From bf9d779f210f6e45798bbe1c60ee23f0d421ff90 Mon Sep 17 00:00:00 2001
From: Geoff Genz <geoff@clickhouse.com>
Date: Sun, 5 Nov 2023 18:36:16 -0700
Subject: [PATCH] Fix s3 bucket bug

---
 CHANGELOG.md                         |  7 +++++++
 dbt/adapters/clickhouse/impl.py      |  5 ++---
 tests/integration/adapter/test_s3.py | 29 ++++++++++++++++++++++++++++
 3 files changed, 38 insertions(+), 3 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 38e6e945..761c7653 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,10 @@
+### Release [1.5.0], TBD
+#### Improvement
+- Compatible with dbt 1.5.x
+
+#### Bug Fix
+- Fix s3 macro when bucket includes `https://` prefix.  Closes https://github.com/ClickHouse/dbt-clickhouse/issues/192.
+
 ### Release [1.4.9], 2023-10-27
 #### Improvement
 - Lots of work on Distributed table materializations.  Big thanks to [gfunc](https://github.com/gfunc) for the additional PR
diff --git a/dbt/adapters/clickhouse/impl.py b/dbt/adapters/clickhouse/impl.py
index 80497fe9..3c9a5b87 100644
--- a/dbt/adapters/clickhouse/impl.py
+++ b/dbt/adapters/clickhouse/impl.py
@@ -167,13 +167,12 @@ def s3source_clause(
         fmt = fmt or s3config.get('fmt')
         bucket = bucket or s3config.get('bucket', '')
         path = path or s3config.get('path', '')
-        url = bucket
+        url = bucket.replace('https://', '')
         if path:
             if bucket and path and not bucket.endswith('/') and not bucket.startswith('/'):
                 path = f'/{path}'
             url = f'{url}{path}'.replace('//', '/')
-        if not url.startswith('http'):
-            url = f'https://{url}'
+        url = f'https://{url}'
         access = ''
         if aws_access_key_id and not aws_secret_access_key:
             raise DbtRuntimeError('S3 aws_access_key_id specified without aws_secret_access_key')
diff --git a/tests/integration/adapter/test_s3.py b/tests/integration/adapter/test_s3.py
index 8fb8727f..10f1289e 100644
--- a/tests/integration/adapter/test_s3.py
+++ b/tests/integration/adapter/test_s3.py
@@ -27,6 +27,10 @@
 select * from {{ clickhouse_s3source('taxi_s3', path='/trips_4.gz') }} LIMIT 5000
 """
 
+s3_taxis_full_source = """
+select * from {{ clickhouse_s3source('taxi_s3', path='/trips_5.gz') }} LIMIT 1000
+"""
+
 s3_taxis_inc = """
 {{ config(
     materialized='incremental',
@@ -84,3 +88,28 @@ def test_s3_incremental(self, project):
         )
         assert 5000 < result[0] < 10000
         assert result[1] > 0
+
+
+class TestS3Bucket:
+    @pytest.fixture(scope="class")
+    def project_config_update(self):
+        return {
+            'vars': {
+                'taxi_s3': {
+                    'bucket': 'https://datasets-documentation.s3.eu-west-3.amazonaws.com/nyc-taxi/',
+                    'fmt': 'TabSeparatedWithNames',
+                }
+            }
+        }
+
+    @pytest.fixture(scope="class")
+    def models(self):
+        return {
+            "s3_taxis_source.sql": s3_taxis_full_source,
+            "schema.yml": schema_yaml,
+        }
+
+    def test_read(self, project):
+        run_dbt(["run", "--select", "s3_taxis_source.sql"])
+        result = project.run_sql("select count() as num_rows from s3_taxis_source", fetch="one")
+        assert result[0] == 1000