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

#![no_std]
//! # rut - A compact byte-wise UTF-8 parser for Rust.
//! 
//! This crate was originally part of [Termiku](https://github.com/ShinySaana/termiku).  
//! As such, the API is pretty bare currently, and some issues are not handled yet (like [overlong sequences](https://wikipedia.org/wiki/UTF-8#Overlong_encodings)).
//! 
//! However, it is usable, and `#[no_std]` compatible.
//! 
//! # Examples
//! 
//! ```rust
//! use rut::Utf8Parser;
//! 
//! let mut p = Utf8Parser::new();
//! 
//! // UTF-8 representation of '€'
//! let bytes = [0xE2, 0x82, 0xAC];
//! 
//! assert_eq!(p.parse_byte(bytes[0]), Ok(None));
//! assert_eq!(p.parse_byte(bytes[1]), Ok(None));
//! assert_eq!(p.parse_byte(bytes[2]), Ok(Some('€')));
//! ```

type Result = ::core::result::Result<Option<char>, Utf8ParserError>;

#[derive(Copy, Clone, Debug, Eq, PartialEq)]
pub enum Utf8ParserError {
    InvalidByte,
    InvalidContinuationByte,
    InvalidCodePoint(u32),
    UnexpectedContinuationByte,
}

#[derive(Clone, Default)]
pub struct Utf8Parser {
    // Code point parsed up to this point.
    value: u32,
    // Number of continuation bytes in this sequence.
    length: u32,
}

// Bitmask used for leading bytes.
// 110x xxxx for 2 byte,
// 1110 xxxx for 3 byte, and
// 1111 0xxx for 4 byte sequences.
const LEAD_MASK: [u8; 3] = [0x1F, 0x0F, 0x07];

// Bitmask used for continuation bytes.
// Is always 11xx xxxx.
const CONT_MASK: u8 = 0x3F;

impl Utf8Parser {
    /// Crates a new parser.
    pub fn new() -> Utf8Parser {
        Utf8Parser::default()
    }

    /// Tries to parse an UTF-8 byte.
    /// Returns `Ok(Some(char))` if a full character was parsed,
    /// `Ok(None)` if the byte was parsed but no full character yet,
    /// and `Err(Utf8ParserError)` otherwise.
    pub fn parse_byte(&mut self, byte: u8) -> Result {
        use core::convert::TryFrom;
        use Utf8ParserError::*;

        if !is_valid_utf8_byte(byte) {
            // If we get an invalid byte, reset parsing state.
            self.length = 0;
            return Err(InvalidByte);
        }

        // Start parsing a new sequence.
        if self.length == 0 {
            if byte < 0x80 {
                return Ok(Some(char::from(byte)));
            } else if is_utf8_continuation_byte(byte) {
                return Err(UnexpectedContinuationByte);
            }

            // We subtract 1 and treat it as the number of bytes following this one.
            self.length = utf8_length(byte) - 1;
            self.value = (byte & LEAD_MASK[self.length as usize]) as u32;

            // Parsing is Ok, but we don't have a full char yet.
            return Ok(None);
        } else {
            // Continue parsing the current sequence
            if !is_utf8_continuation_byte(byte) {
                // If we get an invalid continuation byte, reset parsing state.
                self.length = 0;
                return Err(InvalidContinuationByte);
            }

            self.value = (self.value << 6) | (byte & CONT_MASK) as u32;
            self.length -= 1;

            // We're done
            if self.length == 0 {
                match char::try_from(self.value) {
                    Ok(c) => return Ok(Some(c)),
                    Err(_) => return Err(InvalidCodePoint(self.value)),
                }
            } else {
                // Parsing is Ok, but we don't have a full char yet.
                return Ok(None);
            }
        }
    }

    // TODO: Implement this!
    #[allow(unused)]
    fn parse_bytes(&mut self, bytes: &[u8]) {
        unimplemented!("Utf8Parser::parse_bytes is unimplemented due to uncertainties as to what its return type should be.");
    }
}

#[inline]
fn is_valid_utf8_byte(byte: u8) -> bool {
    match byte {
        0xC0 | 0xC1 | 0xF5..=0xFF => false,
        _ => true,
    }
}

#[inline]
fn is_utf8_continuation_byte(byte: u8) -> bool {
    // Continuation bytes look like 10xx xxxx,
    // so we look at the top 2 bits and see if they match.
    (byte >> 6) == 0b10
}

#[inline]
fn utf8_length(byte: u8) -> u32 {
    // This function assumes that `byte` is both a valid UTF-8 byte and *not* a continuation byte.
    debug_assert!(is_valid_utf8_byte(byte) && !is_utf8_continuation_byte(byte));

    if byte < 0x80 {
        1
    } else {
        (!byte).leading_zeros()
    }
}