fcode 1.1.0

A binary serialization/deserialization strategy for Serde that supports schema evolution
Documentation 
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

use crate::error::{Error, Result};
use std::io::Write;

// A tag byte has the wire type in the low 3 bits. If the wire type is a varint
// (Int, Sequence, Bytes, Variant), then it additionally has 4 bits of value,
// and a stop bit as bit 7.

#[repr(u8)]
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum WireType {
	Int = 0, // varint, up to 128 bits of data
	Fixed32 = 1,
	Fixed64 = 2,
	Sequence = 3, // varint length followed by this many encoded items
	Bytes = 4,    // varint length, followed by u8 data
	Variant = 5,  // varint discriminator, followed by single item; for Option it's 0 (None) or 1 (Some)
	_Reserved1 = 6,
	_Reserved2 = 7,
}

#[inline]
pub fn read_wiretype(tagbyte: u8) -> WireType {
	let tag = tagbyte & 7;
	// safety: this is safe because every value from 0 to 7 is a valid wire type
	unsafe { std::mem::transmute(tag) }
}

// write a varint together with the wiretype tag
#[inline]
pub fn write_varint(writer: &mut impl Write, tag: WireType, mut value: u64) -> Result<()> {
	let tag = tag as u8;
	let partial = ((value & 15) << 3) as u8;
	value >>= 4;
	if value == 0 {
		writer.write_all(&[tag | partial])?;
		return Ok(());
	}
	// 10 bytes supports 4 + 9 * 7 = 67 bits of data
	let mut b = [0u8; 10];
	b[0] = tag | partial | 0x80;
	let mut len = 1;
	loop {
		let partial = (value & 0x7f) as u8;
		value >>= 7;
		if value == 0 {
			// safety: we've calculcated that the local array can support the value
			unsafe {
				*b.get_unchecked_mut(len) = partial;
			}
			len += 1;
			break;
		}
		// safety: we've calculcated that the local array can support the value
		unsafe {
			*b.get_unchecked_mut(len) = partial | 0x80;
		}
		len += 1;
	}
	writer.write_all(&b[..len])?;
	Ok(())
}

// read a varint, given a tag byte and remaining data; returns the value and
// the size consumed from data
#[inline]
pub fn read_varint(tagbyte: u8, data: &[u8]) -> Result<(u64, usize)> {
	if tagbyte & 0x80 == 0 {
		let value = tagbyte >> 3;
		return Ok((value as u64, 0));
	}
	let mut value = ((tagbyte & 0x7f) >> 3) as u64;
	let mut shift = 4;
	// note I've tested with a fast/slow variant, where the fast variant doesn't need
	// to check for end of input, but it doesn't make it faster; the test is negligible
	for (i, b) in data.iter().copied().enumerate() {
		if shift >= 64 {
			return Err(Error::ValueOverflow);
		}
		if b & 0x80 == 0 {
			value |= (b as u64) << shift;
			return Ok((value, i + 1));
		}
		value |= ((b & 0x7f) as u64) << shift;
		shift += 7;
	}
	Err(Error::UnexpectedEndOfInput)
}

#[inline]
pub fn skip_varint(tagbyte: u8, data: &[u8]) -> Result<usize> {
	if tagbyte & 0x80 == 0 {
		return Ok(0);
	}
	for (i, b) in data.iter().copied().enumerate() {
		// if we reach byte 18, we've consumed 19 bytes including tag byte, exceeding
		// max encoding of a 128-bit varint
		if i == 18 {
			return Err(Error::ValueOverflow);
		}
		if b & 0x80 == 0 {
			return Ok(i + 1);
		}
	}
	Err(Error::UnexpectedEndOfInput)
}

#[test]
fn test_varint() {
	let mut buf = vec![];

	write_varint(&mut buf, WireType::Int, 15).unwrap();
	assert_eq!(buf.len(), 1);
	assert_eq!(read_varint(buf[0], &buf[1..]).unwrap(), (15, 0));

	buf.clear();
	write_varint(&mut buf, WireType::Int, u64::MAX).unwrap();
	assert_eq!(buf.len(), 10);
	assert_eq!(read_varint(buf[0], &buf[1..]).unwrap(), (u64::MAX, 9));
}

serde::serde_if_integer128! {
	#[inline]
	pub fn write_varint_128(writer: &mut impl Write, tag: WireType, mut value: u128) -> Result<()> {
		let tag = tag as u8;
		let partial = ((value & 15) << 3) as u8;
		value >>= 4;
		if value == 0 {
			writer.write_all(&[tag | partial])?;
			return Ok(());
		}
		// 19 bytes supports 4 + 18 x 7 = 130 bits of data
		let mut b = [0u8; 19];
		b[0] = tag | partial | 0x80;
		let mut len = 1;
		loop {
			let partial = (value & 0x7f) as u8;
			value >>= 7;
			if value == 0 {
				// safety: we've calculcated that the local array can support the value
				unsafe { *b.get_unchecked_mut(len) = partial; }
				len += 1;
				break;
			}
			// safety: we've calculcated that the local array can support the value
			unsafe { *b.get_unchecked_mut(len) = partial | 0x80; }
			len += 1;
		}
		writer.write_all(&b[..len])?;
		Ok(())
	}

	#[inline]
	pub fn read_varint_128(tagbyte: u8, data: &[u8]) -> Result<(u128, usize)> {
		if tagbyte & 0x80 == 0 {
			let value = tagbyte >> 3;
			return Ok((value as u128, 0));
		}
		let mut value = ((tagbyte & 0x7f) >> 3) as u128;
		let mut shift = 4;
		for (i,b) in data.iter().copied().enumerate() {
			if shift >= 128 {
				return Err(Error::ValueOverflow);
			}
			if b & 0x80 == 0 {
				value |= (b as u128) << shift;
				return Ok((value, i + 1));
			}
			value |= ((b & 0x7f) as u128) << shift;
			shift += 7;
		}
		Err(Error::UnexpectedEndOfInput)
	}

	#[test]
	fn test_varint_128() {
		let mut buf = vec![];

		write_varint_128(&mut buf, WireType::Int, u128::MAX).unwrap();
		assert_eq!(buf.len(), 19);
		assert_eq!(read_varint_128(buf[0], &buf[1..]).unwrap(), (u128::MAX, 18));
	}
}

// signed varints use google's zig-zag method

#[inline]
pub fn zigzag_encode(value: i64) -> u64 {
	let encoded = (value << 1) ^ (value >> 63);
	encoded as u64
}

#[inline]
pub fn zigzag_decode(encoded: u64) -> i64 {
	(encoded >> 1) as i64 ^ -(encoded as i64 & 1)
}

#[test]
fn test_zigzag() {
	assert_eq!(zigzag_decode(zigzag_encode(42)), 42);
	assert_eq!(zigzag_decode(zigzag_encode(-42)), -42);
	assert_eq!(zigzag_decode(zigzag_encode(0)), 0);
	assert_eq!(zigzag_decode(zigzag_encode(i64::MAX)), i64::MAX);
	assert_eq!(zigzag_decode(zigzag_encode(i64::MIN)), i64::MIN);

	assert!(zigzag_encode(-10) < zigzag_encode(100));
	assert!(zigzag_encode(10) < zigzag_encode(-100));
}

serde::serde_if_integer128! {
	#[inline]
	pub fn zigzag_encode_128(value: i128) -> u128 {
		let encoded = (value << 1) ^ (value >> 127);
		encoded as u128
	}


	#[inline]
	pub fn zigzag_decode_128(encoded: u128) -> i128 {
		(encoded >> 1) as i128 ^ -(encoded as i128 & 1)
	}

	#[test]
	fn test_zigzag_128() {
		assert_eq!(zigzag_decode_128(zigzag_encode_128(42)), 42);
		assert_eq!(zigzag_decode_128(zigzag_encode_128(-42)), -42);
		assert_eq!(zigzag_decode_128(zigzag_encode_128(i128::MAX)), i128::MAX);
		assert_eq!(zigzag_decode_128(zigzag_encode_128(i128::MIN)), i128::MIN);

		assert_eq!(zigzag_decode_128(zigzag_encode(42) as u128), 42);
		assert_eq!(zigzag_decode_128(zigzag_encode(-42) as u128), -42);

		assert_eq!(zigzag_decode(zigzag_encode_128(42) as u64), 42);
		assert_eq!(zigzag_decode(zigzag_encode_128(-42) as u64), -42);
	}
}