Shell Programming and Scripting

The below sed command does that. Did you try this?

Code:

sed -e 's/               /       /' -e 's/       /,/g' tab_del

Code:

you can try this if the tabs are going to 1 or more than 1 between fields...


sed 's/        [         ]*/,/g' tab_del

after s/.... there is <tab> ( not spaces ) and within [] there is one more tab ( not spaces )....

this will replace any number of tabs to single comma...

Guys...

Could u people let me know on how to find whether
Spaces between the fields in a record is a "Tab" or "Multiple Spaces".

Code:

/export/home/test/mons>cat -vet hh
adasd^Isdfsdf^I^Isdfsdfdsf$
dsfdsf sdf sdfsd sfsdfsdf$

cat -vet will display all the control characters

^I represents tab

and space is shown as " " only.

when i issue cat dmlog.txt i get

ANALYZE DLS_DM COMPLETE GE 24-OCT-05
LOAD_STAGE STAGE_LOAD COMPLETE C0 16-DEC-05



when i issue cat -vet dmlog.txt i get

ANALYZE DLS_DM COMPLETE GE 24-OCT-05 $
LOAD_STAGE STAGE_LOAD COMPLETE C0 16-DEC-05 $


Is that mean only spaces are there b/w fields ?? I dont see any ^I characters..

Here is what i get...

when i issue cat dmlog.txt i get

ANALYZE DLS_DM COMPLETE GE 24-OCT-05
LOAD_STAGE STAGE_LOAD COMPLETE C0 16-DEC-05


when i issue cat -vet dmlog.txt i get

ANALYZE DLS_DM COMPLETE GE 24-OCT-05 $
LOAD_STAGE STAGE_LOAD COMPLETE C0 16-DEC-05 $



Is that mean only spaces are there b/w fields ?? I dont see any ^I characters..

If so,

How to convert those spaces into comma ??

then this command will do

Code:

sed 's/ /,/g' filename

or

tr " " "," < filename

I can see some extra spaces at the last of each line. If you don't want a comma at the last then you can strip the extra space and convert. The below will do that

Code:

sed -e 's/^\(.*\) $/\1/' -e 's/ /,/g' filename