fix: "PostgresMetadataExtractor doesn't discover Redshift late binding views"

databuilder/extractor/base_postgres_metadata_extractor.py

@@ -0,0 +1,116 @@
+# Copyright Contributors to the Amundsen project.
+# SPDX-License-Identifier: Apache-2.0
+
+import abc
+import logging
+from collections import namedtuple
+
+from pyhocon import ConfigFactory, ConfigTree
+from typing import Iterator, Union, Dict, Any
+
+from databuilder import Scoped
+from databuilder.extractor.base_extractor import Extractor
+from databuilder.extractor.sql_alchemy_extractor import SQLAlchemyExtractor
+from databuilder.models.table_metadata import TableMetadata, ColumnMetadata
+from itertools import groupby
+
+
+TableKey = namedtuple('TableKey', ['schema', 'table_name'])
+
+LOGGER = logging.getLogger(__name__)
+
+
+class BasePostgresMetadataExtractor(Extractor):
+    """
+    Extracts Postgres table and column metadata from underlying meta store database using SQLAlchemyExtractor
+    """
+
+    # CONFIG KEYS
+    WHERE_CLAUSE_SUFFIX_KEY = 'where_clause_suffix'
+    CLUSTER_KEY = 'cluster_key'
+    USE_CATALOG_AS_CLUSTER_NAME = 'use_catalog_as_cluster_name'
+    DATABASE_KEY = 'database_key'
+
+    # Default values
+    DEFAULT_CLUSTER_NAME = 'master'
+
+    DEFAULT_CONFIG = ConfigFactory.from_dict(
+        {WHERE_CLAUSE_SUFFIX_KEY: ' ', CLUSTER_KEY: DEFAULT_CLUSTER_NAME, USE_CATALOG_AS_CLUSTER_NAME: True}
+    )
+
+    @abc.abstractmethod
+    def get_sql_statement(self, use_catalog_as_cluster_name: bool, where_clause_suffix: str) -> Any:
+        """
+        :return: Provides a record or None if no more to extract
+        """
+        return None
+
+    def init(self, conf: ConfigTree) -> None:
+        conf = conf.with_fallback(BasePostgresMetadataExtractor.DEFAULT_CONFIG)
+        self._cluster = '{}'.format(conf.get_string(BasePostgresMetadataExtractor.CLUSTER_KEY))
+
+        self._database = conf.get_string(BasePostgresMetadataExtractor.DATABASE_KEY, default='postgres')
+
+        self.sql_stmt = self.get_sql_statement(
+            use_catalog_as_cluster_name=conf.get_bool(BasePostgresMetadataExtractor.USE_CATALOG_AS_CLUSTER_NAME),
+            where_clause_suffix=conf.get_string(BasePostgresMetadataExtractor.WHERE_CLAUSE_SUFFIX_KEY),
+        )
+
+        self._alchemy_extractor = SQLAlchemyExtractor()
+        sql_alch_conf = Scoped.get_scoped_conf(conf, self._alchemy_extractor.get_scope())\
+            .with_fallback(ConfigFactory.from_dict({SQLAlchemyExtractor.EXTRACT_SQL: self.sql_stmt}))
+
+        self.sql_stmt = sql_alch_conf.get_string(SQLAlchemyExtractor.EXTRACT_SQL)
+
+        LOGGER.info('SQL for postgres metadata: {}'.format(self.sql_stmt))
+
+        self._alchemy_extractor.init(sql_alch_conf)
+        self._extract_iter: Union[None, Iterator] = None
+
+    def extract(self) -> Union[TableMetadata, None]:
+        if not self._extract_iter:
+            self._extract_iter = self._get_extract_iter()
+        try:
+            return next(self._extract_iter)
+        except StopIteration:
+            return None
+
+    def _get_extract_iter(self) -> Iterator[TableMetadata]:
+        """
+        Using itertools.groupby and raw level iterator, it groups to table and yields TableMetadata
+        :return:
+        """
+        for key, group in groupby(self._get_raw_extract_iter(), self._get_table_key):
+            columns = []
+
+            for row in group:
+                last_row = row
+                columns.append(ColumnMetadata(row['col_name'], row['col_description'],
+                                              row['col_type'], row['col_sort_order']))
+
+            yield TableMetadata(self._database, last_row['cluster'],
+                                last_row['schema'],
+                                last_row['name'],
+                                last_row['description'],
+                                columns)
+
+    def _get_raw_extract_iter(self) -> Iterator[Dict[str, Any]]:
+        """
+        Provides iterator of result row from SQLAlchemy extractor
+        :return:
+        """
+        row = self._alchemy_extractor.extract()
+        while row:
+            yield row
+            row = self._alchemy_extractor.extract()
+
+    def _get_table_key(self, row: Dict[str, Any]) -> Union[TableKey, None]:
+        """
+        Table key consists of schema and table name
+        :param row:
+        :return:
+        """
+        if row:
+            return TableKey(schema=row['schema'], table_name=row['name'])
+
+        return None

databuilder/extractor/postgres_metadata_extractor.py

@@ -1,132 +1,43 @@
 # Copyright Contributors to the Amundsen project.
 # SPDX-License-Identifier: Apache-2.0
 
-import logging
-from collections import namedtuple
+from pyhocon import ConfigFactory, ConfigTree  # noqa: F401
+from typing import Iterator, Union, Dict, Any  # noqa: F401
 
-from pyhocon import ConfigFactory, ConfigTree
-from typing import Iterator, Union, Dict, Any
+from databuilder.extractor.base_postgres_metadata_extractor import BasePostgresMetadataExtractor
 
-from databuilder import Scoped
-from databuilder.extractor.base_extractor import Extractor
-from databuilder.extractor.sql_alchemy_extractor import SQLAlchemyExtractor
-from databuilder.models.table_metadata import TableMetadata, ColumnMetadata
-from itertools import groupby
 
-
-TableKey = namedtuple('TableKey', ['schema', 'table_name'])
-
-LOGGER = logging.getLogger(__name__)
-
-
-class PostgresMetadataExtractor(Extractor):
+class PostgresMetadataExtractor(BasePostgresMetadataExtractor):
     """
     Extracts Postgres table and column metadata from underlying meta store database using SQLAlchemyExtractor
     """
-    # SELECT statement from postgres information_schema to extract table and column metadata
-    SQL_STATEMENT = """
-    SELECT
-      {cluster_source} as cluster, c.table_schema as schema, c.table_name as name, pgtd.description as description
-      ,c.column_name as col_name, c.data_type as col_type
-      , pgcd.description as col_description, ordinal_position as col_sort_order
-    FROM INFORMATION_SCHEMA.COLUMNS c
-    INNER JOIN
-      pg_catalog.pg_statio_all_tables as st on c.table_schema=st.schemaname and c.table_name=st.relname
-    LEFT JOIN
-      pg_catalog.pg_description pgcd on pgcd.objoid=st.relid and pgcd.objsubid=c.ordinal_position
-    LEFT JOIN
-      pg_catalog.pg_description pgtd on pgtd.objoid=st.relid and pgtd.objsubid=0
-    {where_clause_suffix}
-    ORDER by cluster, schema, name, col_sort_order ;
-    """
-
-    # CONFIG KEYS
-    WHERE_CLAUSE_SUFFIX_KEY = 'where_clause_suffix'
-    CLUSTER_KEY = 'cluster_key'
-    USE_CATALOG_AS_CLUSTER_NAME = 'use_catalog_as_cluster_name'
-    DATABASE_KEY = 'database_key'
 
-    # Default values
-    DEFAULT_CLUSTER_NAME = 'master'
-
-    DEFAULT_CONFIG = ConfigFactory.from_dict(
-        {WHERE_CLAUSE_SUFFIX_KEY: ' ', CLUSTER_KEY: DEFAULT_CLUSTER_NAME, USE_CATALOG_AS_CLUSTER_NAME: True}
-    )
-
-    def init(self, conf: ConfigTree) -> None:
-        conf = conf.with_fallback(PostgresMetadataExtractor.DEFAULT_CONFIG)
-        self._cluster = '{}'.format(conf.get_string(PostgresMetadataExtractor.CLUSTER_KEY))
-
-        if conf.get_bool(PostgresMetadataExtractor.USE_CATALOG_AS_CLUSTER_NAME):
+    def get_sql_statement(self, use_catalog_as_cluster_name, where_clause_suffix):
+        # type: (bool, str) -> str
+        if use_catalog_as_cluster_name:
             cluster_source = "c.table_catalog"
         else:
             cluster_source = "'{}'".format(self._cluster)
 
-        self._database = conf.get_string(PostgresMetadataExtractor.DATABASE_KEY, default='postgres')
-
-        self.sql_stmt = PostgresMetadataExtractor.SQL_STATEMENT.format(
-            where_clause_suffix=conf.get_string(PostgresMetadataExtractor.WHERE_CLAUSE_SUFFIX_KEY),
-            cluster_source=cluster_source
+        return """
+        SELECT
+          {cluster_source} as cluster, c.table_schema as schema, c.table_name as name, pgtd.description as description
+          ,c.column_name as col_name, c.data_type as col_type
+          , pgcd.description as col_description, ordinal_position as col_sort_order
+        FROM INFORMATION_SCHEMA.COLUMNS c
+        INNER JOIN
+          pg_catalog.pg_statio_all_tables as st on c.table_schema=st.schemaname and c.table_name=st.relname
+        LEFT JOIN
+          pg_catalog.pg_description pgcd on pgcd.objoid=st.relid and pgcd.objsubid=c.ordinal_position
+        LEFT JOIN
+          pg_catalog.pg_description pgtd on pgtd.objoid=st.relid and pgtd.objsubid=0
+        {where_clause_suffix}
+        ORDER by cluster, schema, name, col_sort_order ;
+        """.format(
+            cluster_source=cluster_source,
+            where_clause_suffix=where_clause_suffix,
         )
 
-        self._alchemy_extractor = SQLAlchemyExtractor()
-        sql_alch_conf = Scoped.get_scoped_conf(conf, self._alchemy_extractor.get_scope())\
-            .with_fallback(ConfigFactory.from_dict({SQLAlchemyExtractor.EXTRACT_SQL: self.sql_stmt}))
-
-        self.sql_stmt = sql_alch_conf.get_string(SQLAlchemyExtractor.EXTRACT_SQL)
-
-        LOGGER.info('SQL for postgres metadata: {}'.format(self.sql_stmt))
-
-        self._alchemy_extractor.init(sql_alch_conf)
-        self._extract_iter: Union[None, Iterator] = None
-
-    def extract(self) -> Union[TableMetadata, None]:
-        if not self._extract_iter:
-            self._extract_iter = self._get_extract_iter()
-        try:
-            return next(self._extract_iter)
-        except StopIteration:
-            return None
-
-    def get_scope(self) -> str:
+    def get_scope(self):
+        # type: () -> str
         return 'extractor.postgres_metadata'
-
-    def _get_extract_iter(self) -> Iterator[TableMetadata]:
-        """
-        Using itertools.groupby and raw level iterator, it groups to table and yields TableMetadata
-        :return:
-        """
-        for key, group in groupby(self._get_raw_extract_iter(), self._get_table_key):
-            columns = []
-
-            for row in group:
-                last_row = row
-                columns.append(ColumnMetadata(row['col_name'], row['col_description'],
-                                              row['col_type'], row['col_sort_order']))
-
-            yield TableMetadata(self._database, last_row['cluster'],
-                                last_row['schema'],
-                                last_row['name'],
-                                last_row['description'],
-                                columns)
-
-    def _get_raw_extract_iter(self) -> Iterator[Dict[str, Any]]:
-        """
-        Provides iterator of result row from SQLAlchemy extractor
-        :return:
-        """
-        row = self._alchemy_extractor.extract()
-        while row:
-            yield row
-            row = self._alchemy_extractor.extract()
-
-    def _get_table_key(self, row: Dict[str, Any]) -> Union[TableKey, None]:
-        """
-        Table key consists of schema and table name
-        :param row:
-        :return:
-        """
-        if row:
-            return TableKey(schema=row['schema'], table_name=row['name'])
-
-        return None

databuilder/extractor/redshift_metadata_extractor.py

@@ -0,0 +1,71 @@
+# Copyright Contributors to the Amundsen project.
+# SPDX-License-Identifier: Apache-2.0
+
+from pyhocon import ConfigFactory, ConfigTree  # noqa: F401
+from typing import Iterator, Union, Dict, Any  # noqa: F401
+
+from databuilder.extractor.base_postgres_metadata_extractor import BasePostgresMetadataExtractor
+
+
+class RedshiftMetadataExtractor(BasePostgresMetadataExtractor):
+    """
+    Extracts Redshift table and column metadata from underlying meta store database using SQLAlchemyExtractor
+
+    This differs from the PostgresMetadataExtractor because in order to support Redshift's late binding views,
+    we need to join the INFORMATION_SCHEMA data against the function PG_GET_LATE_BINDING_VIEW_COLS().
+    """
+
+    def get_sql_statement(self, use_catalog_as_cluster_name, where_clause_suffix):
+        # type: (bool, str) -> str
+        if use_catalog_as_cluster_name:
+            cluster_source = "CURRENT_DATABASE()"
+        else:
+            cluster_source = "'{}'".format(self._cluster)
+
+        return """
+        SELECT
+            *
+        FROM (
+            SELECT
+              {cluster_source} as cluster,
+              c.table_schema as schema,
+              c.table_name as name,
+              pgtd.description as description,
+              c.column_name as col_name,
+              c.data_type as col_type,
+              pgcd.description as col_description,
+              ordinal_position as col_sort_order
+            FROM INFORMATION_SCHEMA.COLUMNS c
+            INNER JOIN
+              pg_catalog.pg_statio_all_tables as st on c.table_schema=st.schemaname and c.table_name=st.relname
+            LEFT JOIN
+              pg_catalog.pg_description pgcd on pgcd.objoid=st.relid and pgcd.objsubid=c.ordinal_position
+            LEFT JOIN
+              pg_catalog.pg_description pgtd on pgtd.objoid=st.relid and pgtd.objsubid=0
+
+            UNION
+
+            SELECT
+              {cluster_source} as cluster,
+              view_schema as schema,
+              view_name as name,
+              NULL as description,
+              column_name as col_name,
+              data_type as col_type,
+              NULL as col_description,
+              ordinal_position as col_sort_order
+            FROM
+                PG_GET_LATE_BINDING_VIEW_COLS()
+                    COLS(view_schema NAME, view_name NAME, column_name NAME, data_type VARCHAR, ordinal_position INT)
+        )
+
+        {where_clause_suffix}
+        ORDER by cluster, schema, name, col_sort_order ;
+        """.format(
+            cluster_source=cluster_source,
+            where_clause_suffix=where_clause_suffix,
+        )
+
+    def get_scope(self):
+        # type: () -> str
+        return 'extractor.redshift_metadata'