-
Notifications
You must be signed in to change notification settings - Fork 0
/
encoding.zig
258 lines (199 loc) · 8.85 KB
/
encoding.zig
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
const std = @import("std");
const _heap = @import("heap.zig");
const Tree = @import("tree.zig");
const Utils = @import("utils.zig");
const Heap = _heap.Heap;
const TreeNode = Tree.TreeNode;
const BUFFER_SIZE = 10;
fn lessThanTree(A: TreeNode, B: TreeNode) std.math.Order {
// We want to have the lowest probability be highest priority, hence the inversion.
return std.math.order(B.probability, A.probability);
}
//
// Handle the encoding from a file path.
// And write endoding to an output file
//
pub fn encode(file_path: [:0]const u8, output_path: [:0]const u8) !void {
const page_alloc = std.heap.page_allocator;
const findFile = try std.fs.cwd().openFile(file_path, .{});
const max_size: usize = 999999;
var content = try findFile.reader().readAllAlloc(page_alloc, max_size);
const wordSize = content.len;
defer page_alloc.free(content);
// Map between ASCII and frequency
var charMap = std.AutoHashMap(u8, i32).init(page_alloc);
for (content) |char| {
const mapChar = charMap.get(char);
if (mapChar != null) {
try charMap.put(char, mapChar.? + 1);
} else {
try charMap.put(char, 1);
}
}
var arena = std.heap.ArenaAllocator.init(page_alloc);
const allocator = arena.allocator();
defer arena.deinit();
var heap = try Heap(TreeNode, lessThanTree).init(page_alloc);
var mapIter = charMap.iterator();
var item_counter: u16 = 0;
while (mapIter.next()) |entry| {
const node = try allocator.create(TreeNode);
node.* = .{
.value = entry.key_ptr.*,
.probability = entry.value_ptr.*,
.left_child = null,
.right_child = null,
};
item_counter += 1;
heap.insert(node.*);
}
// At least 2 items in heap.
while (heap.pointer > 1) {
const left = heap.remove();
const right = heap.remove();
// Because we would return the items that are in the array of the heap.
// And the pointers in the array don't change, we would end up with
// Recursive relations A -> B -> A
// So to fix this I create copies, seperate.
// I don't think this is ideal.
const copies = try allocator.alloc(TreeNode, 3);
copies[0] = .{
.value = left.?.value,
.probability = left.?.probability,
.left_child = left.?.left_child,
.right_child = left.?.right_child,
};
copies[1] = .{
.value = right.?.value,
.probability = right.?.probability,
.left_child = right.?.left_child,
.right_child = right.?.right_child,
};
copies[2] = .{
.value = 0,
.probability = left.?.probability + right.?.probability,
.left_child = &copies[0],
.right_child = &copies[1],
};
item_counter += 1;
heap.insert(copies[2]);
}
const nodePointer = heap.remove().?;
var my_bitmasks = try allocator.alloc(u64, BUFFER_SIZE);
var bitmask_index: usize = 0;
my_bitmasks[bitmask_index] = 0;
var bitmask: *u64 = &my_bitmasks[bitmask_index];
var bitmask_used: u16 = 0;
for (content) |char| {
// Pattern starting with 1, so that we don't lose the left 0s
var pattern = Tree.walk(nodePointer, char, 1).?;
// -1 because our pattern contains a left most 1, to not lose the 0s.
var bits_used: u16 = Utils.count_bits_used(pattern) - 1;
const remaining = @bitSizeOf(u64) - bitmask_used;
if (remaining < bits_used) {
var clean_pattern = pattern - std.math.pow(u16, 2, bits_used);
var split_pattern = @as(u64, clean_pattern >> @truncate(u4, bits_used - remaining));
bitmask.* |= split_pattern;
bitmask_index += 1;
// We must relloac, as we reached end of our buffer.
if (bitmask_index == my_bitmasks.len) {
my_bitmasks = try allocator.realloc(my_bitmasks, my_bitmasks.len + BUFFER_SIZE);
}
my_bitmasks[bitmask_index] = 0;
bitmask = &my_bitmasks[bitmask_index];
bitmask_used = 0;
// We must take whats left of this current pattern.
var helper: u16 = @as(u16, 0xFFFF) >> @truncate(u4, 16 - (bits_used - remaining));
pattern = clean_pattern & helper;
pattern = pattern + std.math.pow(u16, 2, bits_used - remaining);
bits_used = bits_used - remaining;
}
// Probably a better way to do this.
// I tried bitshifting, but zig doesn't like shifting by non comptime values.
var clean_pattern = pattern - std.math.pow(u16, 2, bits_used);
var shifted_pattern = @as(u64, clean_pattern) << @truncate(u6, (64 - bits_used - bitmask_used));
bitmask.* |= shifted_pattern;
bitmask_used += bits_used;
}
const file = try std.fs.cwd().createFile(output_path, .{ .read = true });
const writer = file.writer();
var preOrderArr = try allocator.alloc(u8, item_counter);
var preOrderIndex: usize = 0;
Tree.pre_order_traversal(nodePointer, preOrderArr, &preOrderIndex);
// Format
// TreeSize (u16) --- PreOrder [TreeSize]u8 --- HuffmanCode []u8
const size_as_u8 = std.mem.asBytes(&item_counter);
const word_size_as_u8 = std.mem.asBytes(&wordSize);
try writer.writeAll(size_as_u8[0..]);
try writer.writeAll(word_size_as_u8[0..]);
try writer.writeAll(preOrderArr[0..]);
var c: usize = 0;
while (c <= bitmask_index) {
try writer.writeAll(std.mem.asBytes(&my_bitmasks[c]));
c += 1;
}
}
pub fn decompress(file_path: [:0]const u8, output_path: [:0]const u8) !void {
const page_alloc = std.heap.page_allocator;
var arena = std.heap.ArenaAllocator.init(page_alloc);
const allocator = arena.allocator();
defer arena.deinit();
const findFile = try std.fs.cwd().openFile(file_path, .{});
const max_size: usize = 999999;
var content = try findFile.reader().readAllAlloc(allocator, max_size);
const tree_size = @intCast(u16, content[0]) + (@intCast(u16, content[1]) << 8);
const word_size = @as(usize, Utils.convert_to_u64(content[2 .. @sizeOf(usize) + 2]));
const offset = @sizeOf(u16) + @sizeOf(usize);
var index: usize = 0;
var tree = try Tree.build_simple_tree(allocator, content[offset..(tree_size + offset)], &index);
// Mirrored so we can do % 2 trick to know to go left or right.
const bitmask_offset = tree_size + offset;
var bitmasks_used: usize = 0;
var bitmask: u64 = 0;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset]) << 0;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 1]) << 8;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 2]) << 16;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 3]) << 24;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 4]) << 32;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 5]) << 40;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 6]) << 48;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 7]) << 56;
var word = try allocator.alloc(u8, word_size);
var char_counter: usize = 0;
var current_node: *const TreeNode = tree;
var mirrored_bitmask = Utils.mirror_bitmask(bitmask);
var counter: u8 = 0;
while (counter < 64 and char_counter < word_size) {
const first_bit = mirrored_bitmask & 1;
if (first_bit == 1) {
current_node = current_node.right_child.?;
} else {
current_node = current_node.left_child.?;
}
if (current_node.value != 0) {
word[char_counter] = current_node.*.value;
current_node = tree;
char_counter += 1;
}
mirrored_bitmask = mirrored_bitmask >> 1;
if (counter != 63) {
counter += 1;
continue;
}
counter = 0;
bitmask = 0;
bitmasks_used += 1;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset]) << 0;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 1]) << 8;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 2]) << 16;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 3]) << 24;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 4]) << 32;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 5]) << 40;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 6]) << 48;
bitmask |= @as(u64, content[bitmasks_used * 8 + bitmask_offset + 7]) << 56;
mirrored_bitmask = Utils.mirror_bitmask(bitmask);
}
const file = try std.fs.cwd().createFile(output_path, .{ .read = true });
const writer = file.writer();
try writer.writeAll(word[0..]);
}