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
|
netencode(5) "netencode manual"
# NAME
netencode - Length-prefixed, type-safe data serialization format
# SYNOPSIS
A netencode record containing user information:
```
{49:<6:active|<4:true|u,<3:age|i:30,<4:name|t5:Alice,}
```
A netencode list of text values:
```
[23:t5:hello,t5:world,i:42,]
```
Binary data with length prefix:
```
b13:Hello, World!,
```
# DESCRIPTION
Netencode is a *bencode* and *netstring* inspired data serialization format
designed for Unix pipelines and command-line tools. It provides type-safe,
length-prefixed encoding that is both human-readable for debugging and
machine-efficient for parsing.
The format is designed to be trivial to generate (requiring only *byte_length()*
and *printf()*) and easy to parse (complete parsers can be written in under
100 lines of code).
Key features:
- Length-prefixed variable data for efficient streaming
- Explicit type information for all values
- ASCII prefixes make format partially human-readable
- No escaping required for embedded quotes or special characters
- Binary data support for arbitrary byte sequences
# SCALAR TYPES
All netencode values follow the pattern: *[type][size]:[data][terminator]*
Where *size* is a natural number without leading zeroes, and *terminator* is
always a comma (,).
## Unit Type
The unit type represents a value with no data content.
```
Format: u,
Example: u,
```
## Numbers
All numbers are 64-bit values. Floating-point numbers are not supported.
```
Natural numbers: n:[value],
Range: 0 to 18446744073709551615
Example: n:42,
Signed integers: i:[value],
Range: -9223372036854775808 to 9223372036854775807
Example: i:-42,
Note: JSON numbers are converted to signed integers
```
## Booleans
Booleans are represented as tagged unit values.
```
Boolean true: <4:true|u,
Boolean false: <5:false|u,
```
## Text Strings
UTF-8 encoded text with byte-length prefix. No escaping is required.
```
Format: t[size]:[text],
Example with regular text: t11:hello world,
Example with quotes: t12:He said "hi",
Empty string: t0:,
```
## Binary Data
Arbitrary byte sequences with length prefix. Can contain null bytes and
any binary data.
```
Format: b[size]:[data],
Example: b5:hello,
Empty binary: b0:,
```
# COMPOSITE TYPES
## Tagged Values (Sum Types)
A tag (<) gives a value a name, creating sum types or tagged unions.
```
Format: <[tag_size]:[tag]|[value]
Example: <3:foo|t5:hello,
Empty tag: <0:|i:42,
Nested tag: <4:Some|<4:data|t6:secret,
```
## Records (Maps)
A record ({) is a concatenation of tagged values, closed with }. Records
start with the byte length of their entire encoded content.
```
Format: {[size]:[tag1][tag2]...}
Multiple fields: {21:<3:foo|u,<1:x|t3:baz,}
Single field: {9:<3:foo|u,}
```
*Note*: If tag names repeat, earlier occurrences are ignored. Tag ordering
within records does not matter.
## Lists
A list ([) imposes ordering on a sequence of values, closed with ]. Lists
start with the byte length of their entire encoded content.
```
Format: [[size]:[value1][value2]...]
Example: [13:t3:foo,i:-42,]
Empty list: [0:]
Nested list: [23:[7:t3:foo,]t5:hello,]
```
# PARSING SECURITY CONSIDERATIONS
The length field in netencode is a decimal number that is not length-restricted.
Attackers could provide extremely long length values to overflow parsers or
cause excessive memory allocation.
*Recommended protection*:
1. Define a maximum value length in bytes for your use case
2. Count the digits in that maximum (e.g., 1024 bytes = 4 digits)
3. Restrict length field parsing to that many decimal digits
4. This implicitly limits the maximum value size
*Example*: If your maximum value is 1024 bytes, limit length fields to 4
digits. This prevents parsing anything longer than 9999 bytes.
*Implementation*: Check length field size before parsing the actual value.
Reject any length field exceeding your digit limit.
# EXAMPLES
## Simple Data Types
```
# Natural number
n:42,
# Negative integer
i:-123,
# Text with embedded quotes
t26:She said "Hello, world!",
# Binary data (3 bytes)
b3:abc,
```
## Structured Data
```
# User record
{49:<6:active|<4:true|u,<3:age|i:30,<4:name|t5:Alice,}
# List of names
[23:t5:Alice,t3:Bob,t7:Charlie,]
# Optional value (Some)
<4:Some|t5:value,
# Optional value (None)
<4:None|u,
```
## Nested Structures
```
# Record containing a list
{25:<5:items|[13:t3:foo,t3:bar,],}
# List of records
[42:{12:<4:name|t3:foo,}{12:<4:name|t3:bar,}]
```
# SEE ALSO
*netencode-pretty*(1), *netencode-record-get*(1), *netencode-to-env*(1),
*env-to-netencode*(1), *netencode-mustache*(1), *json-to-netencode*(1)
Project homepage: https://github.com/Profpatsch/netencode
|