Sed expression that converts some matches to uppercase

Question

This sed expression converts an input string into a two-line output string. Each of the two output lines are composed of substrings from the input. The first line needs to be convered into upper case:

s:random_stuff\(choice1\|choice2\){\([^}]*\)}:\U\1\n\2:

The aim is to convert

random_stuff_choice1{This is a sentence that MAY contain AnyThing}
random_stuff_choice2{This is another similar sentence}

into

CHOICE1
This is a sentence that MAY contain AnyThing
CHOICE2
This is another similar sentence

The problem I have is that \U aplies to everything following it so the second line is also uppercased. Is it possible to make \U apply to the first match only ?

Answer 1

With sed:

$ sed 's/.*\(choice[0-9]\+\){\([^}]*\)}/\U\1\n\E\2/' file
CHOICE1
This is a sentence that MAY contain AnyThing
CHOICE2
This is another similar sentence

With awk:

$ awk -F'{|}' 'gsub(/.*_/,""){print toupper($1)"\n"$2}' file
CHOICE1
This is a sentence that MAY contain AnyThing
CHOICE2
This is another similar sentence