PDF to text

Reading the glyphs a PDF contains, with their coordinates, and emitting them as a string. A PDF does not store paragraphs — the paragraph you see is something your eye assembles — so extraction returns the file’s internal order, which is usually but not always reading order.

• It is labelled as text, not as a document

  Recovering columns, tables and headings means inferring structure that was never recorded. Every free tool claiming to do it returns a flat run of text with the tables destroyed. This gives you that same text and is honest about what it is.

• Runs in the tab, on any size of document

  Nothing is uploaded and there is no page cap, so pulling the words out of a 300-page report is the same operation as pulling them out of a receipt — the only difference is how long your own processor takes.

1. 1

   Run it and read the first page of output

   That tells you immediately whether the document has a text layer and whether its internal order matches reading order.

2. 2

   Expect interleaving in multi-column layouts

   Two columns often interleave, and text in headers, footers or floating boxes can appear anywhere in the stream.

3. 3

   If nothing comes out, the file is a scan

   An image of a page with no text layer. Getting words out of that needs OCR, which is a different and much heavier operation.

• How to extract text from a PDF, and why some PDFs give you nothingPull the text out of a PDF in your browser, and why a scan gives nothing: it has no text layer, and this tool does no OCR. Columns may interleave.

• No OCR. A scanned page has no text layer to extract, and recognising glyphs from pixels is a different operation with a much larger download.
• No Word output. The open-source route produces a document with every paragraph on its own line and the tables gone, which is worse than useless for anything anybody would actually convert.