feat(provenance): nh3 HTML sanitizer for scraped post descriptions

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

backend/app/utils/html_sanitize.py

@@ -0,0 +1,39 @@
+"""Backend HTML sanitization for scraped Post.description.
+
+Provenance descriptions come from gallery-dl (untrusted third-party HTML).
+We render them via v-html in the UI sub-slice, so they MUST be sanitized
+server-side to a tight allowlist. nh3 (Rust/ammonia bindings) is used;
+bleach is deprecated upstream and intentionally not used.
+"""
+
+import nh3
+
+ALLOWED_TAGS = {
+    "p", "a", "br", "em", "strong", "b", "i", "ul", "ol", "li", "blockquote",
+}
+ALLOWED_ATTRIBUTES = {"a": {"href", "title"}}
+ALLOWED_URL_SCHEMES = {"http", "https"}
+
+
+def sanitize_post_html(raw: str | None) -> str | None:
+    """Return sanitized HTML, or None for null/empty/whitespace input.
+
+    - keeps only ALLOWED_TAGS / ALLOWED_ATTRIBUTES
+    - only http/https hrefs survive (javascript: etc. dropped)
+    - <script> content removed entirely
+    - all <a> get rel="noopener noreferrer" (nh3 default link_rel)
+    - never raises
+    """
+    if raw is None:
+        return None
+    stripped = raw.strip()
+    if not stripped:
+        return None
+    return nh3.clean(
+        stripped,
+        tags=ALLOWED_TAGS,
+        attributes=ALLOWED_ATTRIBUTES,
+        url_schemes=ALLOWED_URL_SCHEMES,
+        link_rel="noopener noreferrer",
+        clean_content_tags={"script", "style"},
+    )