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

use std::ops::RangeInclusive;

use nom8::branch::alt;
use nom8::bytes::one_of;
use nom8::bytes::take_while;
use nom8::bytes::take_while1;
use nom8::combinator::eof;
use nom8::combinator::opt;
use nom8::multi::many0_count;
use nom8::multi::many1_count;
use nom8::prelude::*;
use nom8::sequence::terminated;

use crate::parser::prelude::*;

pub(crate) unsafe fn from_utf8_unchecked<'b>(
    bytes: &'b [u8],
    safety_justification: &'static str,
) -> &'b str {
    if cfg!(debug_assertions) {
        // Catch problems more quickly when testing
        std::str::from_utf8(bytes).expect(safety_justification)
    } else {
        std::str::from_utf8_unchecked(bytes)
    }
}

// wschar = ( %x20 /              ; Space
//            %x09 )              ; Horizontal tab
pub(crate) const WSCHAR: (u8, u8) = (b' ', b'\t');

// ws = *wschar
pub(crate) fn ws(input: Input<'_>) -> IResult<Input<'_>, &str, ParserError<'_>> {
    take_while(WSCHAR)
        .map(|b| unsafe { from_utf8_unchecked(b, "`is_wschar` filters out on-ASCII") })
        .parse(input)
}

// non-ascii = %x80-D7FF / %xE000-10FFFF
// - ASCII is 0xxxxxxx
// - First byte for UTF-8 is 11xxxxxx
// - Subsequent UTF-8 bytes are 10xxxxxx
pub(crate) const NON_ASCII: RangeInclusive<u8> = 0x80..=0xff;

// non-eol = %x09 / %x20-7E / non-ascii
pub(crate) const NON_EOL: (u8, RangeInclusive<u8>, RangeInclusive<u8>) =
    (0x09, 0x20..=0x7E, NON_ASCII);

// comment-start-symbol = %x23 ; #
pub(crate) const COMMENT_START_SYMBOL: u8 = b'#';

// comment = comment-start-symbol *non-eol
pub(crate) fn comment(input: Input<'_>) -> IResult<Input<'_>, &[u8], ParserError<'_>> {
    (COMMENT_START_SYMBOL, take_while(NON_EOL))
        .recognize()
        .parse(input)
}

// newline = ( %x0A /              ; LF
//             %x0D.0A )           ; CRLF
pub(crate) fn newline(input: Input<'_>) -> IResult<Input<'_>, u8, ParserError<'_>> {
    alt((
        one_of(LF).value(b'\n'),
        (one_of(CR), one_of(LF)).value(b'\n'),
    ))
    .parse(input)
}
pub(crate) const LF: u8 = b'\n';
pub(crate) const CR: u8 = b'\r';

// ws-newline       = *( wschar / newline )
pub(crate) fn ws_newline(input: Input<'_>) -> IResult<Input<'_>, &str, ParserError<'_>> {
    many0_count(alt((newline.value(&b"\n"[..]), take_while1(WSCHAR))))
        .recognize()
        .map(|b| unsafe {
            from_utf8_unchecked(b, "`is_wschar` and `newline` filters out on-ASCII")
        })
        .parse(input)
}

// ws-newlines      = newline *( wschar / newline )
pub(crate) fn ws_newlines(input: Input<'_>) -> IResult<Input<'_>, &str, ParserError<'_>> {
    (newline, ws_newline)
        .recognize()
        .map(|b| unsafe {
            from_utf8_unchecked(b, "`is_wschar` and `newline` filters out on-ASCII")
        })
        .parse(input)
}

// note: this rule is not present in the original grammar
// ws-comment-newline = *( ws-newline-nonempty / comment )
pub(crate) fn ws_comment_newline(input: Input<'_>) -> IResult<Input<'_>, &[u8], ParserError<'_>> {
    many0_count(alt((
        many1_count(alt((take_while1(WSCHAR), newline.value(&b"\n"[..])))).value(()),
        comment.value(()),
    )))
    .recognize()
    .parse(input)
}

// note: this rule is not present in the original grammar
// line-ending = newline / eof
pub(crate) fn line_ending(input: Input<'_>) -> IResult<Input<'_>, &str, ParserError<'_>> {
    alt((newline.value("\n"), eof.value(""))).parse(input)
}

// note: this rule is not present in the original grammar
// line-trailing = ws [comment] skip-line-ending
pub(crate) fn line_trailing(
    input: Input<'_>,
) -> IResult<Input<'_>, std::ops::Range<usize>, ParserError<'_>> {
    terminated((ws, opt(comment)).span(), line_ending).parse(input)
}

#[cfg(test)]
mod test {
    use super::*;

    #[test]
    fn trivia() {
        let inputs = [
            "",
            r#" "#,
            r#"
"#,
            r#"
# comment

# comment2


"#,
            r#"
        "#,
            r#"# comment
# comment2


   "#,
        ];
        for input in inputs {
            dbg!(input);
            let parsed = ws_comment_newline.parse(new_input(input)).finish();
            assert!(parsed.is_ok(), "{:?}", parsed);
            let parsed = parsed.unwrap();
            assert_eq!(parsed, input.as_bytes());
        }
    }
}