1 //! Detects invalid HTML (like an unclosed `<span>`) in doc comments.
4 use crate::core::DocContext;
5 use crate::html::markdown::main_body_opts;
6 use crate::visit::DocVisitor;
8 use pulldown_cmark::{Event, Parser, Tag};
10 use std::iter::Peekable;
12 use std::str::CharIndices;
14 crate const CHECK_INVALID_HTML_TAGS: Pass = Pass {
15 name: "check-invalid-html-tags",
16 run: check_invalid_html_tags,
17 description: "detects invalid HTML tags in doc comments",
20 struct InvalidHtmlTagsLinter<'a, 'tcx> {
21 cx: &'a mut DocContext<'tcx>,
24 crate fn check_invalid_html_tags(krate: Crate, cx: &mut DocContext<'_>) -> Crate {
25 if cx.tcx.sess.is_nightly_build() {
26 let mut coll = InvalidHtmlTagsLinter { cx };
27 coll.visit_crate(&krate);
32 const ALLOWED_UNCLOSED: &[&str] = &[
33 "area", "base", "br", "col", "embed", "hr", "img", "input", "keygen", "link", "meta", "param",
34 "source", "track", "wbr",
38 tags: &mut Vec<(String, Range<usize>)>,
41 f: &impl Fn(&str, &Range<usize>),
43 let tag_name_low = tag_name.to_lowercase();
44 if let Some(pos) = tags.iter().rposition(|(t, _)| t.to_lowercase() == tag_name_low) {
45 // If the tag is nested inside a "<script>" or a "<style>" tag, no warning should
47 let should_not_warn = tags.iter().take(pos + 1).any(|(at, _)| {
48 let at = at.to_lowercase();
49 at == "script" || at == "style"
51 for (last_tag_name, last_tag_span) in tags.drain(pos + 1..) {
55 let last_tag_name_low = last_tag_name.to_lowercase();
56 if ALLOWED_UNCLOSED.contains(&last_tag_name_low.as_str()) {
59 // `tags` is used as a queue, meaning that everything after `pos` is included inside it.
60 // So `<h2><h3></h2>` will look like `["h2", "h3"]`. So when closing `h2`, we will still
61 // have `h3`, meaning the tag wasn't closed as it should have.
62 f(&format!("unclosed HTML tag `{}`", last_tag_name), &last_tag_span);
64 // Remove the `tag_name` that was originally closed
67 // It can happen for example in this case: `<h2></script></h2>` (the `h2` tag isn't required
68 // but it helps for the visualization).
69 f(&format!("unopened HTML tag `{}`", tag_name), &range);
74 tags: &mut Vec<(String, Range<usize>)>,
78 iter: &mut Peekable<CharIndices<'_>>,
79 f: &impl Fn(&str, &Range<usize>),
81 let mut tag_name = String::new();
82 let mut is_closing = false;
83 let mut prev_pos = start_pos;
86 let (pos, c) = match iter.peek() {
87 Some((pos, c)) => (*pos, *c),
88 // In case we reached the of the doc comment, we want to check that it's an
89 // unclosed HTML tag. For example "/// <h3".
90 None => (prev_pos, '\0'),
93 // Checking if this is a closing tag (like `</a>` for `<a>`).
94 if c == '/' && tag_name.is_empty() {
96 } else if c.is_ascii_alphanumeric() {
99 if !tag_name.is_empty() {
100 let mut r = Range { start: range.start + start_pos, end: range.start + pos };
102 // In case we have a tag without attribute, we can consider the span to
103 // refer to it fully.
107 // In case we have "</div >" or even "</div >".
109 if !c.is_whitespace() {
110 // It seems like it's not a valid HTML tag.
113 let mut found = false;
114 for (new_pos, c) in text[pos..].char_indices() {
115 if !c.is_whitespace() {
117 r.end = range.start + new_pos + 1;
127 drop_tag(tags, tag_name, r, f);
129 tags.push((tag_name, r));
139 tags: &mut Vec<(String, Range<usize>)>,
142 is_in_comment: &mut Option<Range<usize>>,
143 f: &impl Fn(&str, &Range<usize>),
145 let mut iter = text.char_indices().peekable();
147 while let Some((start_pos, c)) = iter.next() {
148 if is_in_comment.is_some() {
149 if text[start_pos..].starts_with("-->") {
150 *is_in_comment = None;
153 if text[start_pos..].starts_with("<!--") {
154 // We skip the "!--" part. (Once `advance_by` is stable, might be nice to use it!)
158 *is_in_comment = Some(Range {
159 start: range.start + start_pos,
160 end: range.start + start_pos + 3,
163 extract_html_tag(tags, text, &range, start_pos, &mut iter, f);
169 impl<'a, 'tcx> DocVisitor for InvalidHtmlTagsLinter<'a, 'tcx> {
170 fn visit_item(&mut self, item: &Item) {
171 let tcx = self.cx.tcx;
172 let hir_id = match DocContext::as_local_hir_id(tcx, item.def_id) {
173 Some(hir_id) => hir_id,
175 // If non-local, no need to check anything.
179 let dox = item.attrs.collapsed_doc_value().unwrap_or_default();
181 let report_diag = |msg: &str, range: &Range<usize>| {
182 let sp = match super::source_span_for_markdown_range(tcx, &dox, range, &item.attrs)
185 None => item.attr_span(tcx),
187 tcx.struct_span_lint_hir(crate::lint::INVALID_HTML_TAGS, hir_id, sp, |lint| {
188 lint.build(msg).emit()
192 let mut tags = Vec::new();
193 let mut is_in_comment = None;
194 let mut in_code_block = false;
196 let p = Parser::new_ext(&dox, main_body_opts()).into_offset_iter();
198 for (event, range) in p {
200 Event::Start(Tag::CodeBlock(_)) => in_code_block = true,
201 Event::Html(text) | Event::Text(text) if !in_code_block => {
202 extract_tags(&mut tags, &text, range, &mut is_in_comment, &report_diag)
204 Event::End(Tag::CodeBlock(_)) => in_code_block = false,
209 for (tag, range) in tags.iter().filter(|(t, _)| {
210 let t = t.to_lowercase();
211 !ALLOWED_UNCLOSED.contains(&t.as_str())
213 report_diag(&format!("unclosed HTML tag `{}`", tag), range);
216 if let Some(range) = is_in_comment {
217 report_diag("Unclosed HTML comment", &range);
221 self.visit_item_recur(item)