Python Data Processing: Merge two messy sources and compute a summary dictionary

Prompt In Python project mode, you receive two data sources from different systems/teams (provided in code files; often as raw strings or lists/dicts). The data contains known issues explicitly listed in the README, such as: Inconsistent event formats / missing fields Invalid/corrupted records (must be skipped; your code must not crash) Orphaned records between the two sources Schema changes causing nulls or renamed fields Implement the function signature specified in the README to: Parse raw data into structured objects. Handle bad data robustly (skip/default/log; but do not crash). Merge the two sources using the required key/linking logic. Compute aggregate metrics grouped by categories/dimensions. Return a clean summary dictionary that directly answers the Section 1 business question. Constraints Handle all edge cases called out in the README. The solution must run and match the expected sample output. This tests real-world data wrangling, code readability, and edge-case handling.