1//! Skip text characters for text token, place those to pending buffer
2//! and increment current pos
3//!
4use regex::{self, Regex};
5
6use crate::parser::inline::{InlineRule, InlineState};
7use crate::{MarkdownIt, Node, NodeValue, Renderer};
8
9#[derive(Debug)]
10/// Plain text AST node.
11pub struct Text {
12 pub content: String
13}
14
15impl NodeValue for Text {
16 fn render(&self, _: &Node, fmt: &mut dyn Renderer) {
17 fmt.text(&self.content);
18 }
19}
20
21#[derive(Debug)]
22/// Escaped text AST node (backslash escapes and entities).
23pub struct TextSpecial {
24 pub content: String,
25 pub markup: String,
26 pub info: &'static str,
27}
28
29impl NodeValue for TextSpecial {
30 fn render(&self, _: &Node, fmt: &mut dyn Renderer) {
31 fmt.text(&self.content);
32 }
33}
34
35pub fn add(md: &mut MarkdownIt) {
36 md.inline.add_rule::<TextScanner>()
37 .before_all();
38}
39
40#[derive(Debug)]
41pub(crate) enum TextScannerImpl {
42 SkipPunct,
43 SkipRegex(Regex),
44}
45
46/// Rule to skip pure text
47/// '{}$%@~+=:' reserved for extensions
48///
49/// !, ", #, $, %, &, ', (, ), *, +, ,, -, ., /, :, ;, <, =, >, ?, @, [, \, ], ^, _, `, {, |, }, or ~
50///
51/// !!!! Don't confuse with "Markdown ASCII Punctuation" chars
52/// http://spec.commonmark.org/0.15/#ascii-punctuation-character
53///
54pub struct TextScanner;
55
56impl TextScanner {
57 fn choose_text_impl(charmap: Vec<char>) -> TextScannerImpl {
58 let mut can_use_punct = true;
59 for ch in charmap.iter() {
60 match ch {
61 '\n' | '!' | '#' | '$' | '%' | '&' | '*' | '+' | '-' |
62 ':' | '<' | '=' | '>' | '@' | '[' | '\\' | ']' | '^' |
63 '_' | '`' | '{' | '}' | '~' => {},
64 _ => {
65 can_use_punct = false;
66 break;
67 }
68 }
69 }
70
71 if can_use_punct {
72 TextScannerImpl::SkipPunct
73 } else {
74 TextScannerImpl::SkipRegex(
75 Regex::new(
76 // [] panics on "unclosed character class", but it cannot happen here
77 // (we'd use punct rule instead)
78 &format!("^[^{}]+", charmap.into_iter().map(
79 |c| regex::escape(&c.to_string())
80 ).collect::<String>())
81 ).unwrap()
82 )
83 }
84 }
85
86 fn find_text_length(state: &mut InlineState) -> usize {
87 let text_impl = state.md.inline.text_impl.get_or_init(
88 || Self::choose_text_impl(state.md.inline.text_charmap.keys().copied().collect())
89 );
90
91 let mut len = 0;
92
93 match text_impl {
94 TextScannerImpl::SkipPunct => {
95 let mut chars = state.src[state.pos..state.pos_max].chars();
96
97 loop {
98 match chars.next() {
99 Some(
100 '\n' | '!' | '#' | '$' | '%' | '&' | '*' | '+' | '-' |
101 ':' | '<' | '=' | '>' | '@' | '[' | '\\' | ']' | '^' |
102 '_' | '`' | '{' | '}' | '~'
103 ) => {
104 break;
105 }
106 Some(chr) => {
107 len += chr.len_utf8();
108 }
109 None => { break; }
110 }
111 }
112 }
113 TextScannerImpl::SkipRegex(re) => {
114 if let Some(capture) = re.find(&state.src[state.pos..state.pos_max]) {
115 len = capture.end();
116 }
117 }
118 }
119
120 len
121 }
122}
123
124impl InlineRule for TextScanner {
125 const MARKER: char = '\0';
126
127 fn check(state: &mut InlineState) -> Option<usize> {
128 let len = Self::find_text_length(state);
129 if len == 0 { return None; }
130 Some(len)
131 }
132
133 fn run(state: &mut InlineState) -> Option<(Node, usize)> {
134 let len = Self::find_text_length(state);
135 if len == 0 { return None; }
136 state.trailing_text_push(state.pos, state.pos + len);
137 state.pos += len;
138 Some((Node::default(), 0))
139 }
140}