From 47de54170c2e140d9c788bead699a68660e59dc0 Mon Sep 17 00:00:00 2001
From: Niklas Schmidtmer <niklas@crate.io>
Date: Fri, 16 Aug 2024 17:37:36 +0200
Subject: [PATCH] DynamoDB: Wrap all values in UPDATE statements in quotes

---
 CHANGES.md                              | 1 +
 src/commons_codec/transform/dynamodb.py | 6 ++++--
 tests/transform/test_dynamodb.py        | 4 +++-
 3 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/CHANGES.md b/CHANGES.md
index 11283d8..d08ef28 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -1,6 +1,7 @@
 # Changelog
 
 ## Unreleased
+- DynamoDB: Fixed a syntax issue with `text` data type in `UPDATE` statements
 
 ## 2024/08/16 v0.0.6
 - Changed `UPDATE` statements from DMS not to write the entire `data`
diff --git a/src/commons_codec/transform/dynamodb.py b/src/commons_codec/transform/dynamodb.py
index 5fbde2a..bc439e2 100644
--- a/src/commons_codec/transform/dynamodb.py
+++ b/src/commons_codec/transform/dynamodb.py
@@ -139,14 +139,16 @@ def values_to_update(self, keys: t.Dict[str, t.Dict[str, str]]) -> str:
         {'humidity': {'N': '84.84'}, 'temperature': {'N': '55.66'}}
 
         OUT:
-        data['humidity] = 84.84, temperature = 55.66
+        data['humidity] = '84.84', temperature = '55.66'
         """
         values_clause = self.deserialize_item(keys)
 
         constraints: t.List[str] = []
         for key_name, key_value in values_clause.items():
-            constraint = f"{self.DATA_COLUMN}['{key_name}'] = {key_value}"
+            key_value = str(key_value).replace("'", "''")
+            constraint = f"{self.DATA_COLUMN}['{key_name}'] = '{key_value}'"
             constraints.append(constraint)
+
         return ", ".join(constraints)
 
     def keys_to_where(self, keys: t.Dict[str, t.Dict[str, str]]) -> str:
diff --git a/tests/transform/test_dynamodb.py b/tests/transform/test_dynamodb.py
index 4cad2a8..e9154f7 100644
--- a/tests/transform/test_dynamodb.py
+++ b/tests/transform/test_dynamodb.py
@@ -68,12 +68,14 @@
             "humidity": {"N": "84.84"},
             "temperature": {"N": "55.66"},
             "device": {"S": "bar"},
+            "location": {"S": "Sydney"},
             "timestamp": {"S": "2024-07-12T01:17:42"},
         },
         "OldImage": {
             "humidity": {"N": "84.84"},
             "temperature": {"N": "42.42"},
             "device": {"S": "foo"},
+            "location": {"S": "Sydney"},
             "timestamp": {"S": "2024-07-12T01:17:42"},
         },
         "SizeBytes": 161,
@@ -145,7 +147,7 @@ def test_decode_cdc_insert_nested():
 def test_decode_cdc_modify():
     assert (
         DynamoCDCTranslatorCrateDB(table_name="foo").to_sql(MSG_MODIFY) == 'UPDATE "foo" '
-        "SET data['humidity'] = 84.84, data['temperature'] = 55.66 "
+        "SET data['humidity'] = '84.84', data['temperature'] = '55.66', data['location'] = 'Sydney' "
         "WHERE data['device'] = 'foo' AND data['timestamp'] = '2024-07-12T01:17:42';"
     )