[DAE-63] Handling exception when adding duplicate partitions

docs/source/getstarted.md

@@ -36,5 +36,6 @@ Beyond the default methods, this library also implements the methods below in
 the [`HiveMetastoreClient`](https://github.com/quintoandar/hive-metastore-client/blob/main/hive_metastore_client/hive_metastore_client.py) class:
 
 - [`add_columns_to_table`](https://hive-metastore-client.readthedocs.io/en/latest/hive_metastore_client.html#hive_metastore_client.hive_metastore_client.HiveMetastoreClient.add_columns_to_table)
-- [`add_partitions_to_table`](https://hive-metastore-client.readthedocs.io/en/latest/hive_metastore_client.html#hive_metastore_client.hive_metastore_client.HiveMetastoreClient.add_partitions_to_table)
+- [`drop_columns_from_table`](https://hive-metastore-client.readthedocs.io/en/latest/hive_metastore_client.html#hive_metastore_client.hive_metastore_client.HiveMetastoreClient.drop_columns_from_table)
+- [`add_partitions_if_not_exists`](https://hive-metastore-client.readthedocs.io/en/latest/hive_metastore_client.html#hive_metastore_client.hive_metastore_client.HiveMetastoreClient.add_partitions_if_not_exists)
 - [`create_database_if_not_exists`](https://hive-metastore-client.readthedocs.io/en/latest/hive_metastore_client.html#hive_metastore_client.hive_metastore_client.HiveMetastoreClient.create_database_if_not_exists)

examples/add_partitions.py

@@ -20,4 +20,4 @@
 
 with HiveMetastoreClient(HIVE_HOST, HIVE_PORT) as hive_client:
     # Adding two set of partitions to specified table
-    hive_client.add_partitions_to_table(DATABASE_NAME, TABLE_NAME, partition_list)
+    hive_client.add_partitions_if_not_exists(DATABASE_NAME, TABLE_NAME, partition_list)

hive_metastore_client/hive_metastore_client.py

@@ -113,16 +113,24 @@ def drop_columns_from_table(
             # call alter table to drop columns removed from list of table columns
             self.alter_table(dbname=db_name, tbl_name=table_name, new_tbl=table)
 
-    def add_partitions_to_table(
+    def add_partitions_if_not_exists(
         self, db_name: str, table_name: str, partition_list: List[Partition]
     ) -> None:
         """
-        Add partitions to a table.
+        Add partitions to a table if it does not exist.
+
+        If the user tries to add a partition twice, the method handles the
+         AlreadyExistsException, not raising an error.
 
         :param db_name: database name where the table is at
         :param table_name: table name which the partitions belong to
         :param partition_list: list of partitions to be added to the table
         """
+        if not partition_list:
+            raise ValueError(
+                "m=add_partitions_if_not_exists, msg=The partition list is empty."
+            )
+
         table = self.get_table(dbname=db_name, tbl_name=table_name)
 
         partition_list_with_correct_location = self._format_partitions_location(
@@ -131,7 +139,10 @@ def add_partitions_to_table(
             table_partition_keys=table.partitionKeys,
         )
 
-        self.add_partitions(partition_list_with_correct_location)
+        try:
+            self.add_partitions(partition_list_with_correct_location)
+        except AlreadyExistsException:
+            pass
 
     def create_database_if_not_exists(self, database: Database) -> None:
         """
@@ -197,4 +208,7 @@ def _validate_lists_length(list_a: List[Any], list_b: List[Any]) -> None:
         :param list_b: second list to be compared
         """
         if len(list_a) != len(list_b):
-            raise ValueError("The length of the two provided lists does not match")
+            raise ValueError(
+                "m=_validate_lists_length, msg=The length of the two provided "
+                "lists does not match"
+            )

tests/unit/hive_metastore_client/test_hive_metastore_client.py

@@ -1,7 +1,8 @@
 from unittest import mock
-from unittest.mock import Mock
+from unittest.mock import Mock, ANY
 
 import pytest
+from pytest import raises
 
 from hive_metastore_client import HiveMetastoreClient
 from thrift_files.libraries.thrift_hive_metastore_client.ThriftHiveMetastore import (
@@ -164,16 +165,53 @@ def test_add_partitions_to_table(
         # arrange
         db_name = "database_name"
         table_name = "table_name"
-        formatted_partitions_location = ["abc"]
+
+        mocked_table = Mock()
+        mocked_get_table.return_value = mocked_table
+
         mocked_partition_list = [Mock()]
+        formatted_partitions_location = ["abc"]
+        mocked__format_partitions.return_value = formatted_partitions_location
+
+        # act
+        hive_metastore_client.add_partitions_if_not_exists(
+            db_name=db_name, table_name=table_name, partition_list=mocked_partition_list
+        )
+
+        # assert
+        mocked_get_table.assert_called_once_with(dbname=db_name, tbl_name=table_name)
+        mocked__format_partitions.assert_called_once_with(
+            partition_list=mocked_partition_list,
+            table_storage_descriptor=mocked_table.sd,
+            table_partition_keys=mocked_table.partitionKeys,
+        )
+        mocked_add_partitions.assert_called_once_with(formatted_partitions_location)
+
+    @mock.patch.object(HiveMetastoreClient, "get_table")
+    @mock.patch.object(HiveMetastoreClient, "_format_partitions_location")
+    @mock.patch.object(HiveMetastoreClient, "add_partitions")
+    def test_add_partitions_to_table_with_duplicated_partitions(
+        self,
+        mocked_add_partitions,
+        mocked__format_partitions,
+        mocked_get_table,
+        hive_metastore_client,
+    ):
+        # arrange
+        db_name = "database_name"
+        table_name = "table_name"
+
         mocked_table = Mock()
-        mocked_table.sd = ""
-        mocked_table.partitionKeys = ""
         mocked_get_table.return_value = mocked_table
+
+        mocked_partition_list = [Mock()]
+        formatted_partitions_location = ["abc"]
         mocked__format_partitions.return_value = formatted_partitions_location
 
+        mocked_add_partitions.side_effect = AlreadyExistsException()
+
         # act
-        hive_metastore_client.add_partitions_to_table(
+        hive_metastore_client.add_partitions_if_not_exists(
             db_name=db_name, table_name=table_name, partition_list=mocked_partition_list
         )
 
@@ -186,6 +224,26 @@ def test_add_partitions_to_table(
         )
         mocked_add_partitions.assert_called_once_with(formatted_partitions_location)
 
+    @mock.patch.object(HiveMetastoreClient, "get_table")
+    @mock.patch.object(HiveMetastoreClient, "_format_partitions_location")
+    @mock.patch.object(HiveMetastoreClient, "add_partitions")
+    def test_add_partitions_to_table_with_invalid_partitions(
+        self,
+        mocked_add_partitions,
+        mocked__format_partitions,
+        mocked_get_table,
+        hive_metastore_client,
+    ):
+        # assert
+        with raises(ValueError):
+            # act
+            hive_metastore_client.add_partitions_if_not_exists(
+                db_name=ANY, table_name=ANY, partition_list=[]
+            )
+        mocked_get_table.assert_not_called()
+        mocked__format_partitions.assert_not_called()
+        mocked_add_partitions.assert_not_called()
+
     @mock.patch.object(HiveMetastoreClient, "_validate_lists_length")
     def test_format_partitions_location(
         self, mocked_validate_lists_length, hive_metastore_client