Why some PDFs produce little or no extracted text

Test whether text exists in the document

A scanned page can contain pixels of printed letters without a text layer. A browser text extractor cannot recover words from those pixels by simply reading PDF text objects. PDFLeafy’s text workflow is not an OCR service. If you can neither select words in a PDF viewer nor obtain meaningful extraction, an image-only scan is a likely explanation, though fonts and document structure can also affect results.

Reading order is not always visual order

Columns, positioned labels and tables can be stored in an order that differs from what a reader sees. Extract one representative page and compare paragraphs, headings and table rows. Do not treat the resulting plain text as a faithful spreadsheet. When a column is missing or lines are mixed, keep the PDF beside the output and reconstruct only the portion you can verify.

Page image → no automatic OCR here
Text objects → extraction may need reading-order review
Table layout → plain text does not preserve cell structure

Choose the next step consciously

For a scan requiring search, use an OCR workflow you trust and review its errors, especially numbers and similar-looking letters. That separate workflow may have different upload and privacy behavior. PDFLeafy does not silently send a failed extraction to another service. Keep the original page as evidence when the extracted text will support a consequential decision, and verify every critical value against the visible source.