TRIM

[panic]

Hi all

I HAVE to trim trailing special character |
or just replace with blank while trimimg
can any one please let me know the syntax



how to use exactly
Trim(%string%,[%stripchar%],[%option%])

thanks

[kris007]

What is your source? The special character you are talking about is LineFeed or are you talking about a PIPE.You cannot use TRIM function in order to trim a LF(LineFeed) it. If you can let us know what your source is we can then go about it.

[avi21st]

Hi Panic :D

First thing - do you know what character you are having. What you see as special character in your Editor is not the true identifier of what is coming in data...If you are sure of the special chatacer- Then yes, you can do a field level modification by using TRIM

Code: Select all

Use TRIM( linkname.columnname, 'Character to replace','A')

OR you can use a before job routine (ExecSh) to change all the characters in the input file.

For example this one is used to remove non ascii fields:

Code: Select all

tr '\0' ' ' < #TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME# > #TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME_CLN#

where

tr is for translate in UNIX
'\0' is the character to replace
' ' is the character by which you want to replace
< #TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME# > is the input file
#TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME_CLN# is the cleaned file

Let me know if you are still panicking!!!!

[meena]

Hi Panic...
Can you let us know what is your source and what your job is doing...The special character looks like a pipe..and if you want to replace the character by blank then I think we can use 'Replace' function..

Hi all

I HAVE to trim trailing special character |
or just replace with blank while trimimg
can any one please let me know the syntax



how to use exactly
Trim(%string%,[%stripchar%],[%option%])

thanks

[ray.wurlod]

Either the on-line Help or the DataStage BASIC manual will give you complete descriptions of the Trim() function. You could also use the string_trim() function in a Modify stage.

[Krazykoolrohit]

Make sure you are not working with a pipe delimited file ....lol

[anu123]

Hi:

How to eliminate a non-ascii charecter dynamically? I mean to say...I do not know what non-ascii char I would get in my source file but i need to search for non- ascii chars and clean up the file before loading into target table.

Basically my source file is from SAP. So, I can say non-ascii char would be a Korean/Japanese/chinese language specific.

I am not very good at Shell scripting. please help me out solving this.

thanks,

Hi Panic :D

First thing - do you know what character you are having. What you see as special character in your Editor is not the true identifier of what is coming in data...If you are sure of the special chatacer- Then yes, you can do a field level modification by using TRIM

Code: Select all

Use TRIM( linkname.columnname, 'Character to replace','A')

OR you can use a before job routine (ExecSh) to change all the characters in the input file.

For example this one is used to remove non ascii fields:

Code: Select all

tr '\0' ' ' < #TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME# > #TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME_CLN#

where

tr is for translate in UNIX
'\0' is the character to replace
' ' is the character by which you want to replace
< #TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME# > is the input file
#TARGET_FILE_DIRECTORY#/#OUTPUT_FILE_NAME_CLN# is the cleaned file

Let me know if you are still panicking!!!!

[ray.wurlod]

If they are Chinese-Japanese-Korean (CJK) characters are you sure you want to eliminate them? They are probably valid data. More likely you need to preserve them, would be my guess.

This means that you need DataStage National Language Support (NLS) installed, configured and the appropriate character map(s) chosen. Read more about it in the DataStage NLS manual. There is overlap of some characters in the CJK encodings.

[anu123]

I am sorry for the confusion Ray. Actually we have NLS installed and configured. I mentioned Chinese-japanese-korean as an example to explain the problem. But I do not know what type of non-ascii I would be getting.

-thanks.

If they are Chinese-Japanese-Korean (CJK) characters are you sure you want to eliminate them? They are probably valid data. More likely you need to preserve them, would be my guess.

This means that you need DataStage National Language Support (NLS) installed, configured and the appropriate character map(s) chosen. Read more about it in the DataStage NLS manual. There is overlap of some characters in the CJK encodings.

[ray.wurlod]

Maybe you could test the rightmost character with IsValid() to determine whether it is a valid string. A character that is valid as a ustring but not as a string might fall into your "non-ASCII" class.