[javier]

Companion example for the blog post "One table, three readers: QuestDB, Polars and DuckDB on the same Parquet" (questdb/questdb.io#3049).

• QuestDB 10.1.0 open source in Docker, with the data directory mounted at ./qdb_root so the Parquet partition files can be read from the host
• walkthrough.py loads 1.5M synthetic FX trades from a Polars DataFrame, builds a one-minute materialized view, converts three days to Parquet, and checks QuestDB SQL, Polars and DuckDB against the DataFrame that was sent
• Covers late writes into a Parquet partition, dedup replays and a COPY ... WITH FORMAT PARQUET export
• Only needs Docker and Python, with pinned dependencies in requirements.txt

Not to be merged until the post is ready to publish.