Tablizer is a free converter for tabular data. It reads 18 formats, writes 39, and has a page for each of the 562 conversions that makes. Between the two sits an editable grid, so you can fix a header or drop a stray column without going back to the source file.
How it works
Everything happens in your browser. When you paste or drop a file, JavaScript on this page parses it into one internal table shape: a list of headers and a list of rows. Every output format is generated from that same shape. That is why the behaviour is the same whichever conversion you pick, and why adding a format is one function rather than a matrix of special cases.
The parsing and generating run in a Web Worker, which is a background thread. That is what keeps the page responsive while a large file is being read.
The libraries doing the heavy lifting are Papa Parse for delimited text, SheetJS for Excel workbooks, js-yaml for YAML, fast-xml-parser for XML and jsPDF for PDF output. They are loaded only on demand, so a page you use to convert CSV to Markdown never downloads the Excel reader.
What it does not do
- It does not read scanned PDFs. A scan is an image of a page, and getting data out of one needs optical character recognition, which is a different job with a different failure rate.
- It does not keep formulas, formatting or charts from a spreadsheet. A table has values. Everything else is a property of the file it came from.
- It does not store your data, which also means it cannot recover it. Reload the page and the table is gone.
Who runs this
Tablizer is built and maintained by Aurangzeb Khan, in Bannu, Pakistan. It is not a company and has no staff; the parsers and serialisers behind every conversion on the site were written by one person, which is also why the behaviour is consistent across all of them.
If a conversion gets something wrong, that is worth hearing about. Mail [email protected] with the input that failed and what you expected instead.
Privacy
Your data is parsed by JavaScript running on this page, so nothing is uploaded and the size of the file is limited only by your own machine. There is no upload endpoint, so there is nothing to breach.