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

use super::pattern::is_valid_pattern_identifier;
use super::ParseCtx;
use super::Parser;
use crate::ast::ClassOrObjectMemberKey;
use crate::error::SyntaxError;
use crate::error::SyntaxErrorType;
use crate::error::SyntaxResult;
use crate::num::JsNumber;
use crate::session::SessionString;
use crate::source::SourceRange;
use crate::token::TokenType;
use core::str::FromStr;
use memchr::memchr;
use std::str::from_utf8_unchecked;

fn parse_radix(raw: &str, radix: u32) -> Result<f64, ()> {
  u64::from_str_radix(raw, radix)
    .map_err(|_| ())
    // TODO This is lossy, but there is no TryFrom for converting from u64 to f64, and u32 cannot represent all possible JS values.
    .map(|v| v as f64)
}

pub fn normalise_literal_number<'a>(raw: SourceRange<'a>) -> SyntaxResult<'a, JsNumber> {
  // TODO We assume that the Rust parser follows ECMAScript spec and that different representations
  // of the same value get parsed into the same f64 value/bit pattern (e.g. `5.1e10` and `0.51e11`).
  match raw.as_str() {
    s if s.starts_with("0b") || s.starts_with("0B") => parse_radix(&s[2..], 2),
    s if s.starts_with("0o") || s.starts_with("0o") => parse_radix(&s[2..], 8),
    s if s.starts_with("0x") || s.starts_with("0X") => parse_radix(&s[2..], 16),
    s => f64::from_str(s).map_err(|_| ()),
  }
  .map(|n| JsNumber(n))
  .map_err(|_| SyntaxError::from_loc(raw, SyntaxErrorType::MalformedLiteralNumber, None))
}

pub fn normalise_literal_bigint<'a>(
  ctx: ParseCtx<'a>,
  raw: SourceRange<'a>,
) -> SyntaxResult<'a, SessionString<'a>> {
  // TODO Use custom type like JsNumber.
  let mut norm = ctx.session.new_string();
  // TODO
  norm.push_str(raw.as_str());
  Ok(norm)
}

pub fn normalise_literal_string_or_template_inner<'a>(
  ctx: ParseCtx<'a>,
  mut raw: &[u8],
) -> Option<&'a str> {
  let mut norm = ctx.session.new_vec();
  while !raw.is_empty() {
    let Some(escape_pos) = memchr(b'\\', raw) else {
      norm.extend_from_slice(raw);
      break;
    };
    norm.extend_from_slice(&raw[..escape_pos]);
    raw = &raw[escape_pos + 1..];
    // https://mathiasbynens.be/notes/javascript-escapes
    // https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/String#escape_sequences
    // https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Template_literals#tagged_templates_and_escape_sequences
    let mut tmp = [0u8; 4];
    let (skip, add): (usize, &[u8]) = match raw[0] {
      b'\n' => (1, b""),
      b'b' => (1, b"\x08"),
      b'f' => (1, b"\x0c"),
      b'n' => (1, b"\n"),
      b'r' => (1, b"\r"),
      b't' => (1, b"\t"),
      b'v' => (1, b"\x0b"),
      b'0'..=b'7' => {
        // Octal escape.
        let mut len = 1;
        if raw.get(len).filter(|&&c| c >= b'0' && c <= b'7').is_some() {
          len += 1;
          if raw.get(len).filter(|&&c| c >= b'0' && c <= b'7').is_some() {
            len += 1;
          };
        };
        char::from_u32(
          u32::from_str_radix(unsafe { from_utf8_unchecked(&raw[..len]) }, 8).unwrap(),
        )
        .unwrap()
        .encode_utf8(&mut tmp);
        (len, tmp.as_slice())
      }
      b'x' => {
        // Hexadecimal escape.
        if raw.len() < 3 || !raw[1].is_ascii_hexdigit() || !raw[2].is_ascii_hexdigit() {
          return None;
        };
        char::from_u32(
          u32::from_str_radix(unsafe { from_utf8_unchecked(&raw[1..3]) }, 16).unwrap(),
        )
        .unwrap()
        .encode_utf8(&mut tmp);
        (3, tmp.as_slice())
      }
      b'u' => match raw.get(1) {
        Some(b'{') => {
          // Unicode code point escape.
          let Some(end_pos) = memchr(b'}', raw) else {
            return None;
          };
          if end_pos < 3 || end_pos > 8 {
            return None;
          };
          let cp =
            u32::from_str_radix(unsafe { from_utf8_unchecked(&raw[2..end_pos]) }, 16).ok()?;
          let c = char::from_u32(cp)?;
          c.encode_utf8(&mut tmp);
          (end_pos + 1, tmp.as_slice())
        }
        Some(_) => {
          // Unicode escape.
          if raw.len() < 5 {
            return None;
          };
          let cp = u32::from_str_radix(unsafe { from_utf8_unchecked(&raw[1..5]) }, 16).ok()?;
          let c = char::from_u32(cp)?;
          c.encode_utf8(&mut tmp);
          (5, tmp.as_slice())
        }
        None => {
          return None;
        }
      },
      c => (1, {
        tmp[0] = c;
        &tmp[..1]
      }),
    };
    norm.extend_from_slice(add);
    raw = &raw[skip..];
  }
  // Copy onto arena so we get immutable references that are cheap to copy.
  // We return str instead of [u8] so that serialisation is easy and str methods are available.
  Some(unsafe { from_utf8_unchecked(ctx.session.get_allocator().alloc_slice_copy(&norm)) })
}

pub fn normalise_literal_string<'a>(
  ctx: ParseCtx<'a>,
  raw: SourceRange<'a>,
) -> SyntaxResult<'a, &'a str> {
  normalise_literal_string_or_template_inner(ctx, &raw.as_slice()[1..raw.len() - 1])
    .ok_or_else(|| raw.error(SyntaxErrorType::InvalidCharacterEscape, None))
}

impl<'a> Parser<'a> {
  pub fn parse_and_normalise_literal_string(
    &mut self,
    ctx: ParseCtx<'a>,
  ) -> SyntaxResult<'a, &'a str> {
    let t = self.require(TokenType::LiteralString)?;
    let s = normalise_literal_string(ctx, t.loc)?;
    Ok(s)
  }
}