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
|
//! This module defines functions to turn a forest of forest labels
//! into a sequence of bytes.
//!
//! To be more specific, a forest consists of two parts: a vector of
//! node labels and a graph specifying the relations between nodes.
//!
//! # Number format (endianness)
//!
//! Every number mentionned in this format is an unsigned integer of
//! 64 bits, or 8 bytes. Every such number is specified in the big
//! endian format.
//!
//! # Parts of the sequence of bytes
//!
//! The sequence of bytes has three parts:
//!
//! ## Header
//!
//! The header specifies metadata of this forest.
//!
//! ### Special mark
//!
//! The first three bytes form the string "rep", as a special mark.
//!
//! ### Number of nodes
//!
//! The next 8 bytes specifies the number of nodes of this forest.
//!
//! ## Graph
//!
//! Next comes the underlying graph for this forest. This part
//! consists of a vector of vectors of numbers. Each vector of
//! numbers represents the list of children of a node. So the number
//! of vectors of numbers is equal to the number of nodes.
//!
//! ### Vector of vectors of numbers
//!
//! The vectors are not simply concatenated one after another, as that
//! way one cannot read a random node in a constant amount of time.
//!
//! Instead, we first specify the number of children of each node
//! first, along with the offset for that node of the vector of
//! children.
//!
//! As an example, if a graph has three nodes, represented as the
//! adjacency list: `[[1, 2], [2], []]`, then its representation as a
//! sequence of bytes is as follows:
//! ```text
//! 2, x, 1, y, 0, z, 1, 2, 2,
//! ^ ^ ^
//! x y z
//! ```
//!
//! This has the advantage that we can read the children of the `n`-th
//! node in constant time.
//!
//! ## Vector of labels
//!
//! Each label occupies a fixed number of bytes, so we simply put the
//! labels one after another. The only thing to note here is the
//! format of the labels.
//!
//! ### Labels
//!
//! Each label has 3 parts:
//!
//! 1. Status: either Packed, Cloned, or Plain. If the node is
//! cloned, it has a clone index. So in total this part occupies 1
//! byte for the status and 8 bytes for the clone index.
//! 2. Start and end: the range in the input sentence. We just store
//! two numbers here. Hence this part occupies 16 bytes.
//! 3. Grammar label: either a terminal, a non-terminal, or a rule.
//! Each variant needs a number to speify its index. So in total this
//! part occupies 1 byte for the variant discriminant and 8 bytes for
//! the number.
//!
//! To sum up, each label occupies 34 bytes.
use super::Forest;
use grammar::{GrammarLabel, GrammarLabelType, TNT};
/// Convert any type that implements the `Forest` trait into a
/// sequence of bytes.
pub fn forest_to_bytes<F: Forest<GrammarLabel>>(forest: &F) -> Vec<u8> {
// First calculate the total number of bytes.
let nodes_len = forest.nodes_len();
let degrees: Vec<_> = forest
.nodes()
.map(|node| forest.degree(node).unwrap_or(0))
.collect();
let total_degree: usize = degrees.iter().copied().sum();
let len: usize = 8 // total number of bytes at the start
+ 3 // special mark
+ 8 // number of nodes
+ 8 // offset of labels
+ 16 * nodes_len // degree & offset for each node
+ 8 * total_degree // children of each node
+ 34 * nodes_len // labels
;
// Then fill in the bytes.
let mut bytes: Vec<u8> = Vec::with_capacity(len);
// First the headers
bytes.extend(len.to_be_bytes());
bytes.extend([114, 101, 112]); // rep
bytes.extend(nodes_len.to_be_bytes());
bytes.extend((len - 34 * nodes_len).to_be_bytes());
let mut accumulated: usize = 0;
// Then degrees and offsets
for degree in degrees.iter().copied() {
bytes.extend(degree.to_be_bytes());
bytes.extend((8 + 3 + 8 + 8 + 16 * nodes_len + 8 * accumulated).to_be_bytes());
accumulated += degree;
}
// Then the children
bytes.extend(forest.nodes().flat_map(|node| {
forest
.children_of(node)
.unwrap()
.flat_map(|child| child.to_be_bytes())
}));
// Finally the labels
'nodes_loop: for node in forest.nodes() {
let label = match forest.vertex_label(node) {
Ok(Some(label)) => label,
_ => continue 'nodes_loop,
};
if label.is_plain() {
bytes.extend(std::iter::repeat(0).take(9));
} else if label.is_packed() {
bytes.extend(std::iter::once(1).chain(std::iter::repeat(0).take(8)));
} else if let Some(index) = label.clone_index() {
bytes.extend(std::iter::once(2).chain(index.to_be_bytes()));
}
let label = label.label();
bytes.extend(label.start().to_be_bytes());
bytes.extend(label.end().unwrap_or(0).to_be_bytes());
let label = label.label();
match label {
GrammarLabelType::TNT(TNT::Ter(t)) => {
bytes.extend(std::iter::once(0).chain(t.to_be_bytes()));
}
GrammarLabelType::TNT(TNT::Non(n)) => {
bytes.extend(std::iter::once(1).chain(n.to_be_bytes()));
}
GrammarLabelType::Rule(r) => {
bytes.extend(std::iter::once(2).chain(r.to_be_bytes()));
}
}
}
if bytes.len() != len {
dbg!();
}
bytes
}
|
