ci(validate): detect and deny HTML in README

.github/workflows/validate-plugins.py

@@ -93,6 +93,19 @@ SETTING_FIELDS = {
 OPTION_FIELDS = {"value", "label_key"}
 VISIBLE_WHEN_FIELDS = {"key", "values"}
 
+# Raw HTML is not supported on plugin pages. Markdown autolinks such as
+# <https://example.com> do not match this expression.
+HTML_RE = re.compile(
+    r"<!--|<\?|<!\[CDATA\[|<![A-Z]|"
+    r"</[A-Za-z][A-Za-z0-9-]*\s*>|"
+    r"<[A-Za-z][A-Za-z0-9-]*"
+    r"(?:\s+[A-Za-z_:][A-Za-z0-9_.:-]*"
+    r"(?:\s*=\s*(?:[^\s\"'=<>`]+|'[^']*'|\"[^\"]*\"))?)*\s*/?>",
+    re.DOTALL,
+)
+INLINE_CODE_RE = re.compile(r"(?<!`)(`+)(?!`)(.*?)(?<!`)\1(?!`)", re.DOTALL)
+FENCE_OPEN_RE = re.compile(r"^ {0,3}(`{3,}|~{3,})")
+
 
 def is_non_empty_string(value: Any) -> bool:
     return isinstance(value, str) and value.strip() != ""
@@ -113,6 +126,42 @@ def rel(root: Path, path: Path) -> str:
         return path.as_posix()
 
 
+def raw_html_line(markdown: str) -> int | None:
+    """Return the first line containing raw HTML outside Markdown code, if any."""
+    visible: list[str] = []
+    fence_char = ""
+    fence_length = 0
+
+    for line in markdown.splitlines(keepends=True):
+        if fence_char:
+            closing = rf"^ {{0,3}}{re.escape(fence_char)}{{{fence_length},}}\s*$"
+            if re.match(closing, line.rstrip("\r\n")):
+                fence_char = ""
+                fence_length = 0
+            visible.append("\n" if line.endswith(("\n", "\r")) else "")
+            continue
+
+        opening = FENCE_OPEN_RE.match(line)
+        if opening:
+            fence = opening.group(1)
+            fence_char = fence[0]
+            fence_length = len(fence)
+            visible.append("\n" if line.endswith(("\n", "\r")) else "")
+            continue
+
+        visible.append(line)
+
+    text = "".join(visible)
+    text = INLINE_CODE_RE.sub(
+        lambda match: "".join("\n" if char == "\n" else " " for char in match.group(0)),
+        text,
+    )
+    match = HTML_RE.search(text)
+    if match is None:
+        return None
+    return text.count("\n", 0, match.start()) + 1
+
+
 def webp_dimensions(header: bytes) -> tuple[int, int] | None:
     """Width and height from a WebP header, or None if it is not one we can read.
 
@@ -749,6 +798,21 @@ class Validator:
                 f"Export one with {THUMBNAIL_GENERATOR_URL}",
             )
 
+    def validate_readme(self, plugin_dir: Path) -> None:
+        readme = plugin_dir / "README.md"
+        if not readme.is_file():
+            return
+
+        try:
+            contents = readme.read_text(encoding="utf-8")
+        except UnicodeDecodeError:
+            self.add_error(readme, "must be UTF-8 text")
+            return
+
+        line = raw_html_line(contents)
+        if line is not None:
+            self.add_error(readme, f"raw HTML on line {line} is not allowed; use Markdown instead")
+
     def validate_no_symlinks(self, manifest_path: Path, plugin_dir: Path) -> None:
         for path in plugin_dir.rglob("*"):
             if path.is_symlink():
@@ -765,6 +829,7 @@ class Validator:
         self.validate_root_fields(manifest_path, manifest)
         self.validate_required_files(manifest_path, plugin_dir)
         self.validate_thumbnail(manifest_path, plugin_dir)
+        self.validate_readme(plugin_dir)
         self.validate_no_symlinks(manifest_path, plugin_dir)
 
         if "setting" in manifest: