regex - How could I delete sql functions format in awk? -

i've got sql query looks this:

select test1(func1(myfield)),         test2(max(myfield), lower("nope")),        test3(max(myfield), 1234),        avg(test1(test2(myfield, func1(4)))),        func2(upper("stack"))        substr(myfield, 2, 4),        test2(min(myfield), substr(lower(upper("nope")), 1, 7)),        substr('func1(', 2, 4) mytable; 

then i'm trying remove functions called:

• test1
• test2
• test3
• func1
• func2

but preserving avg, max, upper, substr... , native functions.

so desired output be:

select myfield,         max(myfield),        max(myfield),        avg(myfield),        upper("stack")        substr(myfield, 2, 4),        min(myfield)        substr('func1(', 2, 4) mytable; 

i want remove lower of second line because, argument of 1 of functions delete, in case test2, has 2 parameters. if delete function, should delete params well.

i've tried way in awk:

{     print gensub(/(test1|test2|func1|func2)\(/,"","gi", $0); } 

but output doesn't have account right parentheses, doesn't delete rest of parameters of custom functions:

select myfield)),        max(myfield), lower("nope")),        max(myfield), 1234),        avg(myfield, 4)))),        upper("stack"))        substr(myfield, 2, 4),        min(myfield), substr(lower(upper("nope")), 1, 7)),        substr('', 2, 4) mytable; 

any idea or clue handle situation?

you rename functions' names built-in functioncoalesce while keep brakets ( ) , other params of users' functions. produce same result, not syntactically, work same unless built-in functions don't return null values. easier achieve because don't have worry brakets.

if file input provide, then:

cat file | sed 's#\(test1\|test2\|func1\|func2\)(#coalesce(#g'

will produce:

select coalesce(coalesce(myfield)),     coalesce(max(myfield), 4),     avg(coalesce(coalesce(myfield, coalesce(4)))),     coalesce(upper("stack")) mytable;