Merge pull request #702 from jaceksan/working

gooddata-dbt: Add support for Redshift

gooddata-dbt/gooddata_dbt/dbt/profiles.py

@@ -15,6 +15,7 @@
     CatalogDataSourceVertica,
     MotherDuckAttributes,
     PostgresAttributes,
+    RedshiftAttributes,
     SnowflakeAttributes,
     TokenCredentialsFromEnvVar,
     VerticaAttributes,
@@ -54,6 +55,37 @@ def to_gooddata(self, data_source_id: str, schema_name: str) -> CatalogDataSourc
         )
 
 
+@attrs.define(auto_attribs=True, kw_only=True)
+class DbtOutputRedshift(Base):
+    name: str
+    title: str
+    host: str
+    port: str
+    user: str
+    password: str = attrs.field(repr=lambda value: "***")
+    dbname: str
+    database: str = attrs.field(default=attrs.Factory(lambda self: self.dbname, takes_self=True))
+    schema: str
+
+    def to_gooddata(self, data_source_id: str, schema_name: str) -> CatalogDataSourcePostgres:
+        return CatalogDataSourcePostgres(
+            id=data_source_id,
+            name=self.title,
+            db_specific_attributes=RedshiftAttributes(
+                host=self.host,
+                port=self.port,
+                # TODO - adopt this in Python SDK
+                db_name=quote_plus(self.dbname),
+            ),
+            # Schema name is collected from dbt manifest from relevant tables
+            schema=schema_name,
+            credentials=BasicCredentials(
+                username=self.user,
+                password=self.password,
+            ),
+        )
+
+
 @attrs.define(auto_attribs=True, kw_only=True)
 class DbtOutputSnowflake(Base):
     name: str
@@ -152,7 +184,7 @@ def to_gooddata(self, data_source_id: str, schema_name: str) -> CatalogDataSourc
         )
 
 
-DbtOutput = Union[DbtOutputPostgreSQL, DbtOutputSnowflake, DbtOutputVertica, DbtOutputMotherDuck]
+DbtOutput = Union[DbtOutputPostgreSQL, DbtOutputRedshift, DbtOutputSnowflake, DbtOutputVertica, DbtOutputMotherDuck]
 
 
 @attrs.define(auto_attribs=True, kw_only=True)
@@ -189,16 +221,21 @@ def inject_env_vars(output_def: Dict) -> None:
             # else do nothing, real value seems to be stored in dbt profile
 
     @staticmethod
-    def to_data_class(output: str, output_def: Dict) -> DbtOutput:
+    def to_data_class(output: str, output_def: Dict) -> Optional[DbtOutput]:
         db_type = output_def["type"]
         if db_type == "postgres":
             return DbtOutputPostgreSQL.from_dict({"name": output, **output_def})
+        elif db_type == "redshift":
+            return DbtOutputRedshift.from_dict({"name": output, **output_def})
         elif db_type == "snowflake":
             return DbtOutputSnowflake.from_dict({"name": output, **output_def})
         elif db_type == "vertica":
             return DbtOutputVertica.from_dict({"name": output, **output_def})
-        elif db_type == "duckdb":
+        elif db_type == "duckdb" and output_def["path"].startswith("md:"):
             return DbtOutputMotherDuck.from_dict({"name": output, **output_def})
+        elif db_type == "duckdb":
+            # No logging available here. Pass because GoodData cannot connect to DuckDB file.
+            return None
         else:
             raise Exception(f"Unsupported database type {output=} {db_type=}")
 
@@ -210,7 +247,8 @@ def profiles(self) -> List[DbtProfile]:
             for output, output_def in profile_def["outputs"].items():
                 self.inject_env_vars(output_def)
                 dbt_output = self.to_data_class(output, output_def)
-                outputs.append(dbt_output)
+                if dbt_output:
+                    outputs.append(dbt_output)
             profiles.append(DbtProfile(name=profile, outputs=outputs))
         return profiles