Can I convert a PDF with tables to EPUB?

Why tables are hard

A PDF does not store a table. It stores pieces of text at positions and, sometimes, lines drawn around them. A converter has to guess which pieces form a row, which form a column and where each cell starts, and a cell can contain several lines. Getting that wrong scrambles numbers across rows, which is worse than no table at all.

What EpubCC does today

• The text of the table is kept, so nothing disappears.
• The cells are read as text, column by column, not as an HTML table.
• Bold header cells can be taken for headings.
• Running text around the table, footnotes, chapters and headers are handled as usual.

What to do with a table-heavy PDF

• Run the free sample on a page with your hardest table and look at it in the side-by-side preview.
• If the book is mostly prose with a few tables, the EPUB is usually still worth it: the text reads well and the tables can be looked up in the PDF.
• If the tables are the content (financial reports, data appendices, timetables), keep the PDF. Wide tables do not fit a 6-inch screen even in a well-made EPUB.
• Read the quality report of the sample: it lists the problems the engine found on those pages, and the supported PDFs page lists what we have and have not tested.