DSXchange

I seem to recall that when loading oracle, the empty string is interpreted as a null?

kumar_s wrote:Replace each pipe with CHAR(13):AccountId:','
And AccountId:',' to the first Delimited field of WORDLIST.

Here teh CHAR(13) is the new line character.

How would you do that? There is no eReplace in parallel and convert only works for one char.

Does it work correctly when BUSINESS_FG = 'U'?

IF BUSINESS_FG = "G" THEN IF (PLAN_CD1 = "CES" OR PLAN_CD2 = "CES") THEN VAL = "ES" END ELSE VAL = "DM" END END IF (PLAN_CD1 = "CAT" OR PLAN_CD2 = "CAT") THEN IF ( RIDER_FORM_CD = "ES" OR RIDER_FORM_CD = "09" ...

chulett wrote:It should be fine as long as you know the max ... And it doesn't need to be a fixed number

I was unsure whether the column import needed to find a delimiter for say the 10th column, or whether after the 4th delimiter it is just implied and the column generated anyway.

You could use the column import stage, setting the delimiter to | to separate the second column into multiple columns. Follow this with a pivot stage.

This might only work though if the number of delimiters is constant.

Are you sure your input = 'G' and not 'G ' or some other hidden character?

Search the forum for "Vertical Pivot".

I think there is an explanation in the FAQ section also

Could you also use something like: Index("1,2,3,4,5,6,7,8,9",InputCol.String,1) to get the first required digit then use the field function with the newly acquired digit as your starting point to select 1 char? Would this carry more processing overhead? Addendum : Actually this won't work ...

Is there any message in the Director log? Perhaps something about Job AppMon not running?

While i tried to write the data into another sequential file of DOS style format from the sequential file of UNIX style(aforementioned file), the data in that file is not as same as the source sequential file(UNIX style).

How is it "not the same". What differences are there?

ray.wurlod wrote:If you are running this from an after-job subroutine, then that job is not finished so you can never get the "job finished" message from the job log.

Oh yeah. You could try DSJ.JOBINTERIMSTATUS then?

Meanwhile, is there a way that I can get the "Control" statement which says "Job X Finished" or "Job X aborted" by editing the existing code above. Wouldn't it be something to do with Result = DSGetJobInfo (JobHandle, DSJ.JOBSTATUS) Then you can use the status that are...

It would probably be easier to use Unix (in the filter command field) and do a split on on the output file ,based on a certain number of rows that would be under 500MB, after is finished. That way it doesn't matter how big the output is there will always be an overflow file. **Post Script Edit** Dam...

Please delete one of your posts on this topic. Firstly - the actual value of one field "1234, 1234-01, 1234-02, 236789, 45467, 45456-03.5645.45456-02, 45456-01, 45456" or does that represent multiple rows? If multiple rows then using Field with the delimiter set to "-" and select...