I need working code that reliably extracts plain-text fields directly from one or more databases I will provide. The job is purely text-based scraping—no images or video content are involved—and the source is the database layer itself, not public websites or third-party APIs. Structure, size, and engine details of each database (MySQL and one SQLite dump) will be shared after kickoff along with specific tables and columns to target. I’m happy for you to choose the most efficient stack (Python + SQLAlchemy, Node.js, Go, etc.); what matters is that the final script can be run from the command line, produces a clean UTF-8 CSV or JSON file, and logs any records it skips so nothing is lost unnoticed. There’s no hard deadline, so you can take the time you need to write clear, well-documented code and a concise README explaining setup, dependencies, and usage examples. Once the script runs end-to-end on my sample database and the extracted text matches the original records, the project will be considered complete.