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
/// `ObjectCache` provides a way to calculate and cache values for each node
/// in a clvm object tree. It can be used to calculate the sha256 tree hash
/// for an object and save the hash for all the child objects for building
/// usage tables, for example.
///
/// It also allows a function that's defined recursively on a clvm tree to
/// have a non-recursive implementation (as it keeps a stack of uncached
/// objects locally).
use crate::allocator::{Allocator, NodePtr, SExp};
use std::collections::HashMap;
type CachedFunction<T> = fn(&mut ObjectCache<T>, &Allocator, NodePtr) -> Option<T>;
use super::bytes32::{Bytes32, hash_blobs};
use crate::serde::serialized_length_atom;
pub struct ObjectCache<T> {
cache: HashMap<NodePtr, T>,
/// The function `f` is expected to calculate its T value recursively based
/// on the T values for the left and right child for a pair. For an atom, the
/// function f must calculate the T value directly.
///
/// If a pair is passed and one of the children does not have its T value cached
/// in `ObjectCache` yet, return `None` and f will be called with each child in turn.
/// Don't recurse in f; that's the point of this structure.
f: CachedFunction<T>,
}
impl<T: Clone> ObjectCache<T> {
pub fn new(f: CachedFunction<T>) -> Self {
Self {
cache: HashMap::new(),
f,
}
}
/// return the function value for this node, either from cache
/// or by calculating it. If the stop_token is specified and is found in the
/// CLVM tree below node, traversal will stop and `None` is returned.
pub fn get_or_calculate(
&mut self,
allocator: &Allocator,
node: &NodePtr,
stop_token: Option<NodePtr>,
) -> Option<&T> {
self.calculate(allocator, node, stop_token);
self.get_from_cache(node)
}
/// return the cached value for this node, or `None`
fn get_from_cache(&self, node: &NodePtr) -> Option<&T> {
self.cache.get(node)
}
/// set the cached value for a node
fn set(&mut self, node: &NodePtr, v: T) {
self.cache.insert(*node, v);
}
/// calculate the function's value for the given node, traversing uncached children
/// as necessary. If, the optional, stop_token NodePtr is encountered in the
/// sub tree of root_node, we stop calculations and don't add the the value
/// for root_node to the cache. This is / used for accessing incrementally
/// built trees, where the stop_token indicates an unfinished part of the
/// structure.
fn calculate(
&mut self,
allocator: &Allocator,
root_node: &NodePtr,
stop_token: Option<NodePtr>,
) {
let mut obj_list = vec![*root_node];
while let Some(node) = obj_list.pop() {
if stop_token == Some(node) {
// we must terminate the search if we hit the stop_token. We can't
// traverse past it (since we're serializing incrementally).
return;
}
let v = self.get_from_cache(&node);
match v {
Some(_) => {}
None => match (self.f)(self, allocator, node) {
None => match allocator.sexp(node) {
SExp::Pair(left, right) => {
obj_list.push(node);
obj_list.push(left);
obj_list.push(right);
}
_ => panic!("f returned `None` for atom"),
},
Some(v) => {
self.set(&node, v);
}
},
}
}
}
}
/// calculate the standard `sha256tree` has for a node
pub fn treehash(
cache: &mut ObjectCache<Bytes32>,
allocator: &Allocator,
node: NodePtr,
) -> Option<Bytes32> {
match allocator.sexp(node) {
SExp::Pair(left, right) => match cache.get_from_cache(&left) {
None => None,
Some(left_value) => cache
.get_from_cache(&right)
.map(|right_value| hash_blobs(&[&[2], left_value, right_value])),
},
SExp::Atom => Some(hash_blobs(&[&[1], allocator.atom(node).as_ref()])),
}
}
/// calculate the serialized length (without backrefs) of a node. This is used
/// to check if using backrefs is actually smaller.
pub fn serialized_length(
cache: &mut ObjectCache<u64>,
allocator: &Allocator,
node: NodePtr,
) -> Option<u64> {
match allocator.sexp(node) {
SExp::Pair(left, right) => match cache.get_from_cache(&left) {
None => None,
Some(left_value) => cache.get_from_cache(&right).map(|right_value| {
1_u64
.saturating_add(*left_value)
.saturating_add(*right_value)
}),
},
SExp::Atom => {
let buf = allocator.atom(node);
Some(serialized_length_atom(buf.as_ref()) as u64)
}
}
}
#[cfg(test)]
mod tests {
use super::*;
use hex::FromHex;
use std::cmp::max;
use std::fmt::Debug;
use std::io::Cursor;
use crate::serde::de::node_from_stream;
/// calculate the depth of a node. Used for tests
fn calculate_depth_simple(
cache: &mut ObjectCache<usize>,
allocator: &Allocator,
node: NodePtr,
) -> Option<usize> {
match allocator.sexp(node) {
SExp::Pair(left, right) => match cache.get_from_cache(&left) {
None => None,
Some(left_value) => cache
.get_from_cache(&right)
.map(|right_value| 1 + max(*left_value, *right_value)),
},
SExp::Atom => Some(0),
}
}
fn check_cached_function<T>(obj_as_hex: &str, expected_value: T, f: CachedFunction<T>)
where
T: Clone + Eq + Debug,
{
let mut allocator = Allocator::new();
let blob: Vec<u8> = Vec::from_hex(obj_as_hex).unwrap();
let mut cursor: Cursor<&[u8]> = Cursor::new(&blob);
let obj = node_from_stream(&mut allocator, &mut cursor).unwrap();
let mut oc = ObjectCache::new(f);
assert_eq!(oc.get_from_cache(&obj), None);
oc.calculate(&allocator, &obj, None);
assert_eq!(oc.get_from_cache(&obj), Some(&expected_value));
assert_eq!(
oc.get_or_calculate(&allocator, &obj, None).unwrap().clone(),
expected_value
);
assert_eq!(oc.get_from_cache(&obj), Some(&expected_value));
// do it again, but the simple way
let mut oc = ObjectCache::new(f);
assert_eq!(
oc.get_or_calculate(&allocator, &obj, None).unwrap().clone(),
expected_value
);
}
#[test]
fn test_depths_cache() {
let check = |a, b| check_cached_function(a, b, calculate_depth_simple);
check("01", 0); // 1
check("ff83666f6f83626172", 1); // (foo . bar)
check("ff83666f6fff8362617280", 2); // (foo bar)
check("ffff0102ff0304", 2); // ((1 . 2) . (3 . 4))
check("ff01ff02ff03ff04ff05ff0680", 6); // (1 2 3 4 5 6)
}
#[test]
fn test_treehash() {
let check = |a, b| check_cached_function(a, Bytes32::from_hex(b).unwrap(), treehash);
check(
"ff83666f6f83626172",
"c518e45ae6a7b4146017b7a1d81639051b132f1f5572ce3088a3898a9ed1280b",
); // (foo . bar)
check(
"ff83666f6fff8362617280",
"c97d97cc81100a4980080ba81ff1ba3985f7cff1db9d41d904b9d512bb875144",
); // (foo bar)
check(
"ffff0102ff0304",
"2824018d148bc6aed0847e2c86aaa8a5407b916169f15b12cea31fa932fc4c8d",
); // ((1 . 2) . (3 . 4))
check(
"ff01ff02ff03ff04ff05ff0680",
"65de5098d18bebd62aee37de32f0b62d1803d9c7c48f10dca25501243d7a0392",
); // (1 2 3 4 5 6)
}
#[test]
fn test_serialized_length() {
let check = |a, b| check_cached_function(a, b, serialized_length);
check("ff83666f6f83626172", 9); // (foo . bar)
check("ff83666f6fff8362617280", 11); // (foo bar)
check("ffff0102ff0304", 7); // ((1 . 2) . (3 . 4))
check("ff01ff02ff03ff04ff05ff0680", 13); // (1 2 3 4 5 6)
}
#[test]
#[ignore = "slow: run with `cargo test -- --include-ignored`"]
fn test_very_long_list() {
// in this test, we check that `treehash` and `serialized_length` can handle very deep trees that
// would normally blow out the stack. It's expensive to create such a long list, so we do both
// tests here so we only have to to create the list once
const LIST_SIZE: u64 = 20_000_000;
let mut allocator = Allocator::new();
let mut top = allocator.nil();
for _ in 0..LIST_SIZE {
let atom = allocator.one();
top = allocator.new_pair(atom, top).unwrap();
}
let expected_value = LIST_SIZE * 2 + 1;
let mut oc = ObjectCache::new(serialized_length);
assert_eq!(
oc.get_or_calculate(&allocator, &top, None).unwrap().clone(),
expected_value
);
let expected_value = <[u8; 32]>::from_hex(
"a168fce695099a30c0745075e6db3722ed7f059e0d7cc4d7e7504e215db5017b",
)
.unwrap();
let mut oc = ObjectCache::new(treehash);
assert_eq!(
oc.get_or_calculate(&allocator, &top, None).unwrap().clone(),
expected_value
);
}
fn do_check_token(
allocator: &Allocator,
stop_token: NodePtr,
poisoned_nodes: &[NodePtr],
good_nodes: &[NodePtr],
) {
let mut cache = ObjectCache::new(treehash);
for n in poisoned_nodes {
assert!(
cache
.get_or_calculate(allocator, n, Some(stop_token))
.is_none()
);
}
for n in good_nodes {
assert!(
cache
.get_or_calculate(allocator, n, Some(stop_token))
.is_some()
);
}
}
#[test]
fn test_stop_token() {
// we build a tree and insert a stop_token and ensure we get `None` in
// the appropriate places in the tree
// A
// / \
// B C
// / \ / \
// D E F G
// if F is made the stop-token F, C and A should return None.
let mut allocator = Allocator::new();
let d = allocator.new_atom(b"d").unwrap();
let e = allocator.new_atom(b"e").unwrap();
let f = allocator.new_atom(b"f").unwrap();
let g = allocator.new_atom(b"g").unwrap();
let b = allocator.new_pair(d, e).unwrap();
let c = allocator.new_pair(f, g).unwrap();
let a = allocator.new_pair(b, c).unwrap();
// if d is the stop token; d,b and a should return None
do_check_token(&allocator, d, &[d, b, a], &[e, c, f, g]);
do_check_token(&allocator, e, &[e, b, a], &[d, c, f, g]);
do_check_token(&allocator, f, &[f, c, a], &[d, e, g, b]);
do_check_token(&allocator, g, &[g, c, a], &[d, e, b, f]);
}
}