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
|
//! Heading (setext) is a construct that occurs in the [flow] content type.
//!
//! They’re formed with the following BNF:
//!
//! ```bnf
//! heading_setext ::= line *(eol line) eol whitespace_optional (1*'-' | 1*'=') whitespace_optional
//!
//! whitespace ::= 1*space_or_tab
//! whitespace_optional ::= [ whitespace ]
//! line ::= code - eol
//! eol ::= '\r' | '\r\n' | '\n'
//! ```
//!
//! Heading (setext) in markdown relates to the `<h1>` and `<h2>` elements in
//! HTML.
//! See [*§ 4.3.6 The `h1`, `h2`, `h3`, `h4`, `h5`, and `h6` elements* in the
//! HTML spec][html] for more info.
//!
//! In markdown, it is also possible to create headings with a
//! [heading (atx)][heading_atx] construct.
//! The benefit of setext headings is that their text can include line endings,
//! and by extensions also hard breaks (e.g., with
//! [hard break (escape)][hard_break_escape]).
//! However, their limit is that they cannot form `<h3>` through `<h6>`
//! headings.
//! Due to this limitation, it is recommended to use atx headings.
//!
//! [Thematic breaks][thematic_break] formed with dashes and without whitespace
//! could be interpreted as a heading (setext).
//! Which one forms depends on whether there is text directly in fron of the
//! sequence.
//!
//! > 🏛 **Background**: the word *setext* originates from a small markup
//! > language by Ian Feldman from 1991.
//! > See [*§ Setext* on Wikipedia][wiki-setext] for more info.
//! > The word *atx* originates from a tiny markup language by Aaron Swartz
//! > from 2002.
//! > See [*§ atx, the true structured text format* on `aaronsw.com`][atx] for
//! > more info.
//!
//! ## Tokens
//!
//! * [`HeadingSetext`][Token::HeadingSetext]
//! * [`HeadingSetextText`][Token::HeadingSetextText]
//! * [`HeadingSetextUnderline`][Token::HeadingSetextUnderline]
//!
//! ## References
//!
//! * [`setext-underline.js` in `micromark`](https://github.com/micromark/micromark/blob/main/packages/micromark-core-commonmark/dev/lib/setext-underline.js)
//! * [*§ 4.3 Setext headings* in `CommonMark`](https://spec.commonmark.org/0.30/#setext-headings)
//!
//! [flow]: crate::content::flow
//! [heading_atx]: crate::construct::heading_atx
//! [thematic_break]: crate::construct::thematic_break
//! [hard_break_escape]: crate::construct::hard_break_escape
//! [html]: https://html.spec.whatwg.org/multipage/sections.html#the-h1,-h2,-h3,-h4,-h5,-and-h6-elements
//! [wiki-setext]: https://en.wikipedia.org/wiki/Setext
//! [atx]: http://www.aaronsw.com/2002/atx/
use crate::constant::TAB_SIZE;
use crate::construct::partial_space_or_tab::{space_or_tab, space_or_tab_min_max};
use crate::token::Token;
use crate::tokenizer::{Code, Event, EventType, State, StateFnResult, Tokenizer};
use crate::util::{edit_map::EditMap, skip::opt_back as skip_opt_back};
/// Kind of underline.
#[derive(Debug, Clone, PartialEq)]
pub enum Kind {
/// Dash (rank 2) heading.
///
/// ## Example
///
/// ```markdown
/// alpha
/// -----
/// ```
Dash,
/// Equals to (rank 1) heading.
///
/// ## Example
///
/// ```markdown
/// alpha
/// =====
/// ```
EqualsTo,
}
impl Kind {
/// Turn the kind into a [char].
fn as_char(&self) -> char {
match self {
Kind::Dash => '-',
Kind::EqualsTo => '=',
}
}
/// Turn a [char] into a kind.
///
/// ## Panics
///
/// Panics if `char` is not `-` or `=`.
fn from_char(char: char) -> Kind {
match char {
'-' => Kind::Dash,
'=' => Kind::EqualsTo,
_ => unreachable!("invalid char"),
}
}
}
/// At a line ending, presumably an underline.
///
/// ```markdown
/// | aa
/// > | ==
/// ^
/// ```
pub fn start(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
let max = if tokenizer.parse_state.constructs.code_indented {
TAB_SIZE - 1
} else {
usize::MAX
};
let paragraph_before = !tokenizer.events.is_empty()
&& tokenizer.events[skip_opt_back(
&tokenizer.events,
tokenizer.events.len() - 1,
&[Token::LineEnding, Token::SpaceOrTab],
)]
.token_type
== Token::Paragraph;
// Require a paragraph before and do not allow on a lazy line.
if paragraph_before && !tokenizer.lazy && tokenizer.parse_state.constructs.heading_setext {
tokenizer.go(space_or_tab_min_max(0, max), before)(tokenizer, code)
} else {
(State::Nok, None)
}
}
/// After optional whitespace, presumably an underline.
///
/// ```markdown
/// | aa
/// > | ==
/// ^
/// ```
fn before(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
match code {
Code::Char(char) if char == '-' || char == '=' => {
tokenizer.enter(Token::HeadingSetextUnderline);
inside(tokenizer, code, Kind::from_char(char))
}
_ => (State::Nok, None),
}
}
/// In an underline sequence.
///
/// ```markdown
/// | aa
/// > | ==
/// ^
/// ```
fn inside(tokenizer: &mut Tokenizer, code: Code, kind: Kind) -> StateFnResult {
match code {
Code::Char(char) if char == kind.as_char() => {
tokenizer.consume(code);
(State::Fn(Box::new(move |t, c| inside(t, c, kind))), None)
}
_ => {
tokenizer.exit(Token::HeadingSetextUnderline);
tokenizer.attempt_opt(space_or_tab(), after)(tokenizer, code)
}
}
}
/// After an underline sequence, after optional whitespace.
///
/// ```markdown
/// | aa
/// > | ==
/// ^
/// ```
fn after(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
match code {
Code::None | Code::CarriageReturnLineFeed | Code::Char('\n' | '\r') => {
// Feel free to interrupt.
tokenizer.interrupt = false;
tokenizer.register_resolver("heading_setext".to_string(), Box::new(resolve));
(State::Ok, Some(vec![code]))
}
_ => (State::Nok, None),
}
}
/// Resolve heading (setext).
pub fn resolve(tokenizer: &mut Tokenizer) -> Vec<Event> {
let mut edit_map = EditMap::new();
let mut index = 0;
let mut paragraph_enter: Option<usize> = None;
let mut paragraph_exit: Option<usize> = None;
while index < tokenizer.events.len() {
let event = &tokenizer.events[index];
// Find paragraphs.
if event.event_type == EventType::Enter {
if event.token_type == Token::Paragraph {
paragraph_enter = Some(index);
}
} else if event.token_type == Token::Paragraph {
paragraph_exit = Some(index);
}
// We know this is preceded by a paragraph.
// Otherwise we don’t parse.
else if event.token_type == Token::HeadingSetextUnderline {
let enter = paragraph_enter.take().unwrap();
let exit = paragraph_exit.take().unwrap();
// Change types of Enter:Paragraph, Exit:Paragraph.
tokenizer.events[enter].token_type = Token::HeadingSetextText;
tokenizer.events[exit].token_type = Token::HeadingSetextText;
// Add Enter:HeadingSetext, Exit:HeadingSetext.
let mut heading_enter = tokenizer.events[enter].clone();
heading_enter.token_type = Token::HeadingSetext;
let mut heading_exit = tokenizer.events[index].clone();
heading_exit.token_type = Token::HeadingSetext;
edit_map.add(enter, 0, vec![heading_enter]);
edit_map.add(index + 1, 0, vec![heading_exit]);
}
index += 1;
}
edit_map.consume(tokenizer.events.split_off(0))
}
|
