pub fn html_detect(input: &str) -> bool {
let chars: Vec<char> = input.chars().collect();
let mut i = 0;
while i < chars.len() {
match chars[i] {
'<' => {
if scan_tag(&chars, i).is_some() {
return true;
}
i += 1;
}
'&' => {
if scan_entity(&chars, i).is_some() {
return true;
}
i += 1;
}
_ => i += 1,
}
}
false
}
pub fn html_strip(input: &str) -> String {
let chars: Vec<char> = input.chars().collect();
let mut out = String::with_capacity(input.len());
let mut i = 0;
while i < chars.len() {
match chars[i] {
'<' => {
let Some(tag) = scan_tag(&chars, i) else {
out.push('<');
i += 1;
continue;
};
if let Some(end) = skip_raw_element(&chars, &tag) {
i = end;
continue;
}
if tag.breaks_line {
out.push('\n');
}
i = tag.end;
}
'&' => {
if let Some((decoded, end)) = scan_entity(&chars, i) {
out.push_str(&decoded);
i = end;
} else {
out.push('&');
i += 1;
}
}
c => {
out.push(c);
i += 1;
}
}
}
out
}
struct Tag {
name: String,
closing: bool,
end: usize,
breaks_line: bool,
}
fn scan_tag(chars: &[char], start: usize) -> Option<Tag> {
if chars.get(start + 1) == Some(&'!') {
if chars.get(start + 2) == Some(&'-') && chars.get(start + 3) == Some(&'-') {
let mut i = start + 4;
while i + 2 < chars.len() {
if chars[i] == '-' && chars[i + 1] == '-' && chars[i + 2] == '>' {
return Some(Tag { name: String::new(), closing: false, end: i + 3, breaks_line: false });
}
i += 1;
}
return None;
}
let mut i = start + 2;
while i < chars.len() {
if chars[i] == '>' {
return Some(Tag { name: String::new(), closing: false, end: i + 1, breaks_line: false });
}
i += 1;
}
return None;
}
let mut i = start + 1;
let closing = chars.get(i) == Some(&'/');
if closing {
i += 1;
}
let name_start = i;
while i < chars.len() && (chars[i].is_ascii_alphanumeric()) {
i += 1;
}
if i == name_start || !chars[name_start].is_ascii_alphabetic() {
return None;
}
let name: String = chars[name_start..i].iter().collect::<String>().to_ascii_lowercase();
let mut quote: Option<char> = None;
while i < chars.len() {
let c = chars[i];
match quote {
Some(q) => {
if c == q {
quote = None;
}
}
None => match c {
'"' | '\'' => quote = Some(c),
'>' => {
let breaks_line = matches!(
name.as_str(),
"br" | "hr"
) || (closing
&& matches!(
name.as_str(),
"p" | "div" | "li" | "tr" | "table" | "ul" | "ol"
| "h1" | "h2" | "h3" | "h4" | "h5" | "h6"
));
return Some(Tag { name, closing, end: i + 1, breaks_line });
}
'<' => return None, _ => {}
},
}
i += 1;
}
None
}
fn skip_raw_element(chars: &[char], tag: &Tag) -> Option<usize> {
if tag.closing || !matches!(tag.name.as_str(), "script" | "style") {
return None;
}
let closer: Vec<char> = format!("</{}", tag.name).chars().collect();
let mut i = tag.end;
while i < chars.len() {
if chars[i] == '<' && matches_ignore_case(chars, i, &closer) {
let mut j = i + closer.len();
while j < chars.len() && chars[j] != '>' {
j += 1;
}
return Some((j + 1).min(chars.len()));
}
i += 1;
}
Some(chars.len())
}
fn matches_ignore_case(chars: &[char], at: usize, pat: &[char]) -> bool {
if at + pat.len() > chars.len() {
return false;
}
pat.iter()
.enumerate()
.all(|(k, p)| chars[at + k].to_ascii_lowercase() == *p)
}
fn scan_entity(chars: &[char], start: usize) -> Option<(String, usize)> {
let mut i = start + 1;
let numeric = chars.get(i) == Some(&'#');
if numeric {
i += 1;
}
let hex = numeric && matches!(chars.get(i), Some('x') | Some('X'));
if hex {
i += 1;
}
let body_start = i;
while i < chars.len() && i - body_start < 10 {
let c = chars[i];
if c == ';' {
let body: String = chars[body_start..i].iter().collect();
let decoded = if numeric {
let code = if hex {
u32::from_str_radix(&body, 16).ok()?
} else {
body.parse::<u32>().ok()?
};
let c = char::from_u32(code)?;
c.to_string()
} else {
match body.as_str() {
"amp" => "&".to_string(),
"lt" => "<".to_string(),
"gt" => ">".to_string(),
"quot" => "\"".to_string(),
"apos" => "'".to_string(),
"nbsp" => " ".to_string(),
"copy" => "©".to_string(),
"yen" => "¥".to_string(),
"middot" => "・".to_string(),
_ => return None,
}
};
return Some((decoded, i + 1));
}
if !(c.is_ascii_alphanumeric()) {
return None;
}
i += 1;
}
None
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn detect() {
assert!(html_detect("<p>本文</p>"));
assert!(html_detect("改行<br/>あり"));
assert!(html_detect("A & B"));
assert!(html_detect("<!-- c -->"));
assert!(!html_detect("a < b かつ b > c"));
assert!(!html_detect("Q&A、R&D"));
assert!(!html_detect("プレーンテキスト"));
}
#[test]
fn strip_tags() {
assert_eq!(html_strip("<p>本文</p>"), "本文\n");
assert_eq!(html_strip("1行目<br>2行目"), "1行目\n2行目");
assert_eq!(html_strip("<a href=\"https://x.jp/?a=1&b=2\">リンク</a>先"), "リンク先");
assert_eq!(html_strip("a < b はそのまま"), "a < b はそのまま");
}
#[test]
fn strip_entities() {
assert_eq!(html_strip("A&B C"), "A&B C");
assert_eq!(html_strip("ああ"), "ああ");
assert_eq!(html_strip("Q&A のまま"), "Q&A のまま");
}
#[test]
fn strip_raw_elements() {
assert_eq!(html_strip("前<script>var a=1;</script>後"), "前後");
assert_eq!(html_strip("前<style>p{}</style>後"), "前後");
assert_eq!(html_strip("<STYLE>p{}</STYLE>後"), "後");
}
#[test]
fn question_mark_survives() {
assert_eq!(html_strip("<p>いかがですか?</p>"), "いかがですか?\n");
}
}