feat: Add batch replayer (#227)

* Add BatchReplayer and PartitionedDataset classes

BatchReplayer has the same function as the current data-generator. However, it adds the missing file header and use Step Function + Lambda instead of other Athena. This is more cost-effective, but it requires data to be pre-partitioned with a manifest file.

* Restore projen tasks

* feat: Split lambda functions into 2 steps and write output files in paraellel

* fix: Fix bug that Map index isn't passed to Lambda correctly

* feat: write dataframe into multiple files that doesn't exceed given max size

* fix: remove custom workspace setting

* fix: main merge

Co-authored-by: Vincent Gromakowski <gromav@amazon.com>

Author: ijemmy

core/.projen/deps.json

@@ -22,7 +22,7 @@ const project = new AwsCdkConstructLibrary({
     'analytics',
   ],
 
-  cdkVersion: '1.130',
+  cdkVersion: '1.134',
   defaultReleaseBranch: 'main',
   license: 'MIT-0',
   name: 'aws-analytics-reference-architecture',