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>
2026-07-31 09:50:01 +02:00
parent 141b84bae1
commit c8775ac713
2 changed files with 147 additions and 5 deletions
@@ -259,3 +259,92 @@ class TestCoverImage:
    def test_requires_a_source(self):
        with pytest.raises(ValueError):
            server.set_cover_image("x")
+
+    def test_a_recipe_page_is_resolved_to_its_hero_image(self, mock_mealie, monkeypatch):
+        # Verified live: Mealie answers 400 "Url is not an image" for a recipe page,
+        # and 200 for that same page's og:image.
+        page = "https://www.ica.se/recept/kalsoppa-722069/"
+        hero = "https://assets.icanet.se/imagevaultfiles/kalsoppa.jpg"
+        posted: list[str] = []
+
+        def handler(request: httpx.Request) -> httpx.Response:
+            import json as _json
+
+            if request.method == "POST" and request.url.path.endswith("/image"):
+                url = _json.loads(request.content)["url"]
+                posted.append(url)
+                if url == page:
+                    return httpx.Response(
+                        400, json={"detail": {"message": "Url is not an image", "error": True}}
+                    )
+                return httpx.Response(200, json=None)
+            return httpx.Response(200, json={"slug": "soppa", "image": "abcd"})
+
+        mock_mealie(handler)
+        monkeypatch.setattr(
+            server, "_hero_image_url", lambda url: hero if url == page else None
+        )
+
+        result = server.set_cover_image("soppa", source_url=page)
+
+        # The page is tried first, then retried with the image it advertises.
+        assert posted == [page, hero]
+        assert result["method"] == "scrape (resolved og:image)"
+        assert result["resolved_image_url"] == hero
+
+    def test_a_direct_image_url_is_sent_unchanged(self, mock_mealie, monkeypatch):
+        recorded = mock_mealie(lambda r: httpx.Response(200, json={"slug": "x", "image": "img"}))
+        monkeypatch.setattr(server, "_hero_image_url", lambda url: pytest.fail("not needed"))
+
+        result = server.set_cover_image("x", source_url="https://example.com/hero.jpg")
+
+        assert result["method"] == "scrape"
+        assert "resolved_image_url" not in result
+        assert any(r.method == "POST" for r in recorded)
+
+    def test_a_page_with_no_hero_image_fails_loudly(self, mock_mealie, monkeypatch):
+        def handler(request: httpx.Request) -> httpx.Response:
+            if request.method == "POST" and request.url.path.endswith("/image"):
+                return httpx.Response(400, json={"detail": {"message": "Url is not an image"}})
+            return httpx.Response(200, json={"slug": "x"})
+
+        mock_mealie(handler)
+        monkeypatch.setattr(server, "_hero_image_url", lambda url: None)
+
+        with pytest.raises(ValueError, match="no og:image"):
+            server.set_cover_image("x", source_url="https://example.com/page")
+
+
+class TestHeroImageUrl:
+    def test_reads_og_image_in_either_attribute_order(self, monkeypatch):
+        pages = {
+            "a": '<meta property="og:image" content="https://x.test/a.jpg">',
+            "b": '<meta content="https://x.test/b.jpg" property="og:image">',
+            "c": '<meta name="twitter:image" content="https://x.test/c.jpg">',
+            "d": "<html><body>no image here</body></html>",
+        }
+
+        def fake_get(url, **kwargs):
+            return httpx.Response(200, text=pages[url], request=httpx.Request("GET", "http://t"))
+
+        monkeypatch.setattr(server.httpx, "get", fake_get)
+        assert server._hero_image_url("a") == "https://x.test/a.jpg"
+        assert server._hero_image_url("b") == "https://x.test/b.jpg"
+        assert server._hero_image_url("c") == "https://x.test/c.jpg"
+        assert server._hero_image_url("d") is None
+
+    def test_unescapes_entities_in_the_url(self, monkeypatch):
+        # Query strings in og:image arrive HTML-escaped and must be decoded.
+        page = '<meta property="og:image" content="https://x.test/a.jpg?w=1&amp;h=2">'
+        monkeypatch.setattr(
+            server.httpx, "get",
+            lambda url, **kw: httpx.Response(200, text=page, request=httpx.Request("GET", "http://t")),
+        )
+        assert server._hero_image_url("p") == "https://x.test/a.jpg?w=1&h=2"
+
+    def test_an_unreachable_page_is_not_an_error(self, monkeypatch):
+        def boom(url, **kwargs):
+            raise httpx.ConnectError("no route")
+
+        monkeypatch.setattr(server.httpx, "get", boom)
+        assert server._hero_image_url("https://example.com/page") is None