1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
|
//: Extend 'new' to handle a unicode string literal argument or 'text'.
//: A Mu text is an address to an array of characters.
:(before "End Mu Types Initialization")
put(Type_abbreviations, "text", new_type_tree("&:@:character"));
:(scenario new_string)
def main [
10:text <- new [abc def]
20:char <- index *10:text, 5
]
# number code for 'e'
+mem: storing 101 in location 20
:(scenario new_string_handles_unicode)
def main [
10:text <- new [a«c]
20:num <- length *10:text
21:char <- index *10:text, 1
]
+mem: storing 3 in location 20
# unicode for '«'
+mem: storing 171 in location 21
:(before "End NEW Check Special-cases")
if (is_literal_text(inst.ingredients.at(0))) break;
:(before "Convert 'new' To 'allocate'")
if (inst.name == "new" && !inst.ingredients.empty() && is_literal_text(inst.ingredients.at(0))) continue;
:(after "case NEW" following "Primitive Recipe Implementations")
if (is_literal_text(current_instruction().ingredients.at(0))) {
products.resize(1);
products.at(0).push_back(/*alloc id*/0);
products.at(0).push_back(new_mu_text(current_instruction().ingredients.at(0).name));
trace(Callstack_depth+1, "mem") << "new string alloc: " << products.at(0).at(0) << end();
break;
}
:(code)
int new_mu_text(const string& contents) {
// allocate an array just large enough for it
int string_length = unicode_length(contents);
//? Total_alloc += string_length+1;
//? ++Num_alloc;
int result = allocate(/*array length*/1 + string_length);
int curr_address = result;
++curr_address; // skip alloc id
trace(Callstack_depth+1, "mem") << "storing string length " << string_length << " in location " << curr_address << end();
put(Memory, curr_address, string_length);
++curr_address; // skip length
int curr = 0;
const char* raw_contents = contents.c_str();
for (int i = 0; i < string_length; ++i) {
uint32_t curr_character;
assert(curr < SIZE(contents));
tb_utf8_char_to_unicode(&curr_character, &raw_contents[curr]);
trace(Callstack_depth+1, "mem") << "storing string character " << curr_character << " in location " << curr_address << end();
put(Memory, curr_address, curr_character);
curr += tb_utf8_char_length(raw_contents[curr]);
++curr_address;
}
// Mu strings are not null-terminated in memory.
return result;
}
//: a new kind of typo
:(scenario literal_text_without_instruction)
% Hide_errors = true;
def main [
[abc]
]
+error: main: instruction '[abc]' has no recipe in '[abc]'
//: stash recognizes texts
:(scenario stash_text)
def main [
1:text <- new [abc]
stash [foo:], 1:text
]
+app: foo: abc
:(before "End inspect Special-cases(r, data)")
if (is_mu_text(r)) {
return read_mu_text(data.at(/*skip alloc id*/1));
}
:(before "End $print Special-cases")
else if (is_mu_text(current_instruction().ingredients.at(i))) {
cout << read_mu_text(ingredients.at(i).at(/*skip alloc id*/1));
}
:(scenario unicode_text)
def main [
1:text <- new [♠]
stash [foo:], 1:text
]
+app: foo: ♠
:(scenario stash_space_after_text)
def main [
1:text <- new [abc]
stash 1:text, [foo]
]
+app: abc foo
:(scenario stash_text_as_array)
def main [
1:text <- new [abc]
stash *1:text
]
+app: 3 97 98 99
//: fixes way more than just stash
:(before "End Preprocess is_mu_text(reagent x)")
if (!canonize_type(x)) return false;
//: Allocate more to routine when initializing a literal text
:(scenario new_text_overflow)
% Initial_memory_per_routine = 3;
def main [
10:&:num/raw <- new number:type
20:text/raw <- new [a] # not enough room in initial page, if you take the array length into account
]
+new: routine allocated memory from 1000 to 1003
+new: routine allocated memory from 1003 to 1006
//: helpers
:(code)
int unicode_length(const string& s) {
const char* in = s.c_str();
int result = 0;
int curr = 0;
while (curr < SIZE(s)) { // carefully bounds-check on the string
// before accessing its raw pointer
++result;
curr += tb_utf8_char_length(in[curr]);
}
return result;
}
string read_mu_text(int address) {
if (address == 0) return "";
int length = get_or_insert(Memory, address+/*alloc id*/1);
if (length == 0) return "";
return read_mu_characters(address+/*alloc id*/1+/*length*/1, length);
}
string read_mu_characters(int start, int length) {
ostringstream tmp;
for (int curr = start; curr < start+length; ++curr)
tmp << to_unicode(static_cast<uint32_t>(get_or_insert(Memory, curr)));
return tmp.str();
}
//:: some miscellaneous helpers now that we have text
//: assert: perform sanity checks at runtime
:(scenario assert_literal)
% Hide_errors = true; // '%' lines insert arbitrary C code into tests before calling 'run' with the lines below. Must be immediately after :(scenario) line.
def main [
assert 0, [this is an assert in Mu]
]
+error: this is an assert in Mu
:(scenario assert)
% Hide_errors = true; // '%' lines insert arbitrary C code into tests before calling 'run' with the lines below. Must be immediately after :(scenario) line.
def main [
1:text <- new [this is an assert in Mu]
assert 0, 1:text
]
+error: this is an assert in Mu
:(before "End Primitive Recipe Declarations")
ASSERT,
:(before "End Primitive Recipe Numbers")
put(Recipe_ordinal, "assert", ASSERT);
:(before "End Primitive Recipe Checks")
case ASSERT: {
if (SIZE(inst.ingredients) != 2) {
raise << maybe(get(Recipe, r).name) << "'assert' takes exactly two ingredients rather than '" << to_original_string(inst) << "'\n" << end();
break;
}
if (!is_mu_address(inst.ingredients.at(0)) && !is_mu_scalar(inst.ingredients.at(0))) {
raise << maybe(get(Recipe, r).name) << "'assert' requires a scalar or address for its first ingredient, but got '" << inst.ingredients.at(0).original_string << "'\n" << end();
break;
}
if (!is_literal_text(inst.ingredients.at(1)) && !is_mu_text(inst.ingredients.at(1))) {
raise << maybe(get(Recipe, r).name) << "'assert' requires a text as its second ingredient, but got '" << inst.ingredients.at(1).original_string << "'\n" << end();
break;
}
break;
}
:(before "End Primitive Recipe Implementations")
case ASSERT: {
if (!scalar_ingredient(ingredients, 0)) {
if (is_literal_text(current_instruction().ingredients.at(1)))
raise << current_instruction().ingredients.at(1).name << '\n' << end();
else
raise << read_mu_text(ingredients.at(1).at(/*skip alloc id*/1)) << '\n' << end();
if (!Hide_errors) exit(1);
}
break;
}
//: 'cheating' by using the host system
:(before "End Primitive Recipe Declarations")
_READ,
:(before "End Primitive Recipe Numbers")
put(Recipe_ordinal, "$read", _READ);
:(before "End Primitive Recipe Checks")
case _READ: {
break;
}
:(before "End Primitive Recipe Implementations")
case _READ: {
skip_whitespace(cin);
string result;
if (has_data(cin))
cin >> result;
products.resize(1);
products.at(0).push_back(new_mu_text(result));
break;
}
:(code)
void skip_whitespace(istream& in) {
while (true) {
if (!has_data(in)) break;
if (isspace(in.peek())) in.get();
else break;
}
}
|