UWN
ISO/IEC 13211-2 was published 2000. Even in 2006, many of the aspects of relevance where not yet understood. Think of compatibility with call/N. And the only system that implemented 6.4.4.2...
You are "importing" the module `modop` via `'.'/2`, that is `[File]`. Nowhere in 13211-2 is this operation specified. And then, you are using explicit qualification, which when used directly is...
It really is a good idea to just stick to SICStus. In addition, local operators and the ability to export operators from a module. So like in Scryer. But no...
This usage is not the most common one. So I would not be surprised if there are differences maybe also within the same system. In fact, use_module should be seen...
SICStus has the notion of a type-in module which can be changed with `set_module/1` (since 4.0.0, before it was `module/1`) or the corresponding flag `typein_module` (not sure when this flag...
Works in SICStus. Needed to replace `//1` by `/3` to get rid of that warning. But then, SICStus even detected a real error in your code: | ?- use_module(meta_dcg_test). %...
``` | ?- case1(Ls). calling_non_terminal(user:test). Ls = [a,b,c] ? ; no | ?- case2(Ls). calling_non_terminal(user:test_meta(test)). calling_non_terminal(user:test). Ls = [a,b,c] ? ; no | ?- case3(Ls). calling_non_terminal(user:(test_meta(test),test)). calling_non_terminal(user:test). Ls = [a,b,c,a,b,c]...
There is one microscopic misunderstanding here: `calling_non_terminal` is a bit of a misnomer. For `phrase/2/3`, the first argument is not a non-terminal but rather a grammar rule body. After all,...
And here it is perfect: ``` ?- length(L,N), write_term(L,[max_depth(3)]). [] L = [], N = 0 ; [_13] L = [_A], N = 1 ; [_13,_18] L = [_A,_B], N...
``` Trealla Prolog (c) Infradig 2020-2023, v2.21.31 ?- write_term(t,[max_depth(-1)]). error(domain_error(write_option,max_depth(-1)),write_term/2). t?- ^^ unexpected t