Resolve recipe pages to their hero image for cover scraping

Mealie's POST /api/recipes/{slug}/image downloads the URL it is given and
rejects anything that is not an image, so handing it a recipe page answered
400 {"message": "Url is not an image"}. This was read as a source-specific
failure on ica.se, but that page's own og:image uploads fine (verified live,
200 + 74608 bytes of image/webp), and httpbin's JPEG works too — the endpoint
simply never accepted pages.

Callers naturally pass the recipe page they just imported from, so resolve a
page to its advertised og:image and retry rather than making every caller know
the distinction. A page with no discoverable image now raises instead of
leaving the recipe silently coverless.

Co-authored-by: fredamn76 <fredrik.fallman@gmail.com>
Signed-off-by: fredamn76 <fredrik.fallman@gmail.com>

mealie_mcp/server.py

@@ -1,6 +1,7 @@
 """Small, confirmation-friendly MCP facade over the Mealie API."""
 from __future__ import annotations
 
+import html
 import json
 import os
 import re
@@ -546,17 +547,53 @@ def parse_ingredients(slug_or_id: str) -> dict[str, Any]:
     }
 
 
+#: Hero image declared by a recipe page, in the order worth trusting.
+_HERO_META = (
+    re.compile(r'<meta[^>]+property=["\']og:image["\'][^>]+content=["\']([^"\']+)', re.I),
+    re.compile(r'<meta[^>]+content=["\']([^"\']+)["\'][^>]+property=["\']og:image["\']', re.I),
+    re.compile(r'<meta[^>]+name=["\']twitter:image["\'][^>]+content=["\']([^"\']+)', re.I),
+)
+
+
+def _hero_image_url(page_url: str) -> str | None:
+    """Find the image a recipe page advertises as its own.
+
+    Mealie's cover endpoint downloads the URL it is given and rejects anything
+    that is not an image, so a page has to be reduced to a real image URL first.
+    Fetched with a browser user-agent because recipe sites refuse default ones.
+    """
+    try:
+        response = httpx.get(
+            page_url,
+            headers={"User-Agent": "Mozilla/5.0 (X11; Linux x86_64)"},
+            timeout=30,
+            follow_redirects=True,
+        )
+        response.raise_for_status()
+    except httpx.HTTPError:
+        return None
+    for pattern in _HERO_META:
+        match = pattern.search(response.text)
+        if match:
+            return html.unescape(match.group(1))
+    return None
+
+
 @mcp.tool()
 def set_cover_image(
     slug_or_id: str,
     source_url: str | None = None,
     image_path: str | None = None,
 ) -> dict[str, Any]:
-    """Set a recipe cover image by scraping a URL or uploading a local file.
+    """Set a recipe cover image from a URL or a local file.
+
+    ``source_url`` may be a recipe page or a direct image URL. Mealie's endpoint
+    only accepts the latter — handing it a page answers
+    ``400 {"message": "Url is not an image"}`` — so a page is resolved to its
+    ``og:image`` first and that is what gets sent.
 
     Local uploads must send the multipart ``extension`` field or the upload fails
-    validation silently on this instance. When a scrape does not produce a visible
-    cover (observed on Allrecipes), download the hero image and upload it instead.
+    validation silently on this instance.
     """
     if not source_url and not image_path:
         raise ValueError("Provide either source_url or image_path")
@@ -575,11 +612,27 @@ def set_cover_image(
             _raise_for_status(response)
         method = f"upload ({extension})"
     else:
-        _request("POST", f"/api/recipes/{slug_or_id}/image", json={"url": source_url})
+        used_url = source_url
         method = "scrape"
+        try:
+            _request("POST", f"/api/recipes/{slug_or_id}/image", json={"url": used_url})
+        except httpx.HTTPStatusError as exc:
+            if exc.response.status_code != 400 or "not an image" not in exc.response.text:
+                raise
+            used_url = _hero_image_url(source_url)
+            if not used_url:
+                raise ValueError(
+                    f"{source_url} is not an image and no og:image was found on the page; "
+                    "pass a direct image URL or use image_path"
+                ) from exc
+            _request("POST", f"/api/recipes/{slug_or_id}/image", json={"url": used_url})
+            method = "scrape (resolved og:image)"
 
     recipe = get_recipe(slug_or_id)
-    return {"method": method, "image": recipe.get("image"), "slug": recipe.get("slug")}
+    result = {"method": method, "image": recipe.get("image"), "slug": recipe.get("slug")}
+    if not image_path and used_url != source_url:
+        result["resolved_image_url"] = used_url
+    return result
 
 
 # --- Verify -----------------------------------------------------------------