Fixing several things in one PDF, in the right order

One grant application, five separate problems

Tomás is the volunteer treasurer of a community garden. The council's grant portal closes on Thursday at noon, and the file he has to upload, garden_grant_2027.pdf, was stitched together from three people's exports and a few phone scans. It runs to 14 pages and has five problems: the budget page was scanned sideways, page 12 is blank, the project start date on page 2 says March when it should say April, the pages are not numbered, and the contact sheet on page 14 shows his home phone number, which he does not want in a public council file.

No single button fixes all of that, and nothing here needs a desktop editor either. Each problem belongs to a different kind of change, and once they are sorted the work takes a lunch break.

Which tool handles which change

The sequence that avoids doing anything twice

Each PDFGym tool reopens its own output and checks it before the download appears. Download after every step, give the file a short suffix such as _2 or _3, and open that copy in the next tool. Keeping the in-between files means a mistake costs one step, not the whole job.

1. Fill any form fields first, then flatten them when the answers are final. Later steps can drop interactive fields, so they must be settled before anything else.
2. Fix the structure in Organize pages: remove the blank page 12 and drag any misplaced pages into order.
3. Turn the sideways budget page upright in Rotate & crop, applying a 90, 180 or 270 degree turn to the page in view only.
4. Correct the start date in Edit PDF: drag a white-out over "March", choose Add text, and type "April" in Times or Helvetica to match the surrounding words.
5. Remove the home phone number in Redact PDF by drawing a box over it or searching for the digits.
6. Number the pages last with Stamp PDF, once the page count can no longer change: page numbers mode, bottom centre, all pages.
7. If the portal has an upload limit, run the finished file through Compress PDF and type the limit as the target. Sign only after that, because a signature should go on the version that is actually sent.

Why page changes are easy and word changes are not

A PDF does not store paragraphs. Each page is a list of drawing instructions: this glyph at these coordinates, this image in this rectangle. Moving or removing a whole page means copying that list as a unit, which is why it is quick and why nothing on the page degrades.

Changing a word means reaching inside the list. The letters after it do not slide along to make room, the font is often embedded with only the characters the author used, and a scanned page has no letters at all, only pixels. That is why the dependable fix for a word is to remove it and place a replacement, and why PDFGym does not try to reflow lines. The guide to replacing words on a page goes deeper into matching size and alignment.

Edits that backfire, and the repair

• Page numbers skip a number. They were stamped before a page was deleted. Go back to the copy from before the stamp step, delete the page, then stamp again.
• The "removed" phone number can still be copied. A black box drawn in a viewer only hides text. Use Redact PDF, then open the result in the Redaction checker to confirm nothing is left under the boxes.
• The file grew after the date fix. A page with a white-out is rebuilt from a sharp render, which can add weight. Compress at the end rather than in the middle.
• A clickable contents page stopped working. White-outs and redactions remove links on the rebuilt pages. On link-heavy documents, change the source file instead.
• The cropped margin still holds a note. Cropping only hides what is outside the box. If the margin contains something private, redact it.

When a browser editor is the wrong place to start

If whole paragraphs need rewriting, or the layout has to change, ask whoever made the PDF for the Word, Pages or InDesign file. Editing the source and exporting again beats any amount of patching. When the source is gone, converting with PDF to Word gives you something editable, at the cost of some layout.

Two other limits are worth knowing in advance. PDFGym does not run text recognition, so a scanned page stays a picture; you can cover and type over it, but it will not become searchable. And Edit PDF types with standard fonts that cover English and most Western European letters, so characters such as č or ł, or Cyrillic, Arabic and Asian scripts, need a desktop editor or the original file. A document that already carries a certificate signature should go back to the signer, since any edit invalidates it.