asciidoc_parser/inlines/inline.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 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
use crate::{inlines::InlineMacro, span::MatchedItem, HasSpan, Span};
/// An inline element is a phrase (i.e., span of content) within a block element
/// or one of its attributes in an AsciiDoc document.
#[derive(Clone, Debug, Eq, PartialEq)]
#[non_exhaustive]
pub enum Inline<'src> {
/// Uninterpreted text (i.e., plain text) is text (character data) for which
/// all inline grammar rules fail to match.
Uninterpreted(Span<'src>),
/// A sequence of other inline blocks.
Sequence(Vec<Self>, Span<'src>),
/// An inline macro.
Macro(InlineMacro<'src>),
}
impl<'src> Inline<'src> {
/// Parse a span (typically a line) of any type and return an `Inline` that
/// describes it.
///
/// Returns `None` if input doesn't start with a non-empty line.
pub(crate) fn parse(source: Span<'src>) -> Option<MatchedItem<'src, Self>> {
let line = source.take_non_empty_line()?;
let mut span = line.item;
// Special-case optimization: If the entire span is one
// uninterpreted block, just return that without the allocation
// overhead of the Vec of inlines.
let mut uninterp = parse_uninterpreted(span);
if uninterp.after.is_empty() {
return Some(MatchedItem {
item: Self::Uninterpreted(uninterp.item),
after: line.after,
});
}
let mut inlines: Vec<Self> = vec![];
loop {
if !uninterp.item.is_empty() {
inlines.push(Self::Uninterpreted(uninterp.item));
}
span = uninterp.after;
if span.is_empty() {
break;
}
let interp = parse_interpreted(span)?;
if interp.after.is_empty() && inlines.is_empty() {
return Some(interp);
}
inlines.push(interp.item);
span = interp.after;
uninterp = parse_uninterpreted(span);
}
Some(MatchedItem {
item: Self::Sequence(inlines, source.trim_remainder(line.after)),
after: line.after,
})
}
/// Parse a sequence of non-empty lines as a single `Inline` that
/// describes it.
///
/// Returns `None` if there is not at least one non-empty line at
/// beginning of input.
pub(crate) fn parse_lines(source: Span<'src>) -> Option<MatchedItem<'src, Self>> {
let mut inlines: Vec<Inline<'src>> = vec![];
let mut next = source;
while let Some(inline) = Self::parse(next) {
next = inline.after;
inlines.push(inline.item);
}
if inlines.len() < 2 {
inlines.pop().map(|inline| MatchedItem {
item: inline,
after: next,
})
} else {
let source = source.trim_remainder(next);
Some(MatchedItem {
item: Self::Sequence(inlines, source),
after: next,
})
}
}
}
impl<'src> HasSpan<'src> for Inline<'src> {
fn span(&'src self) -> &'src Span<'src> {
match self {
Self::Uninterpreted(source) => source,
Self::Sequence(_, source) => source,
Self::Macro(m) => m.span(),
}
}
}
// Parse the largest possible block of "uninterpreted" text.
// Remainder is either empty span or first span that requires
// special interpretation.
fn parse_uninterpreted(source: Span<'_>) -> MatchedItem<Span> {
// Optimization: If line doesn't contain special markup chars,
// then it's all uninterpreted.
if !source.contains(':') {
return source.into_parse_result(source.len());
}
let mut after = source;
while !after.is_empty() {
if InlineMacro::parse(after).is_some() {
break;
}
let word = after.take_while(|c| c != ' ' && c != '\t');
let spaces = word.after.take_whitespace();
after = spaces.after;
}
MatchedItem {
item: source.trim_remainder(after),
after,
}
}
// Parse the block as a special "interpreted" inline sequence or error out.
fn parse_interpreted(source: Span<'_>) -> Option<MatchedItem<Inline<'_>>> {
InlineMacro::parse(source).map(|inline| MatchedItem {
item: Inline::Macro(inline.item),
after: inline.after,
})
}