parser/
input.rs

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
//! See [`Input`].

use crate::SyntaxKind;

#[allow(non_camel_case_types)]
type bits = u64;

/// Input for the parser -- a sequence of tokens.
///
/// As of now, parser doesn't have access to the *text* of the tokens, and makes
/// decisions based solely on their classification. Unlike `LexerToken`, the
/// `Tokens` doesn't include whitespace and comments. Main input to the parser.
///
/// Struct of arrays internally, but this shouldn't really matter.
#[derive(Default)]
pub struct Input {
    kind: Vec<SyntaxKind>,
    joint: Vec<bits>,
    contextual_kind: Vec<SyntaxKind>,
}

/// `pub` impl used by callers to create `Tokens`.
impl Input {
    #[inline]
    pub fn push(&mut self, kind: SyntaxKind) {
        self.push_impl(kind, SyntaxKind::EOF)
    }
    #[inline]
    pub fn push_ident(&mut self, contextual_kind: SyntaxKind) {
        self.push_impl(SyntaxKind::IDENT, contextual_kind)
    }
    /// Sets jointness for the last token we've pushed.
    ///
    /// This is a separate API rather than an argument to the `push` to make it
    /// convenient both for textual and mbe tokens. With text, you know whether
    /// the *previous* token was joint, with mbe, you know whether the *current*
    /// one is joint. This API allows for styles of usage:
    ///
    /// ```
    /// // In text:
    /// tokens.was_joint(prev_joint);
    /// tokens.push(curr);
    ///
    /// // In MBE:
    /// token.push(curr);
    /// tokens.push(curr_joint)
    /// ```
    #[inline]
    pub fn was_joint(&mut self) {
        let n = self.len() - 1;
        let (idx, b_idx) = self.bit_index(n);
        self.joint[idx] |= 1 << b_idx;
    }
    #[inline]
    fn push_impl(&mut self, kind: SyntaxKind, contextual_kind: SyntaxKind) {
        let idx = self.len();
        if idx % (bits::BITS as usize) == 0 {
            self.joint.push(0);
        }
        self.kind.push(kind);
        self.contextual_kind.push(contextual_kind);
    }
}

/// pub(crate) impl used by the parser to consume `Tokens`.
impl Input {
    pub(crate) fn kind(&self, idx: usize) -> SyntaxKind {
        self.kind.get(idx).copied().unwrap_or(SyntaxKind::EOF)
    }
    pub(crate) fn contextual_kind(&self, idx: usize) -> SyntaxKind {
        self.contextual_kind.get(idx).copied().unwrap_or(SyntaxKind::EOF)
    }
    pub(crate) fn is_joint(&self, n: usize) -> bool {
        let (idx, b_idx) = self.bit_index(n);
        self.joint[idx] & 1 << b_idx != 0
    }
}

impl Input {
    fn bit_index(&self, n: usize) -> (usize, usize) {
        let idx = n / (bits::BITS as usize);
        let b_idx = n % (bits::BITS as usize);
        (idx, b_idx)
    }
    fn len(&self) -> usize {
        self.kind.len()
    }
}