PDF to Text Extractor

What "extracting text" really does

A PDF does not store paragraphs. It stores drawing instructions such as "show this string at position x, y in this font". This tool uses Mozilla's pdf.js to call getTextContent() on each page, which returns those positioned strings in the order they appear in the file. The tool rebuilds lines from the position of each string: when pdf.js flags the end of a line, or the next string sits at a different vertical position, a line break is inserted. Strings on the same line are joined with a space only when there is a visible gap between them, so words split by kerning are not torn apart.

It does not try to detect paragraphs or columns, so each line on the page becomes one line of text. A paragraph that is hard-wrapped in the PDF stays wrapped in the output.

Page separators

With "Insert page separators" ticked, each page's text is preceded by --- Page N --- on its own line. Pages are separated by a blank line, and the whole result is trimmed. Unticked, pages are simply separated by blank lines. The counter below the box shows the page count and character count.

Worked example

Take a two-page PDF where page 1 holds "Invoice 1042" and "Total: 250.00" on two separate lines, and page 2 holds "Thank you". With separators on, you get:

--- Page 1 ---
Invoice 1042
Total: 250.00

--- Page 2 ---
Thank you

When extraction returns nothing or garbage

• Scanned documents. A scan is a picture. With no text layer in the file, the tool reports that no text was found and that it does not perform OCR. A search-and-copy test in a PDF viewer tells you whether a layer exists.
• Columns and tables. Strings come out in file order, which is usually reading order but can interleave columns or put table cells in odd sequences. Cells in the same table row are separated by a space, not a tab.
• Broken font mappings. Some PDFs embed fonts without a proper Unicode map, so copying produces symbols or stray characters even though the page looks correct. The text layer, not the page's looks, is what counts.
• Ligatures and hyphens. Words split by an end-of-line hyphen stay split, and some fonts store "fi" as a single glyph that may come out unexpectedly.
• Right-to-left scripts may arrive in visual rather than logical order.
• Encrypted PDFs cannot be read.

Once you have the text, tidy it with Remove Line Breaks or measure it with the word counter.