pad string

[RK72]

Hi All,

Convert the source records into target format using transformer. source record should turn into target format even if we get new records in future.

Source records:,................. Target Format:
5 ............... 0005
75 ................ 0075
1275 ................ ...... 1275
189 ................. 0175
in this scenario i am using padstring but i got the 0 records from my right side like 5000 how to get 0 recrds leftside like 0005

Thanks

[chulett]

Show us your syntax.

[RK72]

hi..
my syntax is

If length < 4 Then PadString(DSLink3.SOURCERECORDS,0,4-length) Else DSLink3.SOURCERECORDS

[chulett]

What are the data types for your source and target fields? Strings?

[RK72]

Hi
I am using sequentialfiles

[jwiles]

What are the data types of the source and target columns, not the files!

[Kryt0n]

You could do a Right("0000":<field>,4)

or maybe even a PadString("",0,4-length) : <field>?

According to the help files, padstring does append

[chandra.shekhar@tcs.com]

Or you may use this too.

Code: Select all

 STR(0 , 4 - Len(Trim(Col1))) : Trim(Col1)

[ray.wurlod]

General question to all contributors to this thread: which of the suggested approaches is most efficient (fewest CPU cycles) and which is least efficient?

[asorrell]

I wasn't a contributor, but I believe kryt0n's "Right" solution is the cheapest in terms of CPU cycles.

"If" statements are always expensive, and length statements which need to count the string take up cycles as well.

I usually pad using kryt0n's method as the most effective. If not, I'd love to hear which one is better!

I know I've seen worse...

I saw one job done by a newbie developer that was told they needed to pad about twenty five of the fields to a length of twenty. The job wasn't performing well and when I went in I found each of the fields had code like this...


If (Len(link.column1) = 0) Then "00000000000000000000" Else If (Len(link.column1) = 1) Then "0000000000000000000":link.column1 Else if .... you get the idea....

[priyadarshikunal]

but why not change the data type to number(4) if OP is using sequential file only.

[Kryt0n]

Couldn't say which is the most efficient, never delved in to the inner workings of how it actually defines any function... and it would take a hell of a lot of records to really make a significant difference to really bother figuring out. (Note: not saying I don't consider performance when developing but don't spend much time trying to save a couple of seconds)

If I were to eke out as much performance as possible, I would probably define a stage variable intialised to "0000" and use that in my right() to avoid too many string initialisations... but still wouldn't know if that was the best option

[abc123]

Ray, this is old thread but your question is very interesting. I know that I can use APT_* variables to get some values in the log to check performance or just look at job run time but would like to hear from you.

How would you:

a) ..go around seeing how many CPU cycles it is using?
b) ..find the least/most efficient method?

[ray.wurlod]

I would create jobs to use each, and process a statistically significant number (millions) of rows through each, controlling for things like caching. Use the player timings and other environment variables, and/or DataStage performance analyzer, and/or Operations Console to perform the measurement.

[ssnegi]

You could do a Right("0000":<field>,4)

or maybe even a PadString("",0,4-length) : <field>?

According to the help files, padstring does append

Instead of hardcoding the 0 use Str function to generate the zeros equal to the length of field.

Code: Select all

right(Str('0',4):<field>,4)