Datasets - Cloning problems

[P K]

Hi,

I have 3 datasets
d1Main is the main dataset having rows.
I loop through rows in d1main and any row failing validations is moved to
d2Bad dataset. I had to clone d2bad as d1main and then add failed row to
d2bad

so d2bad is created out of d1main.

Now comes d3good. any records valid in d1main are moved to d3good. again
this was done by cloning d1main for d3good.

Now i use d3good to update the database. The moment database error occurs. I
have to move row in d3good to d2bad.

Moving is failing as it says row already exists. Already d3good is cloned
out of d1main. I cannot clone it again out of d2bad.

how do I do this??

[sloan]

I wouldn't clone.

I'd do this
mydataset dsAllData = new mydataset();

mydataset dsPotentiallyGo odDS = new mydataset();
mydataset dsBadData = new mydataset();
Loop over dsAllData

when an error occurs then do this

for each alldatarow in dsAllData.Selec t("");

mydataset.mytab le.myrow badrow = dsBadData.mytab le.Newmyrow();
badrow.a = alldatarow.a;
//continue on all properties

dsBadData.mytab le.Addmyrow( badrow );

next alldatarow

something like that.. the syntax will be wrong.

The key is..... add a NEW row to the bad dataset or potentiallGood dataset
as you encounter them.
...


"P K" <PK@discussions .microsoft.com> wrote in message
news:19******** *************** ***********@mic rosoft.com...

Hi,

I have 3 datasets
d1Main is the main dataset having rows.
I loop through rows in d1main and any row failing validations is moved to
d2Bad dataset. I had to clone d2bad as d1main and then add failed row to
d2bad

so d2bad is created out of d1main.

Now comes d3good. any records valid in d1main are moved to d3good. again
this was done by cloning d1main for d3good.

Now i use d3good to update the database. The moment database error occurs. I have to move row in d3good to d2bad.

Moving is failing as it says row already exists. Already d3good is cloned
out of d1main. I cannot clone it again out of d2bad.

how do I do this??

[P K]

I wanted to avoid setting value of each column.
And so had to clone and Import.
Columns are not defined. The main dataset could have different kind of data.
Not the same columns always.

To be more precise, I load the dataset from client files. Processing is same
for different files. But the columns in the files could be different.
"sloan" wrote:

I wouldn't clone.

I'd do this
mydataset dsAllData = new mydataset();

mydataset dsPotentiallyGo odDS = new mydataset();
mydataset dsBadData = new mydataset();
Loop over dsAllData

when an error occurs then do this

for each alldatarow in dsAllData.Selec t("");

mydataset.mytab le.myrow badrow = dsBadData.mytab le.Newmyrow();
badrow.a = alldatarow.a;
//continue on all properties

dsBadData.mytab le.Addmyrow( badrow );

next alldatarow

something like that.. the syntax will be wrong.

The key is..... add a NEW row to the bad dataset or potentiallGood dataset
as you encounter them.
...


"P K" <PK@discussions .microsoft.com> wrote in message
news:19******** *************** ***********@mic rosoft.com...

Hi,

I have 3 datasets
d1Main is the main dataset having rows.
I loop through rows in d1main and any row failing validations is moved to
d2Bad dataset. I had to clone d2bad as d1main and then add failed row to
d2bad

so d2bad is created out of d1main.

Now comes d3good. any records valid in d1main are moved to d3good. again
this was done by cloning d1main for d3good.

Now i use d3good to update the database. The moment database error occurs.

I

have to move row in d3good to d2bad.

Moving is failing as it says row already exists. Already d3good is cloned
out of d1main. I cannot clone it again out of d2bad.

how do I do this??

[sirfunusa]

Don't do that. Use T-SQL in a stored proc for everything. Load the
client files to a temp table, and process on the server.