UNIX for Advanced & Expert Users

Hello all
I have a file with this type of information:

Code:

data1;data2;data3
data4:data5;data6
data7:data8,data9

As you see, fileds are separated by diferents caracters each time.
I would like to have always the same caracter the first time for exemple :
as this:

Code:

data1:data2;data3
data4:data5;data6
data7:data8,data9

How can I do this, with awk maybe?
Carefull that if there are 2 ; in the same line, I want to change only the first occurence.
Thanks a lot

Try...

Code:

sed 's/\;/\:/' infile

Ok it works.
Can you tell me why the second ; have not been changed?
in this case for exemple
"PRINCETON:PRINCETON UNIVERSITY, PLASMA PHYSICS LABORATORY;1989"

It's exactly what I want, but I do not anderstand why it works !! jaajaj
Cheers

---------- Post updated at 06:34 AM ---------- Previous update was at 06:24 AM ----------

Hello again
all the file have not been changed well, look this line for exemple:

KALSRUHE;KERNFORSCHUNGSZENTRUM KARLSRUHE GMBH:1990

May I pass the sed command twice?

Hi,

As far as i know you should use "g" option (global) to apply the substitution process each lines of your input file:

Code:

sed 's/\;/\:/g' infile

Besides if you need to change more than one field seperator you can do it with "-e" option ( connecting a few SED commands with -e).

Maybe with a real exemple it will be easy to anderstand, sorry...
So part of the file is this:

Code:

MADRID : JUNTA DE ENERGIA NUCLEAR, 1957
MADRID : JUNTA DE ENERGIA NUCLEAR, 1958
MADRID : JUNTA DE ENERGIA NUCLEAR, 1958
MADRI : :CIEMAT, 1988
MADRID : CIEMAT, 1988
MADRID : JUNTA DE ENERGIA NUCLEAR, 1985

As you see, the 3 first lines are good, it's easy to split the sring with excel to obtain 3 colums (that is the final objective)
But the nex one is problematic because we have : :
The idea will be to convert all the FIRST separate caracter, in this case : with an other, never used in the file, such as @ or ##
This way I'll be abble to use it in excel to separate values.

---------- Post updated at 09:38 AM ---------- Previous update was at 09:36 AM ----------

Me again
If I aplly the sed you told me earlier, The result is:

Code:

MADRID ## JUNTA DE ENERGIA NUCLEAR, 1958
MADRID ## JUNTA DE ENERGIA NUCLEAR, 1958
MADRI ## ##CIEMAT, 1988
MADRID ## CIEMAT, 1988
MADRID ## JUNTA DE ENERGIA NUCLEAR, 1985

and again, this line will be impossible to parse by excel: MADRI ## ##CIEMAT, 1988

the correct should be: MADRI ## :CIEMAT, 1988

Code:

awk '{sub(":", "##"); print}' file

Yes it works...
Very good
now last dubt, is it possible to identify a 4 digit field?? (the date in this case)
I ask because in this file, all the dates are separate with , .... but a few (andreds) are separated with : or ;
So the idea will be to identify the date and replace only the separate caracter just before.
Is it possible?
Thanks (this is the last question...)