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
//! This module contains a bunch of traits necessary for processing byte strings.
//!
//! Most notable are:
//! * `Source` - implemented by default for `&str` and `&[u8]`, used by the `Lexer`.
//! * `Slice` - slices of `Source`, returned by `Lexer::slice`.

use std::fmt::Debug;
use std::ops::Range;

/// Trait for types the `Lexer` can read from.
///
/// Most notably this is implemented for `&str`. It is unlikely you will
/// ever want to use this Trait yourself, unless implementing a new `Source`
/// the `Lexer` can use.
pub trait Source {
    /// A type this `Source` can be sliced into.
    type Slice: ?Sized + PartialEq + Eq + Debug;

    /// Length of the source
    fn len(&self) -> usize;

    /// Read a chunk of bytes into an array. Returns `None` when reading
    /// out of bounds would occur.
    ///
    /// This is very useful for matching fixed-size byte arrays, and tends
    /// to be very fast at it too, since the compiler knows the byte lengths.
    ///
    /// ```rust
    /// use logos::Source;
    ///
    /// fn main() {
    ///     let foo = "foo";
    ///
    ///     assert_eq!(foo.read(0), Some(b"foo"));     // Option<&[u8; 3]>
    ///     assert_eq!(foo.read(0), Some(b"fo"));      // Option<&[u8; 2]>
    ///     assert_eq!(foo.read(2), Some(b'o'));       // Option<u8>
    ///     assert_eq!(foo.read::<&[u8; 4]>(0), None); // Out of bounds
    ///     assert_eq!(foo.read::<&[u8; 2]>(2), None); // Out of bounds
    /// }
    /// ```
    fn read<'a, Chunk>(&'a self, offset: usize) -> Option<Chunk>
    where
        Chunk: self::Chunk<'a>;

    /// Read a chunk of bytes into an array without doing bounds checks.
    unsafe fn read_unchecked<'a, Chunk>(&'a self, offset: usize) -> Chunk
    where
        Chunk: self::Chunk<'a>;

    /// Get a slice of the source at given range. This is analogous to
    /// `slice::get(range)`.
    ///
    /// ```rust
    /// use logos::Source;
    ///
    /// fn main() {
    ///     let foo = "It was the year when they finally immanentized the Eschaton.";
    ///
    ///     assert_eq!(<str as Source>::slice(&foo, 51..59), Some("Eschaton"));
    /// }
    /// ```
    fn slice(&self, range: Range<usize>) -> Option<&Self::Slice>;

    /// Get a slice of the source at given range. This is analogous to
    /// `slice::get_unchecked(range)`.
    ///
    /// **Using this method with range out of bounds is undefined behavior!**
    ///
    /// ```rust
    /// use logos::Source;
    ///
    /// fn main() {
    ///     let foo = "It was the year when they finally immanentized the Eschaton.";
    ///
    ///     unsafe {
    ///         assert_eq!(<str as Source>::slice_unchecked(&foo, 51..59), "Eschaton");
    ///     }
    /// }
    /// ```
    unsafe fn slice_unchecked(&self, range: Range<usize>) -> &Self::Slice;

    /// For `&str` sources attempts to find the closest `char` boundary at which source
    /// can be sliced, starting from `index`.
    ///
    /// For binary sources (`&[u8]`) this should just return `index` back.
    #[inline]
    fn find_boundary(&self, index: usize) -> usize {
        index
    }

    /// Check if `index` is valid for this `Source`, that is:
    ///
    /// + It's not larger than the byte length of the `Source`.
    /// + (`str` only) It doesn't land in the middle of a UTF-8 code point.
    fn is_boundary(&self, index: usize) -> bool;
}

impl Source for str {
    type Slice = str;

    #[inline]
    fn len(&self) -> usize {
        self.len()
    }

    #[inline]
    fn read<'a, Chunk>(&'a self, offset: usize) -> Option<Chunk>
    where
        Chunk: self::Chunk<'a>,
    {
        if offset + (Chunk::SIZE - 1) < self.len() {
            Some(unsafe { Chunk::from_ptr(self.as_ptr().add(offset)) })
        } else {
            None
        }
    }

    #[inline]
    unsafe fn read_unchecked<'a, Chunk>(&'a self, offset: usize) -> Chunk
    where
        Chunk: self::Chunk<'a>,
    {
        Chunk::from_ptr(self.as_ptr().add(offset))
    }

    #[inline]
    fn slice(&self, range: Range<usize>) -> Option<&str> {
        self.get(range)
    }

    #[inline]
    unsafe fn slice_unchecked(&self, range: Range<usize>) -> &str {
        debug_assert!(
            range.start <= self.len() && range.end <= self.len(),
            "Reading out of bounds {:?} for {}!",
            range,
            self.len()
        );

        self.get_unchecked(range)
    }

    #[inline]
    fn find_boundary(&self, mut index: usize) -> usize {
        while !self.is_char_boundary(index) {
            index += 1;
        }

        index
    }

    #[inline]
    fn is_boundary(&self, index: usize) -> bool {
        self.is_char_boundary(index)
    }
}

impl Source for [u8] {
    type Slice = [u8];

    #[inline]
    fn len(&self) -> usize {
        self.len()
    }

    #[inline]
    fn read<'a, Chunk>(&'a self, offset: usize) -> Option<Chunk>
    where
        Chunk: self::Chunk<'a>,
    {
        if offset + (Chunk::SIZE - 1) < self.len() {
            Some(unsafe { Chunk::from_ptr(self.as_ptr().add(offset)) })
        } else {
            None
        }
    }

    #[inline]
    unsafe fn read_unchecked<'a, Chunk>(&'a self, offset: usize) -> Chunk
    where
        Chunk: self::Chunk<'a>,
    {
        Chunk::from_ptr(self.as_ptr().add(offset))
    }

    #[inline]
    fn slice(&self, range: Range<usize>) -> Option<&[u8]> {
        self.get(range)
    }

    #[inline]
    unsafe fn slice_unchecked(&self, range: Range<usize>) -> &[u8] {
        debug_assert!(
            range.start <= self.len() && range.end <= self.len(),
            "Reading out of bounds {:?} for {}!",
            range,
            self.len()
        );

        self.get_unchecked(range)
    }

    #[inline]
    fn is_boundary(&self, index: usize) -> bool {
        index <= self.len()
    }
}

/// A fixed, statically sized chunk of data that can be read from the `Source`.
///
/// This is implemented for `u8`, as well as byte arrays `&[u8; 1]` to `&[u8; 16]`.
pub trait Chunk<'source>: Sized + Copy + PartialEq + Eq {
    /// Size of the chunk being accessed in bytes.
    const SIZE: usize;

    /// Create a chunk from a raw byte pointer.
    unsafe fn from_ptr(ptr: *const u8) -> Self;
}

impl<'source> Chunk<'source> for u8 {
    const SIZE: usize = 1;

    #[inline]
    unsafe fn from_ptr(ptr: *const u8) -> Self {
        *ptr
    }
}

macro_rules! impl_array {
    ($($size:expr),*) => ($(
        impl<'source> Chunk<'source> for &'source [u8; $size] {
            const SIZE: usize = $size;

            #[inline]
            unsafe fn from_ptr(ptr: *const u8) -> Self {
                &*(ptr as *const [u8; $size])
            }
        }
    )*);
}

impl_array!(
    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
);