//: Introduce a new transform to perform various checks in instructions before
//: we start running them. It'll be extensible, so that we can add checks for
//: new recipes as we extend 'run' to support them.
//:
//: Doing checking in a separate part complicates things, because the values
//: of variables in memory and the processor (current_recipe_name,
//: current_instruction) aren't available at checking time. If I had a more
//: sophisticated layer system I'd introduce the simpler version first and
//: transform it in a separate layer or set of layers.
:(before "End Checks")
Transform.push_back(check_instruction); // idempotent
:(code)
void check_instruction(const recipe_ordinal r) {
trace(101, "transform") << "--- perform checks for recipe " << get(Recipe, r).name << end();
map<string, vector<type_ordinal> > metadata;
for (int i = 0; i < SIZE(get(Recipe, r).steps); ++i) {
instruction& inst = get(Recipe, r).steps.at(i);
if (inst.is_label) continue;
switch (inst.operation) {
// Primitive Recipe Checks
case COPY: {
if (SIZE(inst.products) > SIZE(inst.ingredients)) {
raise << maybe(get(Recipe, r).name) << "too many products in '" << to_original_string(inst) << "'\n" << end();
break;
}
for (int i = 0; i < SIZE(inst.products); ++i) {
if (!types_coercible(inst.products.at(i), inst.ingredients.at(i))) {
raise << maybe(get(Recipe, r).name) << "can't copy '" << inst.ingredients.at(i).original_string << "' to '" << inst.products.at(i).original_string << "'; types don't match\n" << end();
goto finish_checking_instruction;
}
}
break;
}
// End Primitive Recipe Checks
default: {
// Defined Recipe Checks
// End Defined Recipe Checks
}
}
finish_checking_instruction:;
}
}
void test_copy_checks_reagent_count() {
Hide_errors = true;
run(
"def main [\n"
" 1:num, 2:num <- copy 34\n"
"]\n"
);
CHECK_TRACE_CONTENTS(
"error: main: too many products in '1:num, 2:num <- copy 34'\n"
);
}
void test_write_scalar_to_array_disallowed() {
Hide_errors = true;
run(
"def main [\n"
" 1:array:num <- copy 34\n"
"]\n"
);
CHECK_TRACE_CONTENTS(
"error: main: can't copy '34' to '1:array:num'; types don't match\n"
);
}
void test_write_scalar_to_array_disallowed_2() {
Hide_errors = true;
run(
"def main [\n"
" 1:num, 2:array:num <- copy 34, 35\n"
"]\n"
);
CHECK_TRACE_CONTENTS(
"error: main: can't copy '35' to '2:array:num'; types don't match\n"
);
}
void test_write_scalar_to_address_disallowed() {
Hide_errors = true;
run(
"def main [\n"
" 1:&:num <- copy 34\n"
"]\n"
);
CHECK_TRACE_CONTENTS(
"error: main: can't copy '34' to '1:&:num'; types don't match\n"
);
}
void test_write_address_to_character_disallowed() {
Hide_errors = true;
run(
"def main [\n"
" 1:&:num <- copy 12/unsafe\n"
" 2:char <- copy 1:&:num\n"
"]\n"
);
CHECK_TRACE_CONTENTS(
"error: main: can't copy '1:&:num' to '2:char'; types don't match\n"
);
}
void test_write_number_to_character_allowed() {
run(
"def main [\n"
" 1:num <- copy 97\n"
" 2:char <- copy 1:num\n"
"]\n"
);
CHECK_TRACE_COUNT("error", 0);
}
:(code)
// types_match with some leniency
bool types_coercible(reagent/*copy*/ to, reagent/*copy*/ from) {
// Begin types_coercible(reagent to, reagent from)
if (types_match_sub(to, from)) return true;
if (is_real_mu_number(from) && is_mu_character(to)) return true;
// End types_coercible Special-cases
return false;
}
bool types_match_sub(const reagent& to, const reagent& from) {
// to sidestep type-checking, use /unsafe in the source.
// this will be highlighted in red inside vim. just for setting up some tests.
if (is_unsafe(from)) return true;
if (is_literal(from)) {
if (is_mu_array(to)) return false;
// End Matching Types For Literal(to)
if (!to.type) return false;
// allow writing null to any address
if (is_mu_address(to)) return from.name == "null";
return size_of(to) == 1; // literals are always scalars
}
return types_strictly_match_sub(to, from);
}
// variant for others to call
bool types_match(reagent/*copy*/ to, reagent/*copy*/ from) {
// Begin types_match(reagent to, reagent from)
return types_match_sub(to, from);
}
//: copy arguments for later layers
bool types_strictly_match_sub(const reagent& to, const reagent& from) {
if (to.type == NULL) return false; // error
if (is_literal(from) && to.type->value == Number_type_ordinal) return true;
// to sidestep type-checking, use /unsafe in the source.
// this will be highlighted in red inside vim. just for setting up some tests.
if (is_unsafe(from)) return true;
// '_' never raises type error
if (is_dummy(to)) return true;
if (!to.type) return !from.type;
return types_strictly_match(to.type, from.type);
}
// variant for others to call
bool types_strictly_match(reagent/*copy*/ to, reagent/*copy*/ from) {
// Begin types_strictly_match(reagent to, reagent from)
return types_strictly_match_sub(to, from);
}
bool types_strictly_match(const type_tree* to, const type_tree* from) {
if (from == to) return true;
if (!to) return false;
if (!from) return to->atom && to->value == 0;
if (from->atom != to->atom) return false;
if (from->atom) {
if (from->value == -1) return from->name == to->name;
return from->value == to->value;
}
if (types_strictly_match(to->left, from->left) && types_strictly_match(to->right, from->right))
return true;
// fallback: (x) == x
if (to->right == NULL && types_strictly_match(to->left, from)) return true;
if (from->right == NULL && types_strictly_match(to, from->left)) return true;
return false;
}
void test_unknown_type_does_not_match_unknown_type() {
reagent a("a:foo");
reagent b("b:bar");
CHECK(!types_strictly_match(a, b));
}
void test_unknown_type_matches_itself() {
reagent a("a:foo");
reagent b("b:foo");
CHECK(types_strictly_match(a, b));
}
void test_type_abbreviations_match_raw_types() {
put(Type_abbreviations, "text", new_type_tree("address:array:character"));
// a has type (address buffer (address array character))
reagent a("a:address:buffer:text");
expand_type_abbreviations(a.type);
// b has type (address buffer address array character)
reagent b("b:address:buffer:address:array:character");
CHECK(types_strictly_match(a, b));
delete Type_abbreviations["text"];
put(Type_abbreviations, "text", NULL);
}
//: helpers
bool is_unsafe(const reagent& r) {
return has_property(r, "unsafe");
}
bool is_mu_array(reagent/*copy*/ r) {
// End Preprocess is_mu_array(reagent r)
return is_mu_array(r.type);
}
bool is_mu_array(const type_tree* type) {
if (!type) return false;
if (is_literal(type)) return false;
if (type->atom) return false;
if (!type->left->atom) {
raise << "invalid type " << to_string(type) << '\n' << end();
return false;
}
return type->left->value == Array_type_ordinal;
}
bool is_mu_boolean(reagent/*copy*/ r) {
// End Preprocess is_mu_boolean(reagent r)
if (!r.type) return false;
if (is_literal(r)) return false;
if (!r.type->atom) return false;
return r.type->value == Boolean_type_ordinal;
}
bool is_mu_number(reagent/*copy*/ r) {
if (is_mu_character(r.type)) return true; // permit arithmetic on unicode code points
return is_real_mu_number(r);
}
bool is_real_mu_number(reagent/*copy*/ r) {
// End Preprocess is_mu_number(reagent r)
if (!r.type) return false;
if (!r.type->atom) return false;
if (is_literal(r)) {
return r.type->name == "literal-fractional-number"
|| r.type->name == "literal";
}
return r.type->value == Number_type_ordinal;
}
bool is_mu_character(reagent/*copy*/ r) {
// End Preprocess is_mu_character(reagent r)
return is_mu_character(r.type);
}
bool is_mu_character(const type_tree* type) {
if (!type) return false;
if (!type->atom) return false;
if (is_literal(type)) return false;
return type->value == Character_type_ordinal;
}