Writing a PDF Metadata Editor

2026-05-19

What did I make?

PDF files often carry incorrect or incomplete metadata. A paper downloaded from a server might have a title like draft_final.docx. A document exported from another program might have a completely empty author field. For users who rely on a PDF reader to organize a library of documents, this matters.

I recently wrote a tool called etch that deals with this problem. It is command-line only and is packaged as a script that can be downloaded from GitHub and run without installation.

Running etch with just a file path displays the current metadata. Passing --title or --author or both shows a before-and-after preview and asks for confirmation before making any changes.

For a short usage guide, see the readme here.

Implementation details

Rather than reading and writing PDF metadata directly, etch delegates this work to exiftool, a well-established utility created by Phil Harvey that handles metadata for many file formats. etch checks that it is installed at startup and exits with a clear message if it is not.

When writing changes, etch passes -overwrite_original to exiftool. This flag alters exiftool’s default behaviour of keeping a backup of the original file before modifying it. Instead of keeping a backup, I opted to ask the user to confirm their change.

Limitations of this tool include the following:

1. etch can only edit title and author. Other metadata fields are not supported.
2. etch requires exiftool to be installed.

Keeping a curated document library

The quality of PDF metadata varies widely and is largely determined by the tool used to create the file. Unfortunately, most tools do a poor job. Files exported from word processors tend to carry over whatever document title was last saved, and scanned documents often have no metadata at all. This is mostly outside the user’s control.

Correcting PDF metadata by hand is a low-priority task for most people. But for those building a well-organized document library, it is worth doing.