🔒 FIX: escape colon_fence class and tasklists label_after content

mdit_py_plugins/colon_fence.py

@@ -153,7 +153,7 @@ def _render(
 
     return (
         "<pre><code"
-        + (f' class="block-{block_name}" ' if block_name else "")
+        + (f' class="block-{escapeHtml(block_name)}" ' if block_name else "")
         + ">"
         + content
         + "</code></pre>\n"

mdit_py_plugins/tasklists/__init__.py

@@ -21,6 +21,7 @@
 from uuid import uuid4
 
 from markdown_it import MarkdownIt
+from markdown_it.common.utils import escapeHtml
 from markdown_it.rules_core import StateCore
 from markdown_it.token import Token
 
@@ -132,7 +133,8 @@ def end_label() -> Token:
     def after_label(content: str, checkbox_id: str) -> Token:
         token = Token("html_inline", "", 0)
         token.content = (
-            f'<label class="task-list-item-label" for="{checkbox_id}">{content}</label>'
+            '<label class="task-list-item-label" '
+            f'for="{checkbox_id}">{escapeHtml(content)}</label>'
         )
         token.attrs = {"for": checkbox_id}
         return token

tests/test_colon_fence.py

@@ -36,3 +36,36 @@ def test_plugin_parse(data_regression):
         )
     )
     data_regression.check([t.as_dict() for t in tokens])
+
+
+@pytest.mark.parametrize(
+    "source,expected",
+    [
+        pytest.param(
+            ':::a"onmouseover="alert(1)\nx\n:::\n',
+            '<pre><code class="block-a&quot;onmouseover=&quot;alert(1)" >x\n'
+            "</code></pre>\n",
+            id="raw-quote",
+        ),
+        pytest.param(
+            ":::a&quot;&gt;&lt;xss&gt;\nx\n:::\n",
+            '<pre><code class="block-a&quot;&gt;&lt;xss&gt;" >x\n</code></pre>\n',
+            id="entity-encoded",
+        ),
+        pytest.param(
+            # benign: same attribute value as before, now well-formed
+            ":::a&b\nx\n:::\n",
+            '<pre><code class="block-a&amp;b" >x\n</code></pre>\n',
+            id="ampersand",
+        ),
+    ],
+)
+def test_block_name_is_escaped(source, expected):
+    """The block name must not break out of the ``class`` attribute (#143).
+
+    With ``html=False`` the output must not carry markup from the source,
+    but the name used to be interpolated into the attribute unescaped
+    (and the entity-encoded form was unescaped first, then interpolated).
+    """
+    md = MarkdownIt("commonmark", {"html": False}).use(colon_fence_plugin)
+    assert md.render(source) == expected

tests/test_tasklists.py

@@ -1,4 +1,5 @@
 from pathlib import Path
+import re
 from textwrap import dedent
 
 from markdown_it import MarkdownIt
@@ -32,3 +33,49 @@ def test_all(line, title, input, expected):
     text = md.render(input)
     print(text)
     assert text.rstrip() == expected.rstrip()
+
+
+@pytest.mark.parametrize(
+    "source,label",
+    [
+        pytest.param(
+            "- [ ] <img src=x onerror=alert(1)>\n",
+            " &lt;img src=x onerror=alert(1)&gt;",
+            id="raw-html",
+        ),
+        pytest.param("- [x] a & b\n", " a &amp; b", id="ampersand"),
+        pytest.param("- [ ] plain task\n", " plain task", id="plain"),
+    ],
+)
+def test_label_after_is_escaped(source, label):
+    """The ``label_after`` label must not carry raw HTML from the source (#143).
+
+    The label is built from the item's source text and emitted as an
+    ``html_inline`` token, so it used to bypass ``html=False``.
+    """
+    md = MarkdownIt("commonmark", {"html": False}).use(
+        tasklists_plugin, label=True, label_after=True
+    )
+    text = re.sub(r"task-item-[0-9a-f-]+", "task-item-ID", md.render(source))
+    checked = ' checked="checked"' if source.startswith("- [x]") else ""
+    assert text == (
+        '<ul class="contains-task-list">\n'
+        '<li class="task-list-item"><input class="task-list-item-checkbox"'
+        f'{checked} disabled="disabled" type="checkbox" id="task-item-ID">'
+        f'<label class="task-list-item-label" for="task-item-ID">{label}</label>'
+        "</li>\n</ul>\n"
+    )
+
+
+def test_label_after_entity_is_shown_literally():
+    """The ``label_after`` label is the item's raw source text, escaped (#143).
+
+    So a source entity is shown as written (``&amp;``), not decoded.
+    Only the label is asserted: the inline children before the last one
+    are also rendered outside the label, which is a separate, existing issue.
+    """
+    md = MarkdownIt("commonmark", {"html": False}).use(
+        tasklists_plugin, label=True, label_after=True
+    )
+    labels = re.findall(r"<label [^>]*>(.*?)</label>", md.render("- [ ] a &amp; b\n"))
+    assert labels == [" a &amp;amp; b"]