Filtering special characters

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
bensonian
Participant
Posts: 42
Joined: Tue Nov 22, 2005 2:12 pm

Filtering special characters

Post by bensonian »

The source data pull has special charecters, i was wondering if there is a way we could filter that out in the DataStage job. There could be a lot of different special charecters coming in.

Thanks alot gurus
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

Define "special" in this context.
-craig

"You can never have too many knives" -- Logan Nine Fingers
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

"Alphanumeric" just means a character that is a letter or a number. As Ray notes, you need to tighten down your requirements first.
-craig

"You can never have too many knives" -- Logan Nine Fingers
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

Why not try to get them to load without aborting? Perhaps if you posted the errors generated, we could help make that happen.
-craig

"You can never have too many knives" -- Logan Nine Fingers
senthilt1
Participant
Posts: 134
Joined: Mon Nov 19, 2007 2:17 am

Post by senthilt1 »

Hi,

Even i had the same issue removing the special characters from this input records. We have written a script that search for special characters in the input records and replaces all the special characters with emptyspace. We called this script in external filter or external source depends on the flow. This is how we are doing.

I am not sure whether this is the right way..........FYI.

Thanks,

Senthil P
bensonian
Participant
Posts: 42
Joined: Tue Nov 22, 2005 2:12 pm

Post by bensonian »

chulett wrote:Why not try to get them to load without aborting? Perhaps if you posted the errors generated, we could help make that happen. ...



The error is:

sqlcode is: -12899
esql complaint: ORA-12899: value too large for column "<schema_name>"."<table_name"."<column_name>" (actual: 32, maximum: 30)

the length of <column_name> is varchar(30).

I actually identified the record based upon row_num and
output > file

try to insert into db again. When i was viewing it, from file>view data. DS was actually adding extra bytes to it and it's not the same when i view the file on unix box. The way DS is interpreting it is wierd, so i happened to check the NLS map, which is not turned on at the project level.

Is it the 'NLS map' thats causing the whole issue. Please advise me of how this ('NLS') could be turned on at project level.

Appreciate all your help.
file >
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

NLS is enabled, and default map selected for a project, using the Administrator client NLS button once that project has been selected.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
bensonian
Participant
Posts: 42
Joined: Tue Nov 22, 2005 2:12 pm

Post by bensonian »

ray.wurlod wrote:NLS is enabled, and default map selected for a project, using the Administrator client NLS button once that project has been selected. ...
Ray,

surprisingly for the none of the project's in Administrator 'NLS' tab is neither highlighted or enabled. So how would DS do the charecter mapping between DB and DS engine and how does it interpret the 'accented characters'.
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

What is the value of the NLSMODE parameter in your uvconfig file?
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Post Reply