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
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
use std::mem;
use crate::rubysys::{
constant::{FL_USER_1, FL_USER_17, FL_USER_7},
libc::size_t,
types::{
c_char, c_double, c_int, c_long, CallbackPtr, EncodingType, InternalValue, RBasic, Value,
},
};
pub const STR_TMPLOCK: isize = FL_USER_7;
#[cfg_attr(rutie_dllimport, link(name = "rutie_ruby"))]
extern "C" {
// VALUE
// rb_str_new(const char *ptr, long len)
pub fn rb_str_new(str: *const c_char, len: c_long) -> Value;
// VALUE
// rb_str_new_cstr(const char *ptr)
pub fn rb_str_new_cstr(str: *const c_char) -> Value;
// VALUE
// rb_utf8_str_new(const char *ptr, long len)
pub fn rb_utf8_str_new(str: *const c_char, len: c_long) -> Value;
// VALUE
// rb_utf8_str_new_cstr(const char *ptr)
pub fn rb_utf8_str_new_cstr(str: *const c_char) -> Value;
// char *
// rb_string_value_cstr(volatile VALUE *ptr)
pub fn rb_string_value_cstr(str: *const Value) -> *const c_char;
// char *
// rb_string_value_ptr(volatile VALUE *ptr)
pub fn rb_string_value_ptr(str: *const Value) -> *const c_char;
// long
// rb_str_strlen(VALUE str)
pub fn rb_str_strlen(str: Value) -> c_long;
// int
// rb_enc_str_asciionly_p(VALUE str)
pub fn rb_enc_str_asciionly_p(str: Value) -> c_int;
// VALUE
// rb_enc_str_new(const char *ptr, long len, rb_encoding *enc)
pub fn rb_enc_str_new(str: *const c_char, len: c_long, enc: EncodingType) -> Value;
// VALUE
// rb_str_export_locale(VALUE str)
pub fn rb_str_export_locale(str: Value) -> Value;
// VALUE
// rb_str_cat(VALUE str, const char *ptr, long len)
pub fn rb_str_cat(str: Value, ptr: *const c_char, len: c_long) -> Value;
// VALUE
// rb_check_string_type(VALUE str)
pub fn rb_check_string_type(str: Value) -> Value;
//-------------------------------------------------------------
// LINKER CANNOT FIND
// //
// // call-seq:
// // str.force_encoding(encoding) -> str
// //
// // Changes the encoding to +encoding+ and returns self.
// //
// // static VALUE
// // rb_str_force_encoding(VALUE str, VALUE enc)
// pub fn rb_str_force_encoding(s: Value, enc: Value) -> Value;
//-------------------------------------------------------------
// VALUE
// rb_str_locktmp(VALUE str)
pub fn rb_str_locktmp(str: Value) -> Value;
// VALUE
// rb_str_unlocktmp(VALUE str)
pub fn rb_str_unlocktmp(str: Value) -> Value;
// VALUE
// rb_str_new_frozen(VALUE orig)
pub fn rb_str_new_frozen(orig: Value) -> Value;
// VALUE
// rb_str_freeze(VALUE str)
pub fn rb_str_freeze(string: Value) -> Value;
// VALUE
// rb_str_buf_append(VALUE str, VALUE str2)
pub fn rb_str_buf_append(string: Value, other: Value) -> Value;
// VALUE
// rb_str_buf_new(long capa)
pub fn rb_str_buf_new(capacity: c_long) -> Value;
// size_t
// rb_str_capacity(VALUE str)
pub fn rb_str_capacity(string: Value) -> size_t;
// int
// rb_str_cmp(VALUE str1, VALUE str2)
pub fn rb_str_cmp(string: Value, other: Value) -> c_int;
// VALUE
// rb_str_conv_enc(VALUE str, rb_encoding *from, rb_encoding *to)
pub fn rb_str_conv_enc(string: Value, from: EncodingType, to: EncodingType) -> Value;
// VALUE
// rb_str_dup(VALUE str)
pub fn rb_str_dup(string: Value) -> Value;
// VALUE
// rb_str_ellipsize(VALUE str, long len)
pub fn rb_str_ellipsize(string: Value, len: c_long) -> Value;
// VALUE
// rb_str_equal(VALUE str1, VALUE str2)
pub fn rb_str_equal(string: Value, other: Value) -> Value;
// st_index_t
// rb_str_hash(VALUE str)
pub fn rb_str_hash(string: Value) -> size_t;
// VALUE
// rb_str_inspect(VALUE str)
pub fn rb_str_inspect(string: Value) -> Value;
// VALUE
// rb_str_intern(VALUE str)
pub fn rb_str_intern(string: Value) -> Value;
// VALUE
// rb_str_length(VALUE str)
pub fn rb_str_length(string: Value) -> Value;
// void
// rb_str_modify(VALUE str)
pub fn rb_str_modify(string: Value);
// VALUE
// rb_str_plus(VALUE str1, VALUE str2)
pub fn rb_str_plus(string: Value, other: Value) -> Value;
// VALUE
// rb_str_replace(VALUE str, VALUE str2)
pub fn rb_str_replace(string: Value, other: Value) -> Value;
// VALUE
// rb_str_resize(VALUE str, long len)
//
// Growing leaves the new bytes uninitialized.
pub fn rb_str_resize(string: Value, len: c_long) -> Value;
// VALUE
// rb_str_scrub(VALUE str, VALUE repl)
//
// Returns `Qnil` when `str` has no invalid byte sequences.
pub fn rb_str_scrub(string: Value, replacement: Value) -> Value;
// void
// rb_str_set_len(VALUE str, long len)
pub fn rb_str_set_len(string: Value, len: c_long);
// VALUE
// rb_str_split(VALUE str, const char *sep0)
pub fn rb_str_split(string: Value, separator: *const c_char) -> Value;
// VALUE
// rb_str_subseq(VALUE str, long beg, long len)
//
// Byte offsets, not bounds-checked.
pub fn rb_str_subseq(string: Value, begin: c_long, len: c_long) -> Value;
// VALUE
// rb_str_substr(VALUE str, long beg, long len)
//
// Character offsets; `Qnil` when out of range.
pub fn rb_str_substr(string: Value, begin: c_long, len: c_long) -> Value;
// VALUE
// rb_str_times(VALUE str, VALUE times)
pub fn rb_str_times(string: Value, times: Value) -> Value;
// double
// rb_str_to_dbl(VALUE str, int badcheck)
pub fn rb_str_to_dbl(string: Value, badcheck: c_int) -> c_double;
// VALUE
// rb_str_to_inum(VALUE str, int base, int badcheck)
pub fn rb_str_to_inum(string: Value, base: c_int, badcheck: c_int) -> Value;
// void
// rb_debug_rstring_null_ptr(const char *func)
//
// Warns on stderr that `func` (`"RSTRING_PTR"`) returns NULL, as a debug
// build of Ruby's `RSTRING_PTR` does for a string without a buffer.
pub fn rb_debug_rstring_null_ptr(func: *const c_char);
}
// #[link_name = "ruby_rstring_flags"]
#[derive(Debug, PartialEq)]
#[repr(C)]
enum RStringEmbed {
NoEmbed = FL_USER_1,
LenMax = (mem::size_of::<Value>() as isize * 3) / mem::size_of::<c_char>() as isize - 1,
Fstr = FL_USER_17,
}
#[derive(Copy, Clone)]
#[repr(C)]
union RStringAs {
heap: RStringHeap,
ary: [c_char; RStringEmbed::LenMax as usize + 1],
}
#[derive(Copy, Clone)]
#[repr(C)]
union RStringAux {
capa: c_long,
value: InternalValue,
}
// Ruby 3.2: the length is in `as.heap.len`, or `as.embed.len` at the same
// place for an embedded string.
#[cfg(ruby_3_2)]
#[derive(Copy, Clone)]
#[repr(C)]
struct RStringHeap {
len: c_long,
ptr: *const c_char,
aux: RStringAux,
}
#[cfg(ruby_3_2)]
#[repr(C)]
struct RString {
basic: RBasic,
as_: RStringAs,
}
// Ruby 3.3 moves the length to the top level for every string; `as.heap`
// keeps `ptr` and `aux`, and an embedded string's bytes start at `as`.
#[cfg(ruby_gte_3_3)]
#[derive(Copy, Clone)]
#[repr(C)]
struct RStringHeap {
ptr: *const c_char,
aux: RStringAux,
}
#[cfg(ruby_gte_3_3)]
#[repr(C)]
struct RString {
basic: RBasic,
len: c_long,
as_: RStringAs,
}
unsafe fn rstring_and_flags(value: Value) -> (*const RString, InternalValue) {
let rstring: *const RString = value.value as *const RString;
let flags = (*rstring).basic.flags;
(rstring, flags)
}
unsafe fn embed_check(flags: InternalValue) -> bool {
flags & (RStringEmbed::NoEmbed as size_t) == 0
}
// Ruby 3.2 (`USE_RVARGC`) keeps it in `as.embed.len`, which sits where
// `as.heap.len` does.
#[cfg(ruby_3_2)]
pub unsafe fn rstring_embed_len(value: Value) -> c_long {
let (rstring, _flags) = rstring_and_flags(value);
(*rstring).as_.heap.len
}
#[cfg(ruby_gte_3_3)]
pub unsafe fn rstring_embed_len(value: Value) -> c_long {
let (rstring, _flags) = rstring_and_flags(value);
(*rstring).len
}
// In Ruby 3.2 `as.embed.ary` follows `as.embed.len`.
#[cfg(ruby_3_2)]
unsafe fn rstring_embed_ptr(rstring: *const RString) -> *const c_char {
(&(*rstring).as_ as *const RStringAs as *const c_char).add(mem::size_of::<c_long>())
}
// In Ruby 3.3 `as.embed.ary` is all of `as`.
#[cfg(ruby_gte_3_3)]
unsafe fn rstring_embed_ptr(rstring: *const RString) -> *const c_char {
&(*rstring).as_ as *const RStringAs as *const c_char
}
#[cfg(ruby_3_2)]
pub unsafe fn rstring_len(value: Value) -> c_long {
let (rstring, flags) = rstring_and_flags(value);
if embed_check(flags) {
rstring_embed_len(value)
} else {
(*rstring).as_.heap.len
}
}
#[cfg(ruby_gte_3_3)]
pub unsafe fn rstring_len(value: Value) -> c_long {
let (rstring, _flags) = rstring_and_flags(value);
(*rstring).len
}
pub unsafe fn rstring_ptr(value: Value) -> *const c_char {
let (rstring, flags) = rstring_and_flags(value);
if embed_check(flags) {
rstring_embed_ptr(rstring)
} else {
(*rstring).as_.heap.ptr
}
}
pub unsafe fn rstring_end(value: Value) -> *const c_char {
rstring_ptr(value).add(rstring_len(value) as usize)
}
// ```
// use rutie::VM;
// # VM::init();
//
// use rutie::binding::string::*; // binding not public
//
// let word = new_utf8("word");
// unsafe {
// assert!(!is_locktmp(word), "word should not be locktmp but is");
// locktmp(word);
// assert!(is_locktmp(word), "word should be locktmp but is not");
// unlocktmp(word);
// assert!(!is_locktmp(word), "word should not be locktmp but is");
// }
// ```
pub unsafe fn is_lockedtmp(value: Value) -> bool {
let (_rstring, flags) = rstring_and_flags(value);
flags & STR_TMPLOCK as size_t != 0
}
#[cfg(test)]
mod tests {
use super::{rstring_end, rstring_len, rstring_ptr};
use crate::{Array, Fixnum, Object, RString, VM};
// Strings of every length across the embedded/heap boundary (the slot
// size), made in several ways, read directly and compared with what
// Ruby reports.
#[test]
fn test_direct_rstring_reads() {
crate::on_ruby_thread(|| {
let strings = VM::eval(
"(0..700).step(1).flat_map do |n|
s = (('a'..'z').to_a.join * 30)[0, n]
[s, s.dup, String.new(s, capacity: 1000), \"#{s}\", (\"_\" + s)[1..], s.b]
end",
)
.unwrap();
let strings = Array::from(strings.value());
for i in 0..strings.length() {
let string = strings.at(i as i64);
let value = string.value();
let bytesize = unsafe { string.send("bytesize", &[]) };
let expected = bytesize.try_convert_to::<Fixnum>().unwrap().to_i64();
let len = unsafe { rstring_len(value) };
assert_eq!(len as i64, expected, "length of string #{}", i);
let bytes = unsafe {
std::slice::from_raw_parts(rstring_ptr(value) as *const u8, len as usize)
};
let alphabet: Vec<u8> = (b'a'..=b'z').cycle().take(len as usize).collect();
assert_eq!(bytes, &alphabet[..], "bytes of string #{}", i);
assert_eq!(
unsafe { rstring_end(value).offset_from(rstring_ptr(value)) },
len as isize
);
}
let empty = RString::new_utf8("");
assert_eq!(unsafe { rstring_len(empty.value()) }, 0);
});
}
}