SSIS column mappings must be done in "design time". We humans do this manually in SSIS. To do it programmatically means manipulating a template SSIS pkg and then executing that...

Here's...

Another gah!!! You'll need to grab the columns for a specific table... doh!

foreach(DataRow t in tbls.Rows) {

sql.Clear();

sql.AppendFormat("create table {0}(", t["TABLE_NAME"]);

...

Gah. you'll want this in the first for loop:

foreach (DataRow t in ...) {

sql.Clear();

...

Here's one way, using ADO.Net + ODBC connection... (this is more or less from my working SSIS 2008/2010 ScriptTask C# code...)

The methodology is similar for ADO.Net + SqlConnection. SqlConnection also...

xp_cmdshell has some serious security implications. In that case, it's easier to not deal with that particular set of landmines and just implement that functionality through SSIS.

Last place I worked used DataJunction for the print service bureau side of the business (e.g., bank check printing, etc). While what DJ does is cool, it does store the...

...and if you're using OleDB connection managers, the EzAPI code seems to work to create SSIS packages as well.

Here's the latest source on CodePlex:

http://sqlsrvintegrationsrv.codeplex.com/SourceControl/list/changesets

The problem I had with the 2008...

This link might be helpful too...

http://bifuture.blogspot.com/2011/01/ssis-adding-derived-column-to-ssis.html

Hmm... The features mentioned in the article are done or doable by SSMS Tools Pack (sql snippets)...

I'll bite, as I use regex's much for cases like this (again, on recordsets > million_rows, too).

The patindex is fine, but now, isolate the matched substring.

The regex functions on my...

I've used our own in-house implementations of some of the .Net RexEx functions on seemingly large datasets (several million records each), mostly to help clean up, validate and parse addresses,...

I have no problem RDP'ing into my desktop at work. Then my work stuff can stay at work, and I don't have to worry about the responsibilities and repercussions of...

Hmm... it's a double-edged sword sometimes, too... Just ask our (ex-)spouses...

one way is to just do something like...

[font="Courier New"]select acct_num, icd9_code,

row_number() over (partition by acct_num order by acct_num, icd9_code) as rn

[/font]

Put this in a sub-select or temp...

Hmm... ICD-9 codes... There's a bit more than 1024 different ICD-9 "base" codes...

but maybe he's only going to possibly have no more than 1023 distinct codes over a population (not...