UNIX for Dummies Questions & Answers

I have a data with 4 columns separete with tab delimiter as following:

Code:

A  B  {123}  E{456}
C  D  {789}  T{159}

My expected result as following: (become 5 columns)

Code:

A B {123} E  {456}
C D {789} T  {159}

Is there any way to separate the 4th column to 2 field by using the tab delimiter?
I am using the tr command, but it will add another tab delimiter to the 3rd column as well which is I don't wish to, I just want to touch the last column. See my command.

Code:

tr '{' '\t' < $InFile >> $OutFile

Thanks.

Code:

sed 's/\([^     ]\){/\1 {/' $InFile >> $OutFile

In case it isn't obvious, there are two characters between the [ and the ]: <circumflex> and <tab>.

Code:

sed 's/[\t]*{/\t{/g' infile

No. Inside a bracket expression, backslash is a regular character. This script looks for a string of zero or more backslash and t characters before a { and replaces each one it finds with t{. For the given input file, this produces:

Code:

A	B	t{123}	Et{456}
C	D	t{789}	Tt{159}

instead of:

Code:

A	B	{123}	E	{456}
C	D	{789}	T	{159}

Wrongo. Script works fine.

It works fine with GNU sed. It won't work with the sed implementations used by just about any non-Linux platform.

Regards,
Alister

Not on OS X. And, not if sed handles bracket expressions as defined by the standards.