DSXchange

Here are some more: Begin Case Case SectionName = 'HEADER' and FirstWord = 'Date' NewDsxLine = NewDsxLine[1, BeginOfSecondWord-1] : DQuote(Today) Case OLEType = 'CTrxOutput' and FirstWord = 'Columns' gosub GenTrxOutput Case OLEType = 'CCustomOutput' and SecondWord = '"CCustomProperty&qu...

John Here are the types of things I was looking for: Begin Case Case SectionName = 'HEADER' and FirstWord = 'Date' NewDsxLine = NewDsxLine[1, BeginOfSecondWord-1] : DQuote(Today) Case Identifier = 'ROOT' and FirstWord = 'Description' NewDsxLine = NewDsxLine[1, BeginOfSecondWord-1] : '"'...

narasimha wrote:Did an exact search found 75 matches for @RECORD.
You can find a nice explaination here

You found the right post. You need to work this post in conjunction with a list of all the parameters within a job. This could take a search for each parameter. Lot of work.

We use a script which reads paramaeter values from files in the OS and then runs the job. We built these files from the dsjob command. If you cannot figure out the syntax then I will post it.

I posted some code which would search @RECORD for any value. I think it would be easy to modify this code to get what you want. It does return the link name. It had either upcase or downcase in the code. If you cannot find it let me know.

Ken Bland posted a job on his web site which will set all purge settings to the same value. I expect that might solve your problem.

I wrote my own. When I wrote Parameter Navigator it became difficult to fix all the sequences. We had 500 jobs and over 100 sequences. Each job added at least 3 parameters for each database touched. I wrote one job to fix all ODBC jobs. So I fixed one job for each database touched. Next I copied all...

We are on AIX and our numbers are 3 times too big. Is this a bug or a setting problem?

Now you have all the pieces to put together a solution. You need to figure out the column types. They differ by OLETYPE. OLETYPE represents the stage type. A column type of 1 in ODBC maybe interger but in a Oracle 8 stage this maybe a varchar. You have to figure this out by looking in Designer and s...

Next you need to read the metadata you need from DS_METADATA and substitute it into the DSX for each column. * ------------------------------------------------------------ open 'DS_METADATA' to DsMetadata else ErrMsg = "Error: Unable to open DS_METADATA." gosub ErrRtn goto TheEnd end DsMet...

Next you need to find the column metadata in your DSX. Once you find where columns start in the DSX file then you need to strip out the old columns and insert the new columns. * ------------------------------------------------------------ read DsxRec from KGen, DsxName then NoOfLines = dcount(Ds...

Let me explain how I did this. First you assume a straight table copy. This means source column names equal target column names. Next you assume the metadata stays the same. I save all the table definitons by importing them. These get stored in DS_METADATA. Ray posted a routine which exposed the dat...

John

I have been so busy, I forgot you sent me your example DSX. Hopefully I can work on it tomorrow.

Geez, Craig you could of picked a better post.

Please try again, Ray.