How the Markdown to CSV Tool Works
AnyTool extracts tables using marked's GitHub-Flavored-Markdown lexer rather than a hand-rolled line parser, so it correctly handles tables with or without outer pipes, escaped pipes (\|) inside cells, alignment rows (:--:), ragged rows, and tables nested inside blockquotes or lists. Each cell can be flattened to plain text — dropping inline markdown such as bold, links and inline code while keeping the content — or kept as raw markdown. CSV/TSV is generated with papaparse so quoting and escaping are RFC-4180 correct, and JSON output (array of objects keyed by header) is also available. Extraction runs in real time as you type, every table in the document is found, and multiple tables can be exported together as a zip. Everything runs client-side; the Markdown is never uploaded.
- Runs client-side — Markdown is never uploaded
- GFM lexer (marked): escaped pipes, alignment, loose/outer-pipe tables, nested tables
- Plain-text flattening or keep-markdown cell modes
- RFC-4180 CSV/TSV via papaparse + JSON output
- Real-time extraction of every table; multi-table zip export
- No CDN, no backend
