Shell Programming and Scripting

Hi,

I just wandering how to split a record which has more than one delimiter,
i have a file which contains pattern [999]as group separtor and ~ as field separtor, Ultimately I need consider even the groups as a field, So i need to make this multi-delimited file into ~ delimited file.

My record looks like

Actual record 1

Code:

[001]XXX[002]~D~199~PASSTHRSETT~xx xx~~[003]US[004]USD[005]A[006]A[007]09-DEC-99[008]09-DEC-99[040]5[050]5[060]1[061]0[062]0[063]0[064]0

Actual record 2

Code:

[001]YYY[002]~D~199~PP~xx xx~~[003]US[004]USD[005]A[006]A[007]09-OCT-99[008]09-DEC-99[040]5[050]5[060]1[061]0[062]0[063]0[064]0

Can someone tell me, how can I split the record so that output looks like
Expected record

Code:

XXX~D~199~PASSTHRSETT~xx xx~~US~USD~A~A~09-DEC-99~09-DEC-99~5~5~1~0~0~0~0

Thanks
Brian

sed 's/\[[0-9]\{,3\}\]/~/g' file > newfile.

Hi,

Thanks,
could you tell me what is the significance of {,3\}, is this number of repeatition.

Anyway I modified the SED as

sed 's/\[[0-9][0-9][0-9]\]/~/g' temp > n_temp and it works

That's exactly what it is.

Is this number repition works in a specific version ?
In my system it works like this:
sed 's/\[[0-9]\{3\}\]/~/g'

without the comma. My system is :

SunOS MyServer 5.8 Generic_117350-26 sun4u sparc SUNW,Ultra-60