fix false negative in rustdoc::invalid_html_tags with tags like <img>

previously, this lint did not distinguish between `<img` and `<img>`,
and since the latter should be accepted under html5,
the former was also accepted.

Author: lolbinarycat

src/librustdoc/passes/lint/html_tags.rs

@@ -132,16 +132,16 @@ pub(crate) fn visit_item(cx: &DocContext<'_>, item: &Item, hir_id: HirId, dox: &
         match event {
             Event::Start(Tag::CodeBlock(_)) => in_code_block = true,
             Event::Html(text) | Event::InlineHtml(text) if !in_code_block => {
-                extract_tags(&mut tags, &text, range, &mut is_in_comment, &report_diag)
+                extract_tags(&mut tags, &text, range, dox, &mut is_in_comment, &report_diag)
             }
             Event::End(TagEnd::CodeBlock) => in_code_block = false,
             _ => {}
         }
     }
 
-    for (tag, range) in tags.iter().filter(|(t, _)| {
+    for (tag, range) in tags.iter().filter(|(t, range)| {
         let t = t.to_lowercase();
-        !ALLOWED_UNCLOSED.contains(&t.as_str())
+        !is_implicitly_self_closing(&t, range.clone(), dox)
     }) {
         report_diag(format!("unclosed HTML tag `{tag}`"), range, true);
     }
@@ -151,15 +151,24 @@ pub(crate) fn visit_item(cx: &DocContext<'_>, item: &Item, hir_id: HirId, dox: &
     }
 }
 
+/// These tags are interpreted as self-closing if they lack an explict closing tag.
 const ALLOWED_UNCLOSED: &[&str] = &[
     "area", "base", "br", "col", "embed", "hr", "img", "input", "keygen", "link", "meta", "param",
     "source", "track", "wbr",
 ];
 
+/// Allows constructs like `<img>`, but not `<img`.
+fn is_implicitly_self_closing(tag_name: &str, range: Range<usize>, dox: &str) -> bool {
+    ALLOWED_UNCLOSED.contains(&tag_name) &&
+    // search in reverse order to try to exit early, since `>` appears near the end in the happy case
+        dox[range].bytes().rev().any(|c| c == b'>')
+}
+
 fn drop_tag(
     tags: &mut Vec<(String, Range<usize>)>,
     tag_name: String,
     range: Range<usize>,
+    dox: &str,
     f: &impl Fn(String, &Range<usize>, bool),
 ) {
     let tag_name_low = tag_name.to_lowercase();
@@ -175,7 +184,7 @@ fn drop_tag(
                 continue;
             }
             let last_tag_name_low = last_tag_name.to_lowercase();
-            if ALLOWED_UNCLOSED.contains(&last_tag_name_low.as_str()) {
+            if is_implicitly_self_closing(&last_tag_name_low, last_tag_span.clone(), dox) {
                 continue;
             }
             // `tags` is used as a queue, meaning that everything after `pos` is included inside it.
@@ -256,6 +265,7 @@ fn extract_html_tag(
     tags: &mut Vec<(String, Range<usize>)>,
     text: &str,
     range: &Range<usize>,
+    dox: &str,
     start_pos: usize,
     iter: &mut Peekable<CharIndices<'_>>,
     f: &impl Fn(String, &Range<usize>, bool),
@@ -306,7 +316,7 @@ fn extract_html_tag(
                             break;
                         }
                     }
-                    drop_tag(tags, tag_name, r, f);
+                    drop_tag(tags, tag_name, r, dox, f);
                 } else {
                     let mut is_self_closing = false;
                     let mut quote_pos = None;
@@ -374,6 +384,7 @@ fn extract_tags(
     tags: &mut Vec<(String, Range<usize>)>,
     text: &str,
     range: Range<usize>,
+    dox: &str,
     is_in_comment: &mut Option<Range<usize>>,
     f: &impl Fn(String, &Range<usize>, bool),
 ) {
@@ -395,7 +406,7 @@ fn extract_tags(
                     end: range.start + start_pos + 3,
                 });
             } else {
-                extract_html_tag(tags, text, &range, start_pos, &mut iter, f);
+                extract_html_tag(tags, text, &range, dox, start_pos, &mut iter, f);
             }
         }
     }