@n ews20.bellgloba l.com...

Hi all,

Would someone be able to tell me the most 'graceful' way of removing
unwanted rows from a dataset based on a condition prior to update? OR,
resetting the rows all to unchanged after they are initally added to the
recordset.

I create a dataset, which begins empty after the initial .Fill.

Then I create several rows with some default information, leaving one
critical piece of information value as 0;

dr = ds.Tables("myta ble").NewRow()
With dr
.Item("basicdat a1") = myvalue1
.Item("basicdat a2") = myvalue2
.Item("critical data") = 0
.Item("created" ) = Now()
End With
ds.Tables("myta ble").Rows.Add( dr)

Then, my update does something like;

da.Update(ds, "mytable")

All pretty basic stuff.... but what I would like is to ONLY insert new rows into the database that have a "criticalda ta" value <> 0 and discard the
rest. I could use the GetChanges method if I knew how to mark the rows as
'unchanged' immediately after creation - but don't know how.

Any help would be greatly appreciated.
Thanks,

Graham

[Jay B. Harlow [MVP - Outlook]]

Graham,
In addition to GetChanges with the DataRowState you can use
DataTable.Selec t.

I would recommend DataTable.Selec t as you are returned references to the
rows in the existing DataSet, where as GetChanges will return a new DataSet
(with new rows).

David Sceppa explains the difference between GetChanges & Select when you
intend on updating the database in his book "Microsoft ADO.NET - Core
Reference" from MS
Press.

If you are doing a lot with ADO.NET I strongly recommend Sceppa's book,
which is a good tutorial on ADO.NET as well as a good desk reference once
you know ADO.NET.

Hope this helps
Jay
"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:tH******** *************@n ews20.bellgloba l.com...

Hi all,

Would someone be able to tell me the most 'graceful' way of removing
unwanted rows from a dataset based on a condition prior to update? OR,
resetting the rows all to unchanged after they are initally added to the
recordset.

I create a dataset, which begins empty after the initial .Fill.

Then I create several rows with some default information, leaving one
critical piece of information value as 0;

dr = ds.Tables("myta ble").NewRow()
With dr
.Item("basicdat a1") = myvalue1
.Item("basicdat a2") = myvalue2
.Item("critical data") = 0
.Item("created" ) = Now()
End With
ds.Tables("myta ble").Rows.Add( dr)

Then, my update does something like;

da.Update(ds, "mytable")

All pretty basic stuff.... but what I would like is to ONLY insert new rows into the database that have a "criticalda ta" value <> 0 and discard the
rest. I could use the GetChanges method if I knew how to mark the rows as
'unchanged' immediately after creation - but don't know how.

Any help would be greatly appreciated.
Thanks,

Graham

[Graham Blandford]

I don't think I made myself very clear in my original post - either that or
I am reading the response wrongly....

I have an empty dataset.
In code, I add say 3 rows - "row a", "row b" and "row c" with say a single
value which I set to 0 (zero).

I am led to believe that at this point, these rows will all be marked as
"added".

The user - through the interface may or may not change the value of these
rows to say, 1.

Now, if I perform an update on the dataset I ONLY want the rows that have
been changed to a value of 1.

As far as I understand, the GetChanges method isn't going to help me,
because the rowstate on all of the rows will be 'added' and therefore any
'change' to a value will be ignored, as it already has an 'added' state.

Is this making sense?

Thanks,
Graham

"William Ryan eMVP" <do********@com cast.nospam.net > wrote in message
news:uU******** ******@tk2msftn gp13.phx.gbl...

YOu can use GetChanges and just get the rows with RowState or added. Then
call update on the GetChanges table - then acceptChanges on the original
--
W.G. Ryan MVP Windows - Embedded

Have an opinion on the effectiveness of Microsoft Embedded newsgroups?
Let Microsoft know!
https://www.windowsembeddedeval.com/...ity/newsgroups
"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:tH******** *************@n ews20.bellgloba l.com...

Hi all,

Would someone be able to tell me the most 'graceful' way of removing
unwanted rows from a dataset based on a condition prior to update? OR,
resetting the rows all to unchanged after they are initally added to the
recordset.

I create a dataset, which begins empty after the initial .Fill.

Then I create several rows with some default information, leaving one
critical piece of information value as 0;

dr = ds.Tables("myta ble").NewRow()
With dr
.Item("basicdat a1") = myvalue1
.Item("basicdat a2") = myvalue2
.Item("critical data") = 0
.Item("created" ) = Now()
End With
ds.Tables("myta ble").Rows.Add( dr)

Then, my update does something like;

da.Update(ds, "mytable")

All pretty basic stuff.... but what I would like is to ONLY insert new

rows

into the database that have a "criticalda ta" value <> 0 and discard the
rest. I could use the GetChanges method if I knew how to mark the rows as 'unchanged' immediately after creation - but don't know how.

Any help would be greatly appreciated.
Thanks,

Graham

[Jay B. Harlow [MVP - Outlook]]

Graham,

I am led to believe that at this point, these rows will all be marked as
"added". Correct, they will be marked as Added, you can call AcceptChanges on the
dataset & they will no longer be marked as Added.

Thus allowing you to call GetChanges to get the row that the user changed to
a 1.

If you want to look for rows without 0s, you could use DataTable.Selec t to
return the rows without 0s, then pass this array of rows to your
DataAdapter.

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:vr******** *************@n ews20.bellgloba l.com... I don't think I made myself very clear in my original post - either that or I am reading the response wrongly....

I have an empty dataset.
In code, I add say 3 rows - "row a", "row b" and "row c" with say a single
value which I set to 0 (zero).

I am led to believe that at this point, these rows will all be marked as
"added".

The user - through the interface may or may not change the value of these
rows to say, 1.

Now, if I perform an update on the dataset I ONLY want the rows that have
been changed to a value of 1.

As far as I understand, the GetChanges method isn't going to help me,
because the rowstate on all of the rows will be 'added' and therefore any
'change' to a value will be ignored, as it already has an 'added' state.

Is this making sense?

Thanks,
Graham

"William Ryan eMVP" <do********@com cast.nospam.net > wrote in message
news:uU******** ******@tk2msftn gp13.phx.gbl...

YOu can use GetChanges and just get the rows with RowState or added. Then

call update on the GetChanges table - then acceptChanges on the original
--
W.G. Ryan MVP Windows - Embedded

Have an opinion on the effectiveness of Microsoft Embedded newsgroups?
Let Microsoft know!
https://www.windowsembeddedeval.com/...ity/newsgroups
"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:tH******** *************@n ews20.bellgloba l.com...

Hi all,

Would someone be able to tell me the most 'graceful' way of removing
unwanted rows from a dataset based on a condition prior to update? OR,
resetting the rows all to unchanged after they are initally added to the recordset.

I create a dataset, which begins empty after the initial .Fill.

Then I create several rows with some default information, leaving one
critical piece of information value as 0;

dr = ds.Tables("myta ble").NewRow()
With dr
.Item("basicdat a1") = myvalue1
.Item("basicdat a2") = myvalue2
.Item("critical data") = 0
.Item("created" ) = Now()
End With
ds.Tables("myta ble").Rows.Add( dr)

Then, my update does something like;

da.Update(ds, "mytable")

All pretty basic stuff.... but what I would like is to ONLY insert new rows

into the database that have a "criticalda ta" value <> 0 and discard the rest. I could use the GetChanges method if I knew how to mark the

rows as 'unchanged' immediately after creation - but don't know how.

Any help would be greatly appreciated.
Thanks,

Graham

[Graham Blandford]

Thanks Jay,

I tried the former method, as it appears to be a more graceful answer - but
am a little confused. I call the AcceptChanges method once I have created my
new rows.

Then, after modifying one of the rows, I am doing something like this;

mynewdataset = myoriginaldatas et.GetChanges() ' I also tried
GetChanges(Data rowstate.Modifi ed)..
da.Update(mynew dataset, "mytable") ' Also tried .Update(myorigi nalset,
"mytable") ...

Now I get the error;

"An unhandled exception of type 'System.Data.DB ConcurrencyExce ption'
occurred in system.data.dll
Additional information: Concurrency violation: the UpdateCommand affected 0
records."

Am I doing something wrong? - Should I be using a different method of
update?

If I can't do the update using this method, you mention the
DataTable.Selec t - I can use this - but am unsure on how you would process
the array of rows against the dataadaptor... I'll go looking for this, but
if you have a moment to explain I'd very much appreciate it....
Thanks again,
Graham


"Jay B. Harlow [MVP - Outlook]" <Ja************ @msn.com> wrote in message
news:ek******** ******@tk2msftn gp13.phx.gbl...

Graham,

I am led to believe that at this point, these rows will all be marked as
"added". Correct, they will be marked as Added, you can call AcceptChanges on the
dataset & they will no longer be marked as Added.

Thus allowing you to call GetChanges to get the row that the user changed

to a 1.

If you want to look for rows without 0s, you could use DataTable.Selec t to
return the rows without 0s, then pass this array of rows to your
DataAdapter.

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:vr******** *************@n ews20.bellgloba l.com...

I don't think I made myself very clear in my original post - either that

or

I am reading the response wrongly....

I have an empty dataset.
In code, I add say 3 rows - "row a", "row b" and "row c" with say a single

value which I set to 0 (zero).

I am led to believe that at this point, these rows will all be marked as
"added".

The user - through the interface may or may not change the value of these rows to say, 1.

Now, if I perform an update on the dataset I ONLY want the rows that have been changed to a value of 1.

As far as I understand, the GetChanges method isn't going to help me,
because the rowstate on all of the rows will be 'added' and therefore any 'change' to a value will be ignored, as it already has an 'added' state.

Is this making sense?

Thanks,
Graham

"William Ryan eMVP" <do********@com cast.nospam.net > wrote in message
news:uU******** ******@tk2msftn gp13.phx.gbl...

YOu can use GetChanges and just get the rows with RowState or added.

Then call update on the GetChanges table - then acceptChanges on the original

--
W.G. Ryan MVP Windows - Embedded

Have an opinion on the effectiveness of Microsoft Embedded newsgroups?
Let Microsoft know!
https://www.windowsembeddedeval.com/...ity/newsgroups
"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:tH******** *************@n ews20.bellgloba l.com...
> Hi all,
>
> Would someone be able to tell me the most 'graceful' way of removing
> unwanted rows from a dataset based on a condition prior to update? OR, > resetting the rows all to unchanged after they are initally added to the > recordset.
>
> I create a dataset, which begins empty after the initial .Fill.
>
> Then I create several rows with some default information, leaving one > critical piece of information value as 0;
>
> dr = ds.Tables("myta ble").NewRow()
> With dr
> .Item("basicdat a1") = myvalue1
> .Item("basicdat a2") = myvalue2
> .Item("critical data") = 0
> .Item("created" ) = Now()
> End With
> ds.Tables("myta ble").Rows.Add( dr)
>
> Then, my update does something like;
>
> da.Update(ds, "mytable")
>
> All pretty basic stuff.... but what I would like is to ONLY insert new rows
> into the database that have a "criticalda ta" value <> 0 and discard the > rest. I could use the GetChanges method if I knew how to mark the

rows

as

> 'unchanged' immediately after creation - but don't know how.
>
> Any help would be greatly appreciated.
> Thanks,
>
> Graham
>
>

[Jay B. Harlow [MVP - Outlook]]

Graham,
Appearences can be conceiving ;-)

What are you calling AcceptChanges on?

Are you calling it before or after you change the zeros to a one?

Can you post an actual sample of what you are attempting?

I don't have a sample, however you should be able to:
- create your dataset
- add the rows with zeros
- dataset.acceptc hanges
- allow user to change zeros to one
- getchanges or datatable.selec t
- dataadapter.upd ate

Reading the help on DBConcurrencyEx ception, it sounds like you have a bad
update statement.
There is a overload on DataAdapter.Upd ate that accepts an array of DataRows
(from the DataTable.Selec t for example).

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:FW******** *************@n ews20.bellgloba l.com...

Thanks Jay,

I tried the former method, as it appears to be a more graceful answer - but am a little confused. I call the AcceptChanges method once I have created my new rows.

Then, after modifying one of the rows, I am doing something like this;

mynewdataset = myoriginaldatas et.GetChanges() ' I also tried
GetChanges(Data rowstate.Modifi ed)..
da.Update(mynew dataset, "mytable") ' Also tried .Update(myorigi nalset,
"mytable") ...

Now I get the error;

"An unhandled exception of type 'System.Data.DB ConcurrencyExce ption'
occurred in system.data.dll
Additional information: Concurrency violation: the UpdateCommand affected 0 records."

Am I doing something wrong? - Should I be using a different method of
update?

If I can't do the update using this method, you mention the
DataTable.Selec t - I can use this - but am unsure on how you would process
the array of rows against the dataadaptor... I'll go looking for this, but
if you have a moment to explain I'd very much appreciate it....
Thanks again,
Graham


"Jay B. Harlow [MVP - Outlook]" <Ja************ @msn.com> wrote in message
news:ek******** ******@tk2msftn gp13.phx.gbl...

Graham,

I am led to believe that at this point, these rows will all be marked as "added". Correct, they will be marked as Added, you can call AcceptChanges on the
dataset & they will no longer be marked as Added.

Thus allowing you to call GetChanges to get the row that the user changed to

a 1.

If you want to look for rows without 0s, you could use DataTable.Selec t to return the rows without 0s, then pass this array of rows to your
DataAdapter.

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:vr******** *************@n ews20.bellgloba l.com...

I don't think I made myself very clear in my original post - either that

or

I am reading the response wrongly....

I have an empty dataset.
In code, I add say 3 rows - "row a", "row b" and "row c" with say a single value which I set to 0 (zero).

I am led to believe that at this point, these rows will all be marked

as "added".

The user - through the interface may or may not change the value of

these rows to say, 1.

Now, if I perform an update on the dataset I ONLY want the rows that have been changed to a value of 1.

As far as I understand, the GetChanges method isn't going to help me,
because the rowstate on all of the rows will be 'added' and therefore any 'change' to a value will be ignored, as it already has an 'added' state.
Is this making sense?

Thanks,
Graham

"William Ryan eMVP" <do********@com cast.nospam.net > wrote in message
news:uU******** ******@tk2msftn gp13.phx.gbl...
> YOu can use GetChanges and just get the rows with RowState or added.

Then

> call update on the GetChanges table - then acceptChanges on the original >
>
> --
> W.G. Ryan MVP Windows - Embedded
>
> Have an opinion on the effectiveness of Microsoft Embedded newsgroups? > Let Microsoft know!
> https://www.windowsembeddedeval.com/...ity/newsgroups
> "Graham Blandford" <gr************ **@sympatico.ca > wrote in message
> news:tH******** *************@n ews20.bellgloba l.com...
> > Hi all,
> >
> > Would someone be able to tell me the most 'graceful' way of removing > > unwanted rows from a dataset based on a condition prior to update? OR, > > resetting the rows all to unchanged after they are initally added

to the

> > recordset.
> >
> > I create a dataset, which begins empty after the initial .Fill.
> >
> > Then I create several rows with some default information, leaving one > > critical piece of information value as 0;
> >
> > dr = ds.Tables("myta ble").NewRow()
> > With dr
> > .Item("basicdat a1") = myvalue1
> > .Item("basicdat a2") = myvalue2
> > .Item("critical data") = 0
> > .Item("created" ) = Now()
> > End With
> > ds.Tables("myta ble").Rows.Add( dr)
> >
> > Then, my update does something like;
> >
> > da.Update(ds, "mytable")
> >
> > All pretty basic stuff.... but what I would like is to ONLY insert new > rows
> > into the database that have a "criticalda ta" value <> 0 and

discard the

> > rest. I could use the GetChanges method if I knew how to mark the

rows

as
> > 'unchanged' immediately after creation - but don't know how.
> >
> > Any help would be greatly appreciated.
> > Thanks,
> >
> > Graham
> >
> >
>
>

[Graham Blandford]

Hi Jay,

Still no luck - I get the concurrency violation whether I overload the using
the .Select OR the GetChanges method.
The update statement looks fine, and indeed works if I do a straight forward
update without trying to filter out the zero-valued records.

Here's the code - inline with your requests.

Thanks for helping out.
Graham

- create your dataset Dataset is defined in the IDE along with the adaptor.
- add the rows with zeros
' Performed within a couple of nested loops;
drLinkMAN_crop_ operations =
dsLinkMAN_crop_ operations_1.Ta bles("crop_oper ations").NewRow ()
With drLinkMAN_crop_ operations
.Item("summit_g dcid") = summit_gdcid
.Item("summit_f arm_id") = drLinkMAN_field _trans.Item("su mmit_farm_id")
.Item("summit_f ield_id") = summit_field_id
.Item("year") = int_current_yea r
.Item("farm_id" ) = drLinkMAN_field _trans.Item("fa rm_id")
.Item("field_id ") = drLinkMAN_field _trans.Item("fi eld_id")
.Item("section_ id") = drLinkMAN_field _trans.Item("se ction_id")
.Item("crop_id" ) = 0
.Item("created" ) = Now()

dsLinkMAN_crop_ operations_1.Ta bles("crop_oper ations").Rows.A dd(drLinkMAN_cr o
p_operations)
End With
- dataset.acceptc hanges ' After the loop(s).
dsLinkMAN_crop_ operations_1.Ac ceptChanges()
- allow user to change zeros to one Done via the interface - works - data is correct if I don't try to
remove the zero-valued rows.
- getchanges or datatable.selec t
- dataadapter.upd ate
Dim dsMyChanges As New DataSet
dsMyChanges = dsLinkMAN_crop_ operations_1.Ge tChanges(DataRo wState.Modified )
If dsMyChanges.Has Changes Then
Dim intx As Int16
daLinkMAN_crop_ operations.Upda te(dsMyChanges) '<---------------------
HERE IS WHERE I GET THE VIOLATION ERROR
If dsMyChanges.Has Errors Then
Dim e As String = ""
End If
End If

"Jay B. Harlow [MVP - Outlook]" <Ja************ @msn.com> wrote in message
news:u%******** ********@TK2MSF TNGP09.phx.gbl. .. Graham,
Appearences can be conceiving ;-)

What are you calling AcceptChanges on?

Are you calling it before or after you change the zeros to a one?

Can you post an actual sample of what you are attempting?

I don't have a sample, however you should be able to:
- create your dataset
- add the rows with zeros
- dataset.acceptc hanges
- allow user to change zeros to one
- getchanges or datatable.selec t
- dataadapter.upd ate

Reading the help on DBConcurrencyEx ception, it sounds like you have a bad
update statement.
There is a overload on DataAdapter.Upd ate that accepts an array of DataRows (from the DataTable.Selec t for example).

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:FW******** *************@n ews20.bellgloba l.com...

Thanks Jay,

I tried the former method, as it appears to be a more graceful answer - but

am a little confused. I call the AcceptChanges method once I have created my

new rows.

Then, after modifying one of the rows, I am doing something like this;

mynewdataset = myoriginaldatas et.GetChanges() ' I also tried
GetChanges(Data rowstate.Modifi ed)..
da.Update(mynew dataset, "mytable") ' Also tried .Update(myorigi nalset,
"mytable") ...

Now I get the error;

"An unhandled exception of type 'System.Data.DB ConcurrencyExce ption'
occurred in system.data.dll
Additional information: Concurrency violation: the UpdateCommand affected

0

records."

Am I doing something wrong? - Should I be using a different method of
update?

If I can't do the update using this method, you mention the
DataTable.Selec t - I can use this - but am unsure on how you would

process the array of rows against the dataadaptor... I'll go looking for this, but if you have a moment to explain I'd very much appreciate it....
Thanks again,
Graham


"Jay B. Harlow [MVP - Outlook]" <Ja************ @msn.com> wrote in message news:ek******** ******@tk2msftn gp13.phx.gbl...

Graham,
> I am led to believe that at this point, these rows will all be marked as > "added".
Correct, they will be marked as Added, you can call AcceptChanges on

the dataset & they will no longer be marked as Added.

Thus allowing you to call GetChanges to get the row that the user changed

to

a 1.

If you want to look for rows without 0s, you could use DataTable.Selec t

to return the rows without 0s, then pass this array of rows to your
DataAdapter.

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:vr******** *************@n ews20.bellgloba l.com...
> I don't think I made myself very clear in my original post - either that or
> I am reading the response wrongly....
>
> I have an empty dataset.
> In code, I add say 3 rows - "row a", "row b" and "row c" with say a single

> value which I set to 0 (zero).
>
> I am led to believe that at this point, these rows will all be

marked as > "added".
>
> The user - through the interface may or may not change the value of these

> rows to say, 1.
>
> Now, if I perform an update on the dataset I ONLY want the rows that

have

> been changed to a value of 1.
>
> As far as I understand, the GetChanges method isn't going to help

me, > because the rowstate on all of the rows will be 'added' and therefore any

> 'change' to a value will be ignored, as it already has an 'added' state. >
> Is this making sense?
>
> Thanks,
> Graham
>
>
>
> "William Ryan eMVP" <do********@com cast.nospam.net > wrote in message
> news:uU******** ******@tk2msftn gp13.phx.gbl...
> > YOu can use GetChanges and just get the rows with RowState or

added. Then
> > call update on the GetChanges table - then acceptChanges on the

original

> >
> >
> > --
> > W.G. Ryan MVP Windows - Embedded
> >
> > Have an opinion on the effectiveness of Microsoft Embedded

newsgroups? > > Let Microsoft know!
> > https://www.windowsembeddedeval.com/...ity/newsgroups
> > "Graham Blandford" <gr************ **@sympatico.ca > wrote in message > > news:tH******** *************@n ews20.bellgloba l.com...
> > > Hi all,
> > >
> > > Would someone be able to tell me the most 'graceful' way of removing > > > unwanted rows from a dataset based on a condition prior to update? OR,

> > > resetting the rows all to unchanged after they are initally

added to the
> > > recordset.
> > >
> > > I create a dataset, which begins empty after the initial .Fill.
> > >
> > > Then I create several rows with some default information,

leaving

one

> > > critical piece of information value as 0;
> > >
> > > dr = ds.Tables("myta ble").NewRow()
> > > With dr
> > > .Item("basicdat a1") = myvalue1
> > > .Item("basicdat a2") = myvalue2
> > > .Item("critical data") = 0
> > > .Item("created" ) = Now()
> > > End With
> > > ds.Tables("myta ble").Rows.Add( dr)
> > >
> > > Then, my update does something like;
> > >
> > > da.Update(ds, "mytable")
> > >
> > > All pretty basic stuff.... but what I would like is to ONLY

insert new

> > rows
> > > into the database that have a "criticalda ta" value <> 0 and

discard the
> > > rest. I could use the GetChanges method if I knew how to mark

the rows
> as
> > > 'unchanged' immediately after creation - but don't know how.
> > >
> > > Any help would be greatly appreciated.
> > > Thanks,
> > >
> > > Graham
> > >
> > >
> >
> >
>
>

[Graham Blandford]

Jay. I think I know the problem.

Would I be correct in thinking that when then I AcceptChanges on the
dataset, and then make a change, the row is marked as 'Modified' - and
subsequently calls the UpdateCommand, when in fact it needs to be calling
the InsertCommand?? ...

Cant believe it never clicked before... now I have to figure out how to
filter WITHOUT using the AcceptChanges. I guess I could use the
OnRowUpdating Handler - will that allow me to ignore a row if I mark it?

Thanks,
Graham

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:j6******** *************@n ews20.bellgloba l.com...

Hi Jay,

Still no luck - I get the concurrency violation whether I overload the using the .Select OR the GetChanges method.
The update statement looks fine, and indeed works if I do a straight forward update without trying to filter out the zero-valued records.

Here's the code - inline with your requests.

Thanks for helping out.
Graham

- create your dataset Dataset is defined in the IDE along with the adaptor.

- add the rows with zeros

' Performed within a couple of nested loops;
drLinkMAN_crop_ operations =
dsLinkMAN_crop_ operations_1.Ta bles("crop_oper ations").NewRow ()
With drLinkMAN_crop_ operations
.Item("summit_g dcid") = summit_gdcid
.Item("summit_f arm_id") = drLinkMAN_field _trans.Item("su mmit_farm_id")
.Item("summit_f ield_id") = summit_field_id
.Item("year") = int_current_yea r
.Item("farm_id" ) = drLinkMAN_field _trans.Item("fa rm_id")
.Item("field_id ") = drLinkMAN_field _trans.Item("fi eld_id")
.Item("section_ id") = drLinkMAN_field _trans.Item("se ction_id")
.Item("crop_id" ) = 0
.Item("created" ) = Now()

dsLinkMAN_crop_ operations_1.Ta bles("crop_oper ations").Rows.A dd(drLinkMAN_cr o p_operations)
End With

- dataset.acceptc hanges ' After the loop(s).
dsLinkMAN_crop_ operations_1.Ac ceptChanges()

- allow user to change zeros to one

Done via the interface - works - data is correct if I don't try to
remove the zero-valued rows.

- getchanges or datatable.selec t
- dataadapter.upd ate

Dim dsMyChanges As New DataSet
dsMyChanges =

dsLinkMAN_crop_ operations_1.Ge tChanges(DataRo wState.Modified ) If dsMyChanges.Has Changes Then
Dim intx As Int16
daLinkMAN_crop_ operations.Upda te(dsMyChanges) '<---------------------
HERE IS WHERE I GET THE VIOLATION ERROR
If dsMyChanges.Has Errors Then
Dim e As String = ""
End If
End If

"Jay B. Harlow [MVP - Outlook]" <Ja************ @msn.com> wrote in message
news:u%******** ********@TK2MSF TNGP09.phx.gbl. ..

Graham,
Appearences can be conceiving ;-)

What are you calling AcceptChanges on?

Are you calling it before or after you change the zeros to a one?

Can you post an actual sample of what you are attempting?

I don't have a sample, however you should be able to:
- create your dataset
- add the rows with zeros
- dataset.acceptc hanges
- allow user to change zeros to one
- getchanges or datatable.selec t
- dataadapter.upd ate

Reading the help on DBConcurrencyEx ception, it sounds like you have a bad

update statement.
There is a overload on DataAdapter.Upd ate that accepts an array of

DataRows

(from the DataTable.Selec t for example).

Hope this helps
Jay

"Graham Blandford" <gr************ **@sympatico.ca > wrote in message
news:FW******** *************@n ews20.bellgloba l.com...

Thanks Jay,

I tried the former method, as it appears to be a more graceful answer -

but

am a little confused. I call the AcceptChanges method once I have created

my

new rows.

Then, after modifying one of the rows, I am doing something like this;

mynewdataset = myoriginaldatas et.GetChanges() ' I also tried
GetChanges(Data rowstate.Modifi ed)..
da.Update(mynew dataset, "mytable") ' Also tried .Update(myorigi nalset,
"mytable") ...

Now I get the error;

"An unhandled exception of type 'System.Data.DB ConcurrencyExce ption'
occurred in system.data.dll
Additional information: Concurrency violation: the UpdateCommand

affected

0

records."

Am I doing something wrong? - Should I be using a different method of
update?

If I can't do the update using this method, you mention the
DataTable.Selec t - I can use this - but am unsure on how you would

process the array of rows against the dataadaptor... I'll go looking for this, but if you have a moment to explain I'd very much appreciate it....
Thanks again,
Graham


"Jay B. Harlow [MVP - Outlook]" <Ja************ @msn.com> wrote in message news:ek******** ******@tk2msftn gp13.phx.gbl...
> Graham,
> > I am led to believe that at this point, these rows will all be marked

as

> > "added".
> Correct, they will be marked as Added, you can call AcceptChanges on

the > dataset & they will no longer be marked as Added.
>
> Thus allowing you to call GetChanges to get the row that the user

changed

to
> a 1.
>
> If you want to look for rows without 0s, you could use DataTable.Selec t

to

> return the rows without 0s, then pass this array of rows to your
> DataAdapter.
>
> Hope this helps
> Jay
>
> "Graham Blandford" <gr************ **@sympatico.ca > wrote in message
> news:vr******** *************@n ews20.bellgloba l.com...
> > I don't think I made myself very clear in my original post -

either that

> or
> > I am reading the response wrongly....
> >
> > I have an empty dataset.
> > In code, I add say 3 rows - "row a", "row b" and "row c" with say

a single
> > value which I set to 0 (zero).
> >
> > I am led to believe that at this point, these rows will all be

marked

as

> > "added".
> >
> > The user - through the interface may or may not change the value of these
> > rows to say, 1.
> >
> > Now, if I perform an update on the dataset I ONLY want the rows that have
> > been changed to a value of 1.
> >
> > As far as I understand, the GetChanges method isn't going to help

me, > > because the rowstate on all of the rows will be 'added' and therefore any
> > 'change' to a value will be ignored, as it already has an 'added'

state.

> >
> > Is this making sense?
> >
> > Thanks,
> > Graham
> >
> >
> >
> > "William Ryan eMVP" <do********@com cast.nospam.net > wrote in message > > news:uU******** ******@tk2msftn gp13.phx.gbl...
> > > YOu can use GetChanges and just get the rows with RowState or added. > Then
> > > call update on the GetChanges table - then acceptChanges on the
original
> > >
> > >
> > > --
> > > W.G. Ryan MVP Windows - Embedded
> > >
> > > Have an opinion on the effectiveness of Microsoft Embedded

newsgroups?

> > > Let Microsoft know!
> > > https://www.windowsembeddedeval.com/...ity/newsgroups
> > > "Graham Blandford" <gr************ **@sympatico.ca > wrote in message > > > news:tH******** *************@n ews20.bellgloba l.com...
> > > > Hi all,
> > > >
> > > > Would someone be able to tell me the most 'graceful' way of

removing

> > > > unwanted rows from a dataset based on a condition prior to update? OR,
> > > > resetting the rows all to unchanged after they are initally added

to

> the
> > > > recordset.
> > > >
> > > > I create a dataset, which begins empty after the initial ..Fill. > > > >
> > > > Then I create several rows with some default information,

leaving one
> > > > critical piece of information value as 0;
> > > >
> > > > dr = ds.Tables("myta ble").NewRow()
> > > > With dr
> > > > .Item("basicdat a1") = myvalue1
> > > > .Item("basicdat a2") = myvalue2
> > > > .Item("critical data") = 0
> > > > .Item("created" ) = Now()
> > > > End With
> > > > ds.Tables("myta ble").Rows.Add( dr)
> > > >
> > > > Then, my update does something like;
> > > >
> > > > da.Update(ds, "mytable")
> > > >
> > > > All pretty basic stuff.... but what I would like is to ONLY insert new
> > > rows
> > > > into the database that have a "criticalda ta" value <> 0 and

discard

> the
> > > > rest. I could use the GetChanges method if I knew how to mark the > rows
> > as
> > > > 'unchanged' immediately after creation - but don't know how.
> > > >
> > > > Any help would be greatly appreciated.
> > > > Thanks,
> > > >
> > > > Graham
> > > >
> > > >
> > >
> > >
> >
> >
>
>