Filozy

Guide

How much of a PDF can actually be edited, and why

A PDF stores positioned glyphs rather than paragraphs, so changing a word does not reflow the line around it and may be impossible if the font was subset. Placing new objects over the page is the reliable form of editing because it adds content instead of rewriting it.

Tools used: Edit PDFWatermark PDFDelete PDF pages

“Edit PDF” is one phrase covering two very different operations, and the confusion between them is the source of most frustration with PDF editing.

One of them is genuinely hard. The other is straightforward. Knowing which you need saves a great deal of time.

PDF was never an editing format

PostScript and PDF were designed to describe a finished page — to guarantee that a document looks identical everywhere, forever. That guarantee is achieved by throwing away everything that could be recomputed differently.

A word processor file knows it contains a paragraph with a style, made of sentences, made of words. A PDF knows: draw glyph “T” at x=72, y=700 in font F1 at 11pt. Then glyph “h” at x=78.4. And so on.

There is no paragraph. There is no sentence, no word. There is a list of positioned marks that happens to look like prose.

What that means for changing text

Replace the word “may” with “must” and nothing reflows, because nothing in the file records that those glyphs belong to a line that belongs to a paragraph. The new word is longer, so it overlaps whatever comes after it. A tool can try to shuffle the following glyphs along — and now the line is too long for the column, with no column recorded anywhere to notice.

Then there is font subsetting. To keep files small, a PDF typically embeds only the glyphs the document actually uses. A document with no accented characters carries no accented glyph outlines. Type “é” and there is nothing to draw — you get a blank, or a substituted character in a different face.

This is why in-place text editing works acceptably on short, same-length changes and degrades quickly past that. It is not a limitation of any particular tool; it is the format doing what it was designed to do.

If a document needs substantial rewriting, edit the source and re-export. Every hour spent fighting a PDF is an hour that would have been ten minutes in the original.

The reliable kind of editing

The other operation is adding content on top — and here the format cooperates completely.

A page’s content stream is an ordered list of drawing operations. Appending more operations is exactly what the format expects. New text, a signature, a highlight, a shape: all of these are simply more marks, drawn after the existing ones.

Nothing underneath has to be understood, reflowed or recomputed. This is why it is dependable, and why almost everything people actually need from a PDF editor falls into this category:

Edit PDF works this way: text, signatures, highlights, drawings and shapes are placed directly on the page, and every object stays movable, resizable, rotatable and editable until you export. Until that export, you are arranging objects over the page rather than altering it — so nothing is committed while you are still deciding.

Filling a form that is not a form

A large share of PDFs that look like forms have no form fields at all. They are scans, or documents exported without interactive fields, and clicking a line does nothing.

Placing text is the answer, and a few habits make the result look intentional rather than pasted on:

The thing that is not editing

There is a persistent and genuinely dangerous belief that covering text with a black rectangle removes it.

It does not. The rectangle is a drawing operation added after the text. The text is still there, still selectable, still extractable — straight through the rectangle. Select the “redacted” area, copy, paste: the words appear.

It has happened in real court filings. In January 2019 the blacked-out passages of a filing by Paul Manafort’s lawyers could be read by copying the text and pasting it into another document (Columbia Journalism Review). How to redact a PDF properly covers that case and others.

Covering is not removal. If content must be gone:

Selecting and searching shows what a viewer draws, not everything the file contains. If you can still find the text, it is there; if you cannot, that is not proof it is gone.

Marking status rather than changing content

To signal that a document is a draft, superseded or not for circulation, the right instrument is a mark across the page rather than an edit to it. Watermark PDF stamps text diagonally across every page into a separate copy.

It communicates to an honest reader and does not pretend to be a control — the mark is ordinary page content and can be removed by anyone willing to edit the file.

Before you send it

Questions people ask

Why can I not simply retype a word in a PDF?

Because the file has no sentence to retype. Text is stored as individual glyphs placed at coordinates, with no record of which ones form a paragraph. Replacing a short word with a longer one does not push the following words along, because nothing in the file says they belong together. The new text overlaps whatever sits to its right.

Why does typing a character sometimes produce a blank or a wrong letter?

Because of font subsetting. To keep files small, a PDF usually embeds only the glyphs the document actually uses. If no accented character appeared anywhere in the original, its outline is simply not in the file, and asking for it produces a gap or a substitute. This is why heavy text editing usually requires going back to the source document.

Is placing text on top of a PDF a real edit?

It is a real, permanent change to the saved file — the new object becomes part of the page content like anything else. What it is not is a modification of the original text underneath, which remains in the file. That distinction matters when the intent is to conceal something rather than to add something.

Can I cover sensitive text with a black rectangle?

You can, and you should not rely on it. The rectangle is drawn over the text, not instead of it. The words remain in the file, and anyone can select or extract them straight through the covering shape. If content must be gone, make a new PDF from the source document without it.

Related guides