PDF File Generator

A document is built from a format description: a catalog, a page tree, a font, and a stream with text. Readers open it, its text can be extracted, and the number of pages is set separately from the file size.

Pages and size are separate settings

This matters for testing. Document processors usually limit either the size or the number of pages, and the two must not be confused: a 300-page document can weigh less than a single page holding a photo.

Padding goes inside, not after the end

Extra bytes go into a comment inside the page stream. Many generators append junk after the end of the file because that is easier, but strict readers treat this as corruption. Here the offsets in the cross-reference table are recalculated, so the document stays valid.