Editing page structure without an upload
This tool does two jobs that are really the same job. Deleting a page means leaving it out of the output; moving a page means emitting it at a different position. Both are expressed as a single ordered list of source pages, which is then copied into a fresh document with pdf-lib.
The previews come from pdf.js, the same rendering engine Firefox uses to display PDFs. Every page is rasterised in your browser at thumbnail resolution so you can see what you are about to remove. Neither library needs a server, so the document stays on your device throughout.
Why previews matter more than they sound
Tools that ask you to type page numbers work fine when you already know the document. They fail badly the moment you do not — deleting the blank separator pages from a 90-page scan, or pulling the appendix out of a report someone else assembled, means opening the file in another program first and writing numbers down.
Rendering the pages costs a few seconds up front and removes that entire step. It also removes a class of mistake: you cannot accidentally delete page 34 when you meant 43 if you are looking at both.
What rebuilding a PDF carries across
Page content copies faithfully. Text, vector graphics, embedded images and annotations attached to a page all survive the move into the new document, at their original quality — nothing is re-rendered or re-compressed.
Two document-level structures do not survive. Bookmarks live in an outline tree hanging off the document catalog, and interactive form fields live in an AcroForm dictionary in the same place. Neither is attached to the pages being copied, so neither comes along. The tool inspects your file first and warns you when it contains either, rather than letting you discover it after the fact.
Common uses
- Stripping blank pages and separator sheets out of a bulk scan
- Pulling a single signed section out of a long contract
- Moving an appendix from the middle of a report to the end
- Removing pages containing information you do not want to share
- Reversing the order of a document scanned back to front
That last case is worth flagging: if a duplex scanner produced your pages in the wrong order, reordering here is lossless and takes seconds, where re-scanning does not.