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
|
//! Trailing whitespace occurs in [string][] and [text][].
//!
//! It occurs around line endings, and, in the case of text content it also
//! occurs at the start or end of the whole.
//!
//! They’re formed with the following BNF:
//!
//! ```bnf
//! ; Restriction: the start and end here count as an eol in the case of `text`.
//! whitespace ::= 0.*space_or_tab eol 0.*space_or_tab
//! ```
//!
//! Normally this whitespace is ignored.
//! In the case of text content, whitespace before a line ending that
//! consistents solely of spaces, at least 2, forms a hard break (trailing).
//!
//! The minimum number of the spaces is defined in
//! [`HARD_BREAK_PREFIX_SIZE_MIN`][hard_break_prefix_size_min].
//!
//! Hard breaks in markdown relate to the HTML element `<br>`.
//! See [*§ 4.5.27 The `br` element* in the HTML spec][html] for more info.
//!
//! It is also possible to create a hard break with a similar construct: a
//! [hard break (escape)][hard_break_escape] is a backslash followed
//! by a line ending.
//! That construct is recommended because it is similar to a
//! [character escape][character_escape] and similar to how line endings can be
//! “escaped” in other languages.
//! Trailing spaces are typically invisible in editors, or even automatically
//! removed, making hard break (trailing) hard to use.
//! ## Tokens
//!
//! * [`HardBreakTrailing`][Name::HardBreakTrailing]
//! * [`SpaceOrTab`][Name::SpaceOrTab]
//!
//! ## References
//!
//! * [`initialize/text.js` in `micromark`](https://github.com/micromark/micromark/blob/main/packages/micromark/dev/lib/initialize/text.js)
//! * [*§ 6.7 Hard line breaks* in `CommonMark`](https://spec.commonmark.org/0.30/#hard-line-breaks)
//!
//! [string]: crate::construct::string
//! [text]: crate::construct::text
//! [hard_break_escape]: crate::construct::hard_break_escape
//! [character_escape]: crate::construct::character_escape
//! [hard_break_prefix_size_min]: crate::constant::HARD_BREAK_PREFIX_SIZE_MIN
//! [html]: https://html.spec.whatwg.org/multipage/text-level-semantics.html#the-br-element
use crate::constant::HARD_BREAK_PREFIX_SIZE_MIN;
use crate::event::{Event, Kind, Name};
use crate::tokenizer::Tokenizer;
use crate::util::slice::{Position, Slice};
/// Resolve whitespace.
pub fn resolve_whitespace(tokenizer: &mut Tokenizer, hard_break: bool, trim_whole: bool) {
let mut index = 0;
while index < tokenizer.events.len() {
let event = &tokenizer.events[index];
if event.kind == Kind::Exit && event.name == Name::Data {
let trim_start = (trim_whole && index == 1)
|| (index > 1 && tokenizer.events[index - 2].name == Name::LineEnding);
let trim_end = (trim_whole && index == tokenizer.events.len() - 1)
|| (index + 1 < tokenizer.events.len()
&& tokenizer.events[index + 1].name == Name::LineEnding);
trim_data(tokenizer, index, trim_start, trim_end, hard_break);
}
index += 1;
}
}
/// Trim a [`Data`][Name::Data] token.
fn trim_data(
tokenizer: &mut Tokenizer,
exit_index: usize,
trim_start: bool,
trim_end: bool,
hard_break: bool,
) {
let mut slice = Slice::from_position(
tokenizer.parse_state.bytes,
&Position::from_exit_event(&tokenizer.events, exit_index),
);
if trim_end {
let mut index = slice.bytes.len();
let mut spaces_only = slice.after == 0;
while index > 0 {
match slice.bytes[index - 1] {
b' ' => {}
b'\t' => spaces_only = false,
_ => break,
}
index -= 1;
}
let diff = slice.bytes.len() - index;
let name = if hard_break
&& spaces_only
&& diff >= HARD_BREAK_PREFIX_SIZE_MIN
&& exit_index + 1 < tokenizer.events.len()
{
Name::HardBreakTrailing
} else {
Name::SpaceOrTab
};
// The whole data is whitespace.
// We can be very fast: we only change the token types.
if index == 0 {
tokenizer.events[exit_index - 1].name = name.clone();
tokenizer.events[exit_index].name = name;
return;
}
if diff > 0 || slice.after > 0 {
let exit_point = tokenizer.events[exit_index].point.clone();
let mut enter_point = exit_point.clone();
enter_point.index -= diff;
enter_point.column -= diff;
enter_point.vs = 0;
tokenizer.map.add(
exit_index + 1,
0,
vec![
Event {
kind: Kind::Enter,
name: name.clone(),
point: enter_point.clone(),
link: None,
},
Event {
kind: Kind::Exit,
name,
point: exit_point,
link: None,
},
],
);
tokenizer.events[exit_index].point = enter_point;
slice.bytes = &slice.bytes[..index];
}
}
if trim_start {
let mut index = 0;
while index < slice.bytes.len() {
match slice.bytes[index] {
b' ' | b'\t' => index += 1,
_ => break,
}
}
// The whole data is whitespace.
// We can be very fast: we only change the token types.
if index == slice.bytes.len() {
tokenizer.events[exit_index - 1].name = Name::SpaceOrTab;
tokenizer.events[exit_index].name = Name::SpaceOrTab;
return;
}
if index > 0 || slice.before > 0 {
let enter_point = tokenizer.events[exit_index - 1].point.clone();
let mut exit_point = enter_point.clone();
exit_point.index += index;
exit_point.column += index;
exit_point.vs = 0;
tokenizer.map.add(
exit_index - 1,
0,
vec![
Event {
kind: Kind::Enter,
name: Name::SpaceOrTab,
point: enter_point,
link: None,
},
Event {
kind: Kind::Exit,
name: Name::SpaceOrTab,
point: exit_point.clone(),
link: None,
},
],
);
tokenizer.events[exit_index - 1].point = exit_point;
}
}
}
|
