DATA Step, Macro, Functions and more

Symbol conversion to special character

Reply
Occasional Contributor
Posts: 12

Symbol conversion to special character

HI All,

 

The given string "INFECTIONS -TUBO - NEW OR WORSENING" was converted automatically into "INFECTIONS -TUBO €“ NEW OR WORSENING" while connecting SAS to Oracle database. In SAS datasets it is showing right "INFECTIONS -TUBO - NEW OR WORSENING" and in oracle table it is sowing "INFECTIONS -TUBO €“ NEW OR WORSENING".

 

Can you please suggest how to prevent such type of conversion ? 

 

Thank you in advance !!

Super User
Posts: 9,547

Re: Symbol conversion to special character

Posted in reply to singhsahab

Use a $hex. format, so you can see what is actually stored. Your second hyphen is most probably not a standard hyphen, but a UTF character.

---------------------------------------------------------------------------------------------
Maxims of Maximally Efficient SAS Programmers
How to convert datasets to data steps
How to post code
PROC Star
Posts: 2,211

Re: Symbol conversion to special character

[ Edited ]
Posted in reply to singhsahab

So many answers already on the web.

 

Like 

https://stackoverflow.com/questions/9773995/bad-encoding-why-does-my-medium-sized-dash-is-differentl...

http://community.teradata.com/t5/Database/Seeing-Junk-character-when-inserting-hyphen/td-p/27271

 

From the second link: "That combination of characters is the 3-byte UTF-8 representation of EN DASH U+2013, misinterpreted as Windows-1252. In short, you are loading UTF-8 data using ASCII as the session character set."

 

As @KurtBremser said, you are mixing encodings.

Either you match the encodings of your Oracle data and your SAS data, or more likely you just clean your SAS strings to remove any non-latin characters. 

 

This should work, I can't test as my session is WLATIN1.

 

 STR=prxchange('s/[^[:ascii:]]/ /', -1, STR );

 

Ask a Question
Discussion stats
  • 2 replies
  • 95 views
  • 1 like
  • 3 in conversation