feat(site-publish): add split-surface publishing

Authored-By: OpenAI (GPT-5) <noreply@openai.com>

scripts/deploy.py

@@ -78,6 +78,29 @@ def _immutable_head(endpoint, bucket, key, aws_env):
     raise RuntimeError(f"head-object failed for s3://{bucket}/{key}: {error.strip()}")
 
 
+def _existing_immutable_keys(endpoint, bucket, aws_env):
+    """Enumerate publisher-owned immutable keys so route moves cannot delete them."""
+    args = ["aws", "--endpoint-url", endpoint, "s3api", "list-objects-v2",
+            "--bucket", bucket, "--query", "Contents[].Key", "--output", "json"]
+    result = _aws_capture(args, aws_env)
+    if result.returncode != 0:
+        error = f"{result.stdout}\n{result.stderr}".strip()
+        raise RuntimeError(f"list-objects-v2 failed for s3://{bucket}: {error}")
+    keys = json.loads(result.stdout) or []
+    immutable = []
+    for key in keys:
+        info = _immutable_head(endpoint, bucket, key, aws_env)
+        if info is None:
+            continue
+        directives = {
+            part.strip().lower().split("=", 1)[0]
+            for part in (info.get("CacheControl") or "").split(",")
+        }
+        if "immutable" in directives:
+            immutable.append(key)
+    return sorted(immutable)
+
+
 def _immutable_digests(source, cache_control, content_type):
     with source.open("rb") as stream:
         content_digest = hashlib.file_digest(stream, "sha256").hexdigest()
@@ -172,6 +195,7 @@ def s3_sync(artifact, route, site_dir, credential_env_names=None):
     destination = f"{bucket_destination}{object_prefix + '/' if object_prefix else ''}"
     default_cache = next(rule["cache_control"] for rule in artifact["cache_rules"] if not rule["path"])
     immutable_paths = [rule["path"] for rule in artifact["cache_rules"] if _is_immutable(rule)]
+    existing_immutable = _existing_immutable_keys(endpoint, bucket, aws_env)
     # `excludes` are patterns (site.yaml `excludes:` list) that should never
     # be uploaded *and* should never be deleted from the bucket — escape hatch
     # for assets managed out-of-band (e.g. large PDFs uploaded via aws-cli).
@@ -188,6 +212,7 @@ def s3_sync(artifact, route, site_dir, credential_env_names=None):
     stage = None
     sync_source = html_dir
     sync_excludes = [*exclude_args,
+                     *(arg for key in existing_immutable for arg in ("--exclude", key)),
                      *(arg for path in immutable_paths for arg in ("--exclude", f"{path}/*"))]
     if object_prefix:
         stage = tempfile.TemporaryDirectory()
@@ -198,6 +223,7 @@ def s3_sync(artifact, route, site_dir, credential_env_names=None):
         sync_excludes = [
             *(arg for pattern in artifact["excludes"]
               for arg in ("--exclude", f"{object_prefix}/{pattern}")),
+            *(arg for key in existing_immutable for arg in ("--exclude", key)),
             *(arg for path in immutable_paths
               for arg in ("--exclude", f"{object_prefix}/{path}/*")),
         ]

tests/test_contract.py

@@ -389,6 +389,9 @@ class PublishingTests(unittest.TestCase):
             with patch.dict(os.environ, {
                 "DIST_S3_ACCESS_KEY": "dist-key", "DIST_S3_SECRET_KEY": secret
             }, clear=False), patch.object(deploy, "run", side_effect=capture), \
+                    patch.object(deploy, "_existing_immutable_keys", return_value=[
+                        "old-dist/releases/preserved.js"
+                    ]), \
                     patch.object(deploy, "publish_immutable_rule") as immutable_publish, \
                     redirect_stdout(output):
                 deploy.s3_sync(artifact, route, root)
@@ -400,6 +403,7 @@ class PublishingTests(unittest.TestCase):
             rendered = [" ".join(command) for command, _ in commands]
             self.assertIn("s3://baseline-dist/", rendered[0])
             self.assertIn("dist/releases/*", rendered[0])
+            self.assertIn("old-dist/releases/preserved.js", rendered[0])
             self.assertTrue(all("s3://baseline-dist/dist/" in command for command in rendered[1:]))
             self.assertTrue(any("channels/" in command and
                                 "public, max-age=0, must-revalidate" in command