Convert TXT to PDF

Plain text is the format that outlives everything. No version, no proprietary reader, no formatting to break — it will still open in fifty years. That reliability is also its problem the moment a person on the other end expects a document: submission systems ask for PDF, printers want defined pages, and a `.txt` attachment looks like a fragment of something rather than a finished thing.

This conversion gives the text what it is missing. The content is laid out on standard pages with a readable typeface and sensible margins, paragraph breaks and blank lines intact. For notes, a manuscript, a letter or an exported transcript, it is exactly right and there is nothing more to think about.

A text file does not record which encoding it uses — that information simply is not in the file — so any program opening one has to guess. Guess wrong and every accented character comes out as nonsense: a file saved on Windows in the older Western encoding, read as UTF-8, turns é into two meaningless symbols.

If your output has garbled accents, umlauts or quotation marks, that is what happened. The permanent fix is to open the original in any editor and re-save it as UTF-8, which is what everything modern expects. Nothing on this page can guess better than the file allows.

A text file has no page width, and a PDF page does. Lines that run past the margin wrap onto the next line, which is correct for prose and unhelpful for anything structured. A server log, a code listing or a fixed-width report relies on each record occupying one line, and wrapping destroys that at a glance.

The text is set in a proportional typeface as well, so anything aligned with spaces — an ASCII table, columns of figures — will not line up. If your file is structured rather than written, this is the wrong destination for it.

Nothing is compressed, nothing is interpreted, and there is no formatting to lose because there was none to begin with. What goes in as characters comes out as the same characters, on pages.

Windows, macOS and Unix have historically ended lines differently, and a file written on one system and read on another can arrive without its breaks recognised.

All three conventions are handled here, so this should not happen. If it does, the file is probably not plain text at all — a .txt extension on something that is really RTF or a word processor export behaves strangely, and opening it in a text editor shows you immediately.

Standard page size, a readable typeface, sensible margins, and paragraph breaks and blank lines preserved as they were. What you do not get is page numbers. The conversion is one pass through LibreOffice on its default page style, which carries no footer, and nothing runs afterwards to add one — so a printed stack comes back with nothing on it to put the pages in order by. Where that matters, open the text in a word processor, add a footer and export the PDF from there.

There are no headings, because a text file has none to find. Anything that looks like a heading is a line of text, and it is set as one.

A few thousand pages is not a problem in itself, because the work is proportional to the amount of text. What sets the limit is memory, since the document is assembled before it is written out.

An export of several hundred megabytes is better split first. That size is usually also a sign that the text is a log or a dataset rather than a document, in which case a PDF is the wrong destination for it.

Greek, Cyrillic, Devanagari, Japanese, Chinese and Korean are set with the fonts the converter carries, so a UTF-8 file in any of them comes through readable. Devanagari was measured rather than assumed: a Hindi file converted here produces a PDF with `NotoSansDevanagari-Regular` embedded in it, so the characters travel as characters and the matras join as they should.

Right-to-left scripts such as Arabic and Hebrew are the harder case, because correct layout needs more than the right glyphs in the right order. Look at the result before relying on it for anything that matters.

Drop up to 100 files at once and each becomes its own PDF, returned together as a ZIP. They are not joined into one document, because deciding the order would be a decision made on your behalf.

This pair is one of the few that runs on our server rather than in the tab, so the numbers are tighter than elsewhere on the site: 25 MB per file, and every file spends one of the 100 server conversions a day the free plan covers — a full batch is a day’s worth in a single drop.

If a single document is what you want, join the text files first and convert the result. That way the order is one you chose rather than one that fell out of a file listing.