Parametersless metadata bugfix (#393)

* Parametersless metadata bugfix * Update changelog * More validation
awslabs · Jan 23, 2024 · 2d7fd40 · 2d7fd40
1 parent 843f32b
commit 2d7fd40
Show file tree

Hide file tree

Showing 4 changed files with 25 additions and 4 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,11 @@
 # Change Log
 
+## v0.65
+
+- [#393](https://github.com/awslabs/amazon-s3-find-and-forget/issues/393): Fix
+  for a bug affecting Create Data Mapper API with manually created Glue Tables
+  that don't contain SerdeInfo Parameters metadata
+
 ## v0.64
 
 - [#390](https://github.com/awslabs/amazon-s3-find-and-forget/issues/390):

diff --git a/backend/lambdas/data_mappers/handlers.py b/backend/lambdas/data_mappers/handlers.py
@@ -166,9 +166,10 @@ def get_glue_table_location(t):
 
 
 def get_glue_table_format(t):
+    serde_info = t["Table"].get("StorageDescriptor", {}).get("SerdeInfo", {})
     return (
-        t["Table"]["StorageDescriptor"]["SerdeInfo"]["SerializationLibrary"],
-        t["Table"]["StorageDescriptor"]["SerdeInfo"]["Parameters"],
+        serde_info.get("SerializationLibrary", ""),
+        serde_info.get("Parameters", {}),
     )
 
 

diff --git a/templates/template.yaml b/templates/template.yaml
@@ -1,6 +1,6 @@
 AWSTemplateFormatVersion: "2010-09-09"
 Transform: AWS::Serverless-2016-10-31
-Description: Amazon S3 Find and Forget (uksb-1q2j8beb0) (version:v0.64) (tag:main)
+Description: Amazon S3 Find and Forget (uksb-1q2j8beb0) (version:v0.65) (tag:main)
 
 Parameters:
   AccessControlAllowOriginOverride:
@@ -206,7 +206,7 @@ Conditions:
 Mappings:
   Solution:
     Constants:
-      Version: 'v0.64'
+      Version: 'v0.65'
 
 Resources:
   TempBucket:

diff --git a/tests/unit/data_mappers/test_data_mappers.py b/tests/unit/data_mappers/test_data_mappers.py
@@ -559,6 +559,20 @@ def test_it_gets_glue_table_format_info():
     ) == handlers.get_glue_table_format(get_table_stub())
 
 
+def test_it_gets_glue_table_parametersless():
+    table = {
+        "Table": {
+            "StorageDescriptor": {
+                "Location": "s3://bucket/",
+                "InputFormat": "org.apache.hadoop.mapred.TextInputFormat",
+                "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat",
+            }
+        }
+    }
+
+    assert ("", {}) == handlers.get_glue_table_format(table)
+
+
 @patch("backend.lambdas.data_mappers.handlers.glue_client")
 def test_it_gets_details_for_table(mock_client):
     mock_client.get_table.return_value = get_table_stub()