From 36dda46ddd877c750bd0c303f7a6f1fc7602647b Mon Sep 17 00:00:00 2001
From: Abdessamad Derraz <3028866+Abdess@users.noreply.github.com>
Date: Sat, 17 Oct 2026 15:06:50 +0200
Subject: [PATCH] fix: resolve api catalog links from the catalog

---
 scripts/generate_site.py   | 12 ++++++++++--
 tests/test_site_exports.py | 14 ++++++++++++--
 2 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/scripts/generate_site.py b/scripts/generate_site.py
index 0bd1bb00..952f8905 100644
--- a/scripts/generate_site.py
+++ b/scripts/generate_site.py
@@ -959,6 +959,7 @@ def generate_data_exports(
         (downloads / "retrobios.sqlite", "SQLite snapshot", "application/vnd.sqlite3", None),
     ]
     catalog_items: list[dict] = []
+    published: list[dict] = []
     for path, title, media_type, schema in assets:
         relative = path.relative_to(docs).as_posix()
         item = {
@@ -971,12 +972,19 @@ def generate_data_exports(
         if schema:
             item["schema"] = schema
         catalog_items.append(item)
+        # In the catalog both references resolve against the catalog's own
+        # URL, as RFC 3986 clients do: `url` was relative to the site root
+        # and `schema` to api/v1/, so no single base resolved both.
+        published.append({
+            **item,
+            "url": os.path.relpath(path, api).replace(os.sep, "/"),
+        })
     catalog = {
         "schema_version": 1,
         "generated_at": generated_at,
         "kind": "catalog",
-        "count": len(catalog_items),
-        "items": catalog_items,
+        "count": len(published),
+        "items": published,
     }
     write_if_changed(
         str(api / "catalog.json"),
diff --git a/tests/test_site_exports.py b/tests/test_site_exports.py
index 3ef986c9..fecc1ec5 100644
--- a/tests/test_site_exports.py
+++ b/tests/test_site_exports.py
@@ -183,9 +183,19 @@ class SiteReferenceContracts(unittest.TestCase):
         )
         catalog = json.loads((self.docs / "api/v1/catalog.json").read_text())
         self.assertEqual(catalog["count"], len(exports))
+        # Every reference resolves against the catalog's own URL, as an
+        # RFC 3986 client resolves it: the url was relative to the site root
+        # and the schema to api/v1/, so neither base resolved both.
+        import urllib.parse  # noqa: PLC0415
+
+        site = "https://example.org/site/"
+        base = site + "api/v1/catalog.json"
         for item in catalog["items"]:
-            path = self.docs / item["url"]
-            self.assertTrue(path.is_file())
+            for key in ("url", "schema"):
+                if key in item:
+                    target = urllib.parse.urljoin(base, item[key]).removeprefix(site)
+                    self.assertTrue((self.docs / target).is_file(), (key, item[key]))
+            path = self.docs / urllib.parse.urljoin(base, item["url"]).removeprefix(site)
             self.assertEqual(hashlib.sha256(path.read_bytes()).hexdigest(), item["sha256"])
 
         with sqlite3.connect(self.docs / "downloads/retrobios.sqlite") as connection: