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
#![no_std]
#![cfg_attr(docsrs, feature(doc_cfg))]
#![warn(missing_copy_implementations)]
#![warn(missing_debug_implementations)]
#![warn(missing_docs)]
#![warn(trivial_casts)]
#![warn(trivial_numeric_casts)]
#![warn(unused_extern_crates)]
#![warn(unused_import_braces)]
#![warn(variant_size_differences)]
#![doc(test(attr(deny(warnings))))]

//! Another [Base58][] codec implementation.
//!
//! Compared to [`base58`][] this is significantly faster at decoding (about
//! 2.4x as fast when decoding 32 bytes), almost the same speed for encoding
//! (about 3% slower when encoding 32 bytes) and doesn't have the 128 byte
//! limitation.
//!
//! Compared to [`rust-base58`][] this is massively faster (over ten times as
//! fast when decoding 32 bytes, almost 40 times as fast when encoding 32
//! bytes) and has no external dependencies.
//!
//! Compared to both this supports a configurable alphabet and user provided
//! buffers for zero-allocation {en,de}coding.
//!
//! [Base58]: https://en.wikipedia.org/wiki/Base58
//! [`base58`]: https://github.com/debris/base58
//! [`rust-base58`]: https://github.com/nham/rust-base58
//!
//! # Features
//!
//!  Feature | Activation         | Effect
//! ---------|--------------------|--------
//!  `std`   | **on**-by-default  | Implement [`Error`](std::error::Error) for error types
//!  `alloc` | implied by `std`   | Support encoding/decoding to [`Vec`](alloc::vec::Vec) and [`String`](alloc::string::String) as appropriate
//!  `check` | **off**-by-default | Integrated support for [Base58Check][]
//!
//! [Base58Check]: https://en.bitcoin.it/wiki/Base58Check_encoding
//!
//! # Examples
//!
//! ## Basic example
//!
//! ```rust
//! let decoded = bs58::decode("he11owor1d").into_vec()?;
//! let encoded = bs58::encode(decoded).into_string();
//! assert_eq!("he11owor1d", encoded);
//! # Ok::<(), bs58::decode::Error>(())
//! ```
//!
//! ## Changing the alphabet
//!
//! ```rust
//! let decoded = bs58::decode("he11owor1d")
//!     .with_alphabet(bs58::Alphabet::RIPPLE)
//!     .into_vec()?;
//! let encoded = bs58::encode(decoded)
//!     .with_alphabet(bs58::Alphabet::FLICKR)
//!     .into_string();
//! assert_eq!("4DSSNaN1SC", encoded);
//! # Ok::<(), bs58::decode::Error>(())
//! ```
//!
//! ## Decoding into an existing buffer
//!
//! ```rust
//! let (mut decoded, mut encoded) = ([0xFF; 8], String::with_capacity(10));
//! bs58::decode("he11owor1d").into(&mut decoded)?;
//! bs58::encode(decoded).into(&mut encoded)?;
//! assert_eq!("he11owor1d", encoded);
//! # Ok::<(), Box<dyn std::error::Error>>(())
//! ```

#[cfg(feature = "std")]
extern crate std;

#[cfg(feature = "alloc")]
extern crate alloc;

pub mod alphabet;
#[doc(inline)]
pub use alphabet::Alphabet;

pub mod decode;
pub mod encode;

#[cfg(feature = "check")]
const CHECKSUM_LEN: usize = 4;

/// Possible check variants.
enum Check {
    Disabled,
    #[cfg(feature = "check")]
    Enabled(Option<u8>),
}

/// Setup decoder for the given string using the [default alphabet][Alphabet::DEFAULT].
///
/// # Examples
///
/// ## Basic example
///
/// ```rust
/// assert_eq!(
///     vec![0x04, 0x30, 0x5e, 0x2b, 0x24, 0x73, 0xf0, 0x58],
///     bs58::decode("he11owor1d").into_vec()?);
/// # Ok::<(), bs58::decode::Error>(())
/// ```
///
/// ## Changing the alphabet
///
/// ```rust
/// assert_eq!(
///     vec![0x60, 0x65, 0xe7, 0x9b, 0xba, 0x2f, 0x78],
///     bs58::decode("he11owor1d")
///         .with_alphabet(bs58::Alphabet::RIPPLE)
///         .into_vec()?);
/// # Ok::<(), bs58::decode::Error>(())
/// ```
///
/// ## Decoding into an existing buffer
///
/// ```rust
/// let mut output = [0xFF; 10];
/// assert_eq!(8, bs58::decode("he11owor1d").into(&mut output)?);
/// assert_eq!(
///     [0x04, 0x30, 0x5e, 0x2b, 0x24, 0x73, 0xf0, 0x58, 0xFF, 0xFF],
///     output);
/// # Ok::<(), bs58::decode::Error>(())
/// ```
///
/// ## Errors
///
/// ### Invalid Character
///
/// ```rust
/// assert_eq!(
///     bs58::decode::Error::InvalidCharacter { character: 'l', index: 2 },
///     bs58::decode("hello world").into_vec().unwrap_err());
/// ```
///
/// ### Non-ASCII Character
///
/// ```rust
/// assert_eq!(
///     bs58::decode::Error::NonAsciiCharacter { index: 5 },
///     bs58::decode("he11o🇳🇿").into_vec().unwrap_err());
/// ```
///
/// ### Too Small Buffer
///
/// This error can only occur when reading into a provided buffer, when using
/// [`into_vec()`][decode::DecodeBuilder::into_vec] a vector large enough is guaranteed to be
/// used.
///
/// ```rust
/// let mut output = [0; 7];
/// assert_eq!(
///     bs58::decode::Error::BufferTooSmall,
///     bs58::decode("he11owor1d").into(&mut output).unwrap_err());
/// ```
pub fn decode<I: AsRef<[u8]>>(input: I) -> decode::DecodeBuilder<'static, I> {
    decode::DecodeBuilder::from_input(input)
}

/// Setup encoder for the given bytes using the [default alphabet][Alphabet::DEFAULT].
///
/// # Examples
///
/// ## Basic example
///
/// ```rust
/// let input = [0x04, 0x30, 0x5e, 0x2b, 0x24, 0x73, 0xf0, 0x58];
/// assert_eq!("he11owor1d", bs58::encode(input).into_string());
/// ```
///
/// ## Changing the alphabet
///
/// ```rust
/// let input = [0x60, 0x65, 0xe7, 0x9b, 0xba, 0x2f, 0x78];
/// assert_eq!(
///     "he11owor1d",
///     bs58::encode(input)
///         .with_alphabet(bs58::Alphabet::RIPPLE)
///         .into_string());
/// ```
///
/// ## Encoding into an existing string
///
/// ```rust
/// let input = [0x04, 0x30, 0x5e, 0x2b, 0x24, 0x73, 0xf0, 0x58];
/// let mut output = "goodbye world".to_owned();
/// bs58::encode(input).into(&mut output)?;
/// assert_eq!("he11owor1d", output);
/// # Ok::<(), bs58::encode::Error>(())
/// ```
///
/// ## Errors
///
/// ### Too Small Buffer
///
/// This error can only occur when reading into an unresizeable buffer.
///
/// ```rust
/// let input = [0x04, 0x30, 0x5e, 0x2b, 0x24, 0x73, 0xf0, 0x58];
/// let mut output = [0; 7];
/// assert_eq!(
///     bs58::encode::Error::BufferTooSmall,
///     bs58::encode(input).into(&mut output[..]).unwrap_err());
/// ```
pub fn encode<I: AsRef<[u8]>>(input: I) -> encode::EncodeBuilder<'static, I> {
    encode::EncodeBuilder::from_input(input)
}