YJL

func1() { :;}
func2() { :;}

1   declare or typeset

declare and typeset are synonyms to each other in Bash. If you use with declare -f, you will get

func1 ()
{
    :
}
func2 ()
{
    :
}

From bash(1):

The -f option will restrict the display to shell functions.

Initially, without argument, declare lists all variables and functions with their values and definition. -f instructs to only function, but the definitions are still listed, which are the parts we don’t need and don’t want to have. There is another option -F:

The -F option inhibits the display of function definitions; only the function name and attributes are printed.

With declare -F:

declare -f func1
declare -f func2

Therefore you would do something like the following to extract the function name part:

declare -F | cut -d ' ' -f 3

And you will get the only function names:

func1
func2

2   set

From bash(1):

Without options, the name and value of each shell variable are displayed in a format that can be reused as input for setting or resetting the currently-set variables.

I don’t like this method—though it’s similar to declare without arguments—because the output is mixed with variables. But if you want you can use:

set | sed -n '/[^ ]\+\ ()/ s/ ()//p'

3   compgen

From bash(1):

-A action

function Names of shell functions.

It really looks strange to use completion related functionality to list function names, but it does the job the most elegant way:

compgen -A function

No extraction is needed.

4   Usage of the function names

Whatever you choose to get the function names, the next step is to use it, that is calling them. For my case, I would do like:

funcs=($(compgen -A function))
for f in funcs; do
  # do some filter if the list is mixed with some other functions, you may
  # want to pseudo namespace'd functions you want to invoke.
  $f "$arg1" "$arg2" "..."
done

The list may contain functions you don’t need, so you probably need to check the names before calling them.