Tricky Question

[Champa]

All,

There are two columns as follows in a sequential file.
I am using unix to write this output. And I am reading one line at a time.

col A Col B

a 1
a 2
a 3

Output should look like

Heading a
1
2
3

Can you please help me.

Thanks

[prabu]

All,

There are two columns as follows in a sequential file.
I am using unix to write this output. And I am reading one line at a time.

col A Col B

a 1
a 2
a 3

Output should look like

Heading a
1
2
3

Can you please help me.

Thanks

not so tricky answer

Code: Select all

cat myfilename.dat | awk -F, '{if (NR == 1) {print $1"\n"$2}  else print $2}'

[ray.wurlod]

Similarly in DataStage. Your single output column is derived as

Code: Select all

If @INROWNUM = 1 Then InLink.ColA Else InLink.ColB

[Champa]

Ray & Prabu,

Thanks

I just have bunch of ids in colA & col B.

colA colB
a 1
a 2
a 3
b 1
b 2

like that.

I have to concatenate some other fileds also.

The way I am doing is

Readseq Line

Extract diff fields

unix<-1> = F1
unix<-1> = F2

Heading a
1
2
3

then I append some more data & use mailx to send email.

loop ends here.

And I start all over for b.

Please suggest a solution for this.

In a way i need to put few lines before & after this list only once which is making my approach hard.

Thanks in advance.

[ray.wurlod]

We answered the question you posed.

The "bunch of IDs" in ColA is a new question. It is solved by sorting the input by ColA and using stage variables to detect the changes therein and to accumulate/reset the counter (ColB).

The technique has been described a number of times; search the forum for details.

Importantly, make sure that the data are partitioned using an algorithm that keeps all values in any one group in the same partition as each other.