There are already numerous good PDF libraries for Go, and this one deliberatly takes inspiration from them. However, it is based on a slighty different approach : instead of working with a PDF as a tree of dynamic objects, it starts by modeling the whole specification (at least a good portion of it) with static types: see the package model.
The package model is the corner stone of this library. Then, packages may be divided in two parts:
reader imports a PDF file into memory
fonts provides support to use embeded PDF fonts
contentstream and formfill provides tools to create PDF models
The idea is possibly to provide a complete support of the PDF spec, but more importantly to exposes the different layers (such as parser or content stream operators) so that it can be reusable by other libraries. As such, the first target of this library would be higher levels libraries (such as pdfcpu, gofpdf, oksvg, etc...).
A standard workflow to modify an existing PDF would look like
// load the existing file in memory
fi, _, err := reader.ParsePDFFile(filePath, reader.Options{})
// error handling ...
// process the document model as you wish
err = fi.WriteFile(output, nil)
// error handling ...See decompress and api for more examples.