Convert RTF to PDF

A format from before everything was XML

RTF is a plain-text format with formatting written as control words — readable in any text editor, which is exactly why it survived. Microsoft published it in 1987 as an interchange format between word processors, and it is still what legal document systems, older databases and a good deal of medical and government software emit when they need formatted text that anything can open.

What comes across is everything RTF can express: fonts, sizes, weights, colours, paragraph alignment and spacing, lists, tables with their borders and shading, headers and footers, page size and margins, footnotes and hyperlinks. The text stays real selectable text. Because RTF is far simpler than a word processor's native format, this conversion has fewer edge cases than most — there is less in the file to get wrong.

The awkward part is pictures. RTF from the Windows era frequently embeds images as WMF or EMF — vector drawing instructions for an interface that predates everything current — and those need a real renderer. They are converted normally but left out by the fallback converter, which also skips page headers and footers. Ordinary PNG and JPEG pictures come through on both paths. If a converted document is missing a logo or a diagram, this is almost always why.

Character encoding is the other quirk worth knowing about. RTF predates Unicode and encodes non-ASCII characters through code-page escapes, with a Unicode mechanism bolted on later — so a file written by very old software can produce wrong accented characters if it declares its code page incorrectly. Anything written this century is fine.

When you want something else

LibreOffice does the same conversion locally: soffice --headless --convert-to pdf document.rtf sends nothing anywhere, handles a folder in one command, and renders the WMF and EMF pictures that a local fallback cannot. For a system that generates RTF in bulk — case files, reports, letters — that is the right shape of tool.

For converting RTF at scale in a pipeline, Pandoc — pandoc in.rtf -o out.pdf — is fast and scriptable, though it treats RTF as text and formatting rather than as a page layout, so complex documents come out simplified. And if the RTF came from a system you control, check whether it can emit PDF directly; one conversion fewer is always better than a good converter.