Guides

Task-oriented walkthroughs for specific behaviors and use cases. For class/method signatures, see the API Reference.

• CSV Engines & Rust Acceleration — Choosing between Polars, DuckDB, and PyArrow; the native Rust core.
• Handling Messy CSV — Fail-fast validation, lenient mode, comments, blank lines, legacy \r files, column normalization.
• Excel Worksheets & Options — Selecting sheets, exporting every sheet, Polars read-option passthrough.
• Parquet Options — Polars read/write option passthrough for Parquet.
• PDF Parsing — The typed document model, element rows, schema-driven extraction (rationalize), Markdown and SQL, writing outputs and bundles, hidden text, OCR routes, and page statuses.
• Querying with DuckDB — Running SQL directly against CSV, Excel, and Parquet files.
• Polars to Pandas — Every path from a Datagrunt reader to pandas (or any Arrow-native tool).
• Concurrency & Environments — PDF page workers (no __main__ guard needed), memory-capped render workers, thread safety, and distributed runtimes.
• Migrating to 5.0 (PDF) — Every 4.x PDF option and output mapped to its 5.0 form.