From c771c5ce5f1e05d017cfc1c2e77f69c22e7d3b7e Mon Sep 17 00:00:00 2001
From: jun09 <48198880+jun09@users.noreply.github.com>
Date: Sun, 19 Apr 2026 15:59:25 +0800
Subject: [PATCH] fix(amazon): Support AWS China region endpoints in
 RedshiftSQLHook OpenLineage identifier parsing

The _get_identifier_from_hostname method in RedshiftSQLHook only
handled global AWS endpoints (amazonaws.com) but not AWS China
region endpoints (amazonaws.com.cn). This caused the OpenLineage
authority part to fall back to the full hostname instead of
correctly parsing cluster_identifier.region_name.

Global endpoint format (6 dot-separated parts):
  my-cluster.id.us-east-1.redshift.amazonaws.com

China endpoint format (7 dot-separated parts):
  my-cluster.id.cn-north-1.redshift.amazonaws.com.cn

The same issue affects both provisioned clusters and
Redshift Serverless workgroups in cn-north-1 and cn-northwest-1.
---
 .../providers/amazon/aws/hooks/redshift_sql.py       |  4 ++++
 .../tests/unit/amazon/aws/hooks/test_redshift_sql.py | 12 ++++++++++++
 2 files changed, 16 insertions(+)

diff --git a/providers/amazon/src/airflow/providers/amazon/aws/hooks/redshift_sql.py b/providers/amazon/src/airflow/providers/amazon/aws/hooks/redshift_sql.py
index 0a55c2bdc26c2..d3ce58190426a 100644
--- a/providers/amazon/src/airflow/providers/amazon/aws/hooks/redshift_sql.py
+++ b/providers/amazon/src/airflow/providers/amazon/aws/hooks/redshift_sql.py
@@ -285,6 +285,10 @@ def _get_identifier_from_hostname(self, hostname: str) -> str:
         parts = hostname.split(".")
         if hostname.endswith("amazonaws.com") and len(parts) == 6:
             return f"{parts[0]}.{parts[2]}"
+        # AWS China regions use the amazonaws.com.cn endpoint suffix
+        # e.g. my-cluster.id.cn-north-1.redshift.amazonaws.com.cn (7 parts vs 6 for global)
+        if hostname.endswith("amazonaws.com.cn") and len(parts) == 7:
+            return f"{parts[0]}.{parts[2]}"
         self.log.debug(
             """Could not parse identifier from hostname '%s'.
             You are probably using IP to connect to Redshift cluster.
diff --git a/providers/amazon/tests/unit/amazon/aws/hooks/test_redshift_sql.py b/providers/amazon/tests/unit/amazon/aws/hooks/test_redshift_sql.py
index c8aaa5ac938f6..80a5877e4952c 100644
--- a/providers/amazon/tests/unit/amazon/aws/hooks/test_redshift_sql.py
+++ b/providers/amazon/tests/unit/amazon/aws/hooks/test_redshift_sql.py
@@ -297,6 +297,18 @@ def test_get_conn_iam_does_not_mutate_connection(self, mock_connect, mock_aws_ho
                 {},
                 "1.2.3.4",
             ),
+            # test with AWS China region endpoint (provisioned cluster)
+            (
+                "cluster_identifier_from_host.id.cn-north-1.redshift.amazonaws.com.cn",
+                {"iam": True},
+                "cluster_identifier_from_host.cn-north-1",
+            ),
+            # test with AWS China region endpoint (serverless)
+            (
+                "workgroup-name.account-id.cn-northwest-1.redshift-serverless.amazonaws.com.cn",
+                {"iam": True},
+                "workgroup-name.cn-northwest-1",
+            ),
         ],
     )
     def test_get_openlineage_redshift_authority_part(