[m-users.] higher order issue: `ground` seen but expecting higher-order pred

[Sean Charles (emacstheviking)]

I am fleshing out a VERY loose forth type system, I have the following definitions to allow me to set up a map of word names and there handlers:

:- type lstr == list(string).

:- type word_handler == (pred(fstate, fstate, io, io, lstr, lstr)).
:- inst word_handler == (pred(in, out, di, uo, in, out)is det).

:- type word_map == map(string, word_handler).
:- type stack_entry == stack(int).

    %
    % FBNF operating state
    %
:- type fstate
    --->    fstate(
                fs_dict  :: word_map,
                fs_stack :: stack_entry,
                fs_error :: maybe(string)
            ).

then I initialise the default word list like this:

initial_dictionary = Map :-
    some [!Words] (
        !:Words = map.init,
        map.set(":",     define_word, !Words),
        map.set("words", list_words,  !Words),
        Map = !.Words
   ).

Here are the implementations of the above two default words:

:- pred define_word(fstate::in, fstate::out, io::di, io::uo,
    lstr::in, lstr::out) is det.

define_word(!State, !IO) -->
    {
        io.format("DEFINE WORD\n", [], !IO)
    }.


:- pred list_words(fstate::in, fstate::out, io::di, io::uo,
    lstr::in, lstr::out) is det.

list_words(!State, !IO) -->
    {
        io.format("WORDS\n", [], !IO)
    }.


and eventually when I try to call it, I get the error on the line shown , here I look up the word in the map into Entry, and try to call it:

:- pred execute(string::in, fstate::in, fstate::out, io::di, io::uo,
    lstr::in, lstr::out) is det.

execute(Word, !State, !IO) -->
    { Words = fs_dict(!.State) },
    ( if { map.search(Words, Word, Entry) } then

(191)        Entry(!State, !IO)        <====== compiler hate this line!

    else
        { set_error(
            string.format("word not found: %s\n",
            [s(Word)]), !State) }
    ).


fbnf.m:191: In clause for `execute(in, in, out, di, uo, in, out)':
fbnf.m:191:   in argument 1 (i.e. the predicate term) of higher-order predicate
fbnf.m:191:   call:
fbnf.m:191:   mode error: variable `Entry' has instantiatedness `ground',
fbnf.m:191:   expecting higher-order pred inst of arity 6.



I am not sure the DCG style will remain but it's pure proof of concept / exploratory at the moment. I used getopt.m as inspiration but obviously talent and brains are needed as well... what have I not told the compiler that it needs to know ?

Thanks
Sean

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurylang.org/archives/users/attachments/20221004/0a6d146f/attachment.html>