Linux

Hi,

I have an csv file and there are some non printable characters(extended ascii) so I am trying to create a clean copy of the csv file . I am using
this command:

Code:

tr -cd "[:print:]" < /opt/informatica/PowerCenter8.6.0/server/infa_shared/SrcFiles/ThirdParty/locations.csv  > 
/opt/informatica/PowerCenter8.6.0/server/infa_shared/SrcFiles/locations.csv ...

or

Code:

tr -cd '\11\12\40-\176' < /opt/informatica/PowerCenter8.6.0/server/infa_shared/SrcFiles/ThirdParty/locations.csv 
 > /opt/informatica/PowerCenter8.6.0/server/infa_shared/SrcFiles/locations2.csv

But this command creates just empty output file.
What do you think?

have you tried the [:cntrl:] option?
if you

Code:

man tr

you will probably see a listing of all the special sets, like

Code:

[:print:] & [:alpha:]

Actually I am using informatica command transformation. I am just adding this shell overthere

Not sure of your 2nd command, but your first should send output to a different filename. Unix commands act funny when the input and output filenames are the same.

You might want to pipe it to a pager like more or less to see what it's doing, but definitely use both a distinct output file and note (unless my eyes deceive me...) your tr command seems to be deleting the printing chars you're looking to keep. Try to reverse the set using the full octal values (ie, \000-\011, \013-\014, etc)...

---------- Post updated at 16:52 ---------- Previous update was at 16:50 ----------

Might be useful to pipe it to od -c too, as it would indicate explicitly the chars you're working with...

---------- Post updated at 16:52 ---------- Previous update was at 16:52 ----------

Might be useful to pipe it to od -c too, as it would indicate explicitly the chars you're working with...