about summary refs log tree commit diff stats
path: root/cpp/032array
diff options
context:
space:
mode:
Diffstat (limited to 'cpp/032array')
-rw-r--r--cpp/032array152
1 files changed, 152 insertions, 0 deletions
diff --git a/cpp/032array b/cpp/032array
new file mode 100644
index 00000000..3bdfe95e
--- /dev/null
+++ b/cpp/032array
@@ -0,0 +1,152 @@
+//: Arrays contain a variable number of elements of the same type.
+:(scenario copy_array)
+# Arrays can be copied around with a single instruction just like integers,
+# no matter how large they are.
+recipe main [
+  1:integer <- copy 3:literal
+  2:integer <- copy 14:literal
+  3:integer <- copy 15:literal
+  4:integer <- copy 16:literal
+  5:array:integer <- copy 1:array:integer
+]
++run: instruction main/4
++run: ingredient 0 is 1
++mem: location 1 is 3
++mem: location 2 is 14
++mem: location 3 is 15
++mem: location 4 is 16
++mem: storing 3 in location 5
++mem: storing 14 in location 6
++mem: storing 15 in location 7
++mem: storing 16 in location 8
+
+//: disable the size mismatch check since the destination array need not be initialized
+:(replace "if (size_of(x) != data.size())" following "void write_memory(reagent x, vector<int> data)")
+if (x.types[0] != Type_number["array"] && size_of(x) != data.size())
+:(after "size_t size_of(const reagent& r)")
+  static const int ARRAY = Type_number["array"];
+  if (r.types[0] == ARRAY) {
+    assert(r.types.size() > 1);
+    // skip the 'array' type to get at the element type
+    return 1 + Memory[r.value]*size_of(array_element(r.types));
+  }
+
+//: To access elements of an array, use 'index'
+:(scenario "index")
+recipe main [
+  1:integer <- copy 3:literal
+  2:integer <- copy 14:literal
+  3:integer <- copy 15:literal
+  4:integer <- copy 16:literal
+  5:integer <- index 1:array:integer, 0:literal
+]
++run: instruction main/4
++run: address to copy is 2
++run: its type is 1
++mem: location 2 is 14
++run: product 0 is 14
++mem: storing 14 in location 5
+
+:(scenario "index_direct_offset")
+recipe main [
+  1:integer <- copy 3:literal
+  2:integer <- copy 14:literal
+  3:integer <- copy 15:literal
+  4:integer <- copy 16:literal
+  5:integer <- copy 0:literal
+  6:integer <- index 1:array:integer, 5:integer
+]
++run: instruction main/5
++run: address to copy is 2
++run: its type is 1
++mem: location 2 is 14
++run: product 0 is 14
++mem: storing 14 in location 6
+
+:(before "End Primitive Recipe Declarations")
+INDEX,
+:(before "End Primitive Recipe Numbers")
+Recipe_number["index"] = INDEX;
+:(before "End Primitive Recipe Implementations")
+case INDEX: {
+  static const int ARRAY = Type_number["array"];
+//?   if (Trace_stream) Trace_stream->dump_layer = "run"; //? 1
+  trace("run") << "ingredient 0 is " << instructions[pc].ingredients[0].to_string();
+  reagent base = canonize(instructions[pc].ingredients[0]);
+//?   trace("run") << "ingredient 0 after canonize: " << instructions[pc].ingredients[0].to_string(); //? 1
+  int base_address = base.value;
+  assert(base.types[0] == ARRAY);
+  trace("run") << "ingredient 1 is " << instructions[pc].ingredients[1].to_string();
+  reagent offset = canonize(instructions[pc].ingredients[1]);
+  vector<int> offset_val(read_memory(offset));
+  vector<type_number> element_type = array_element(base.types);
+  int src = base_address + 1 + offset_val[0]*size_of(element_type);
+  trace("run") << "address to copy is " << src;
+  trace("run") << "its type is " << element_type[0];
+  reagent tmp;
+  tmp.set_value(src);
+  copy(element_type.begin(), element_type.end(), inserter(tmp.types, tmp.types.begin()));
+  tmp.properties.push_back(pair<string, vector<string> >("raw", vector<string>()));
+//?   cout << "AAA: " << tmp.to_string() << '\n'; //? 2
+  vector<int> result(read_memory(tmp));
+  trace("run") << "product 0 is " << result[0];
+  write_memory(instructions[pc].products[0], result);
+//?   if (Trace_stream) Trace_stream->dump_layer = ""; //? 1
+  break;
+}
+
+:(code)
+vector<type_number> array_element(const vector<type_number>& types) {
+  return vector<type_number>(++types.begin(), types.end());
+}
+
+:(scenario "index_address")
+recipe main [
+  1:integer <- copy 3:literal
+  2:integer <- copy 14:literal
+  3:integer <- copy 15:literal
+  4:integer <- copy 16:literal
+  5:integer <- index-address 1:array:integer, 0:literal
+]
++run: instruction main/4
++run: address to copy is 2
++mem: storing 2 in location 5
+
+//: To write to elements of containers, you need their address.
+
+:(scenario "index_indirect")
+recipe main [
+  1:integer <- copy 3:literal
+  2:integer <- copy 14:literal
+  3:integer <- copy 15:literal
+  4:integer <- copy 16:literal
+  5:address:array:integer <- copy 1:literal
+  6:integer <- index 5:address:array:integer/deref, 1:literal
+]
++run: instruction main/5
++mem: storing 15 in location 6
+// vim:ft=cpp
+
+:(before "End Primitive Recipe Declarations")
+INDEX_ADDRESS,
+:(before "End Primitive Recipe Numbers")
+Recipe_number["index-address"] = INDEX_ADDRESS;
+:(before "End Primitive Recipe Implementations")
+case INDEX_ADDRESS: {
+  static const int ARRAY = Type_number["array"];
+  trace("run") << "ingredient 0 is " << instructions[pc].ingredients[0].name;
+  reagent base = canonize(instructions[pc].ingredients[0]);
+  int base_address = base.value;
+  assert(base.types[0] == ARRAY);
+  trace("run") << "ingredient 1 is " << instructions[pc].ingredients[1].to_string();
+  reagent offset = canonize(instructions[pc].ingredients[1]);
+  vector<int> offset_val(read_memory(offset));
+  vector<type_number> element_type = array_element(base.types);
+  int src = base_address + 1 + offset_val[0]*size_of(element_type);
+  trace("run") << "address to copy is " << src;
+  vector<int> result;
+  result.push_back(src);
+  trace("run") << "product 0 is " << result[0];
+  write_memory(instructions[pc].products[0], result);
+  break;
+}