Implemented some of the Minor comments in apacheGH-34616

anjakefala · Sep 14, 2023 · 7e1b24c · 7e1b24c
1 parent 65e1e54
commit 7e1b24c
Show file tree

Hide file tree

Showing 4 changed files with 11 additions and 7 deletions.
diff --git a/cpp/src/arrow/dataset/file_parquet.h b/cpp/src/arrow/dataset/file_parquet.h
@@ -241,7 +241,7 @@ class ARROW_DS_EXPORT ParquetFileWriteOptions : public FileWriteOptions {
   /// \brief Parquet Arrow writer properties.
   std::shared_ptr<parquet::ArrowWriterProperties> arrow_writer_properties;
 
-  // A configuration structure that provides per file encryption properties for a dataset
+  // A configuration structure that provides encryption properties for a dataset
   std::shared_ptr<ParquetEncryptionConfig> parquet_encryption_config = NULLPTR;
 
  protected:

diff --git a/cpp/src/arrow/dataset/file_parquet_encryption_test.cc b/cpp/src/arrow/dataset/file_parquet_encryption_test.cc
@@ -35,7 +35,7 @@ constexpr std::string_view kFooterKeyMasterKey = "0123456789012345";
 constexpr std::string_view kFooterKeyMasterKeyId = "footer_key";
 constexpr std::string_view kFooterKeyName = "footer_key";
 constexpr std::string_view kColumnMasterKey = "1234567890123450";
-constexpr std::string_view kColumnMasterKeysId = "col_key";
+constexpr std::string_view kColumnMasterKeyId = "col_key";
 constexpr std::string_view kColumnKeyMapping = "col_key: a";
 constexpr std::string_view kBaseDir = "";
 
@@ -78,7 +78,7 @@ class DatasetEncryptionTest : public ::testing::Test {
 
     // Prepare encryption properties.
     std::unordered_map<std::string, std::string> key_map;
-    key_map.emplace(kColumnMasterKeysId, kColumnMasterKey);
+    key_map.emplace(kColumnMasterKeyId, kColumnMasterKey);
     key_map.emplace(kFooterKeyMasterKeyId, kFooterKeyMasterKey);
 
     crypto_factory = std::make_shared<parquet::encryption::CryptoFactory>();
@@ -177,6 +177,8 @@ TEST_F(DatasetEncryptionTest, WriteReadDatasetWithEncryption) {
 
   // Verify the data was read correctly
   ASSERT_OK_AND_ASSIGN(auto combined_table, read_table->CombineChunks());
+  // Validate the table
+  ASSERT_OK(combined_table->ValidateFull());
   AssertTablesEqual(*combined_table, *table);
 }
 

diff --git a/python/examples/dataset/write_dataset_encrypted.py b/python/examples/dataset/write_dataset_encrypted.py
@@ -15,14 +15,14 @@
 # specific language governing permissions and limitations
 # under the License.
 
+import shutil
+import os
+from datetime import timedelta
 
 import pyarrow as pa
 import pyarrow.dataset as ds
 import pyarrow.parquet.encryption as pe
 from pyarrow.tests.parquet.encryption import InMemoryKmsClient
-from datetime import timedelta
-import shutil
-import os
 
 """ A sample to demonstrate parquet dataset encryption and decryption"""
 

diff --git a/python/pyarrow/_dataset_parquet.pyx b/python/pyarrow/_dataset_parquet.pyx
@@ -77,7 +77,9 @@ IF PARQUET_ENCRYPTION_ENABLED:
         kms_connection_config : KmsConnectionConfig
             Configuration for connecting to Key Management Service.
         encryption_config : EncryptionConfiguration
-            Configuration for encryption settings.
+            :ref:`Configure <encryption-configuration>`_ e.g. which columns to encrypt,
+            length of encryption keys, Parquet Encryption Algorithm
+            and more.
 
         Raises
         ------