Added ability to add fields to an existing PDF (not used yet)

[BryceStevenWilley]

Address the first part of https://github.com/SuffolkLITLab/docassemble-ALWeaver/issues/299, the "Adding AcroForm automatically".

A few key limitations:

• only works on PDFs without any existing AcroForm fields, since we just wholesale copy over the AcroForm object from a newly created PDF. Discussion on https://github.com/pikepdf/pikepdf/issues/207 is interesting, and we would have to follow what happens in https://github.com/torakiki/sejda/blob/00fee714632069b9e330e551efc30d7c2d25d860/sejda-sambox/src/main/java/org/sejda/impl/sambox/component/AcroFormsMerger.java if we do want to fix this.
• All of the field types aren't there, just made this as a proof of concept and haven't filled it in yet.
• Grouped Fields (useful, but only necessary for Radio Buttons I think?): this just adds each field as a flat array, no nesting with parents or anything yet. Not sure if we want to be able to specify fields in a tree or not.

[BryceStevenWilley]

Got something more complete now: given a PDF, we iterate through all pages, trying to find long-ish horizontal lines that don't run up against any vertical lines. It gives descent results on this PDF, with the except of some aesthetic lines at the bottom that are detected as fields. Bringing in some sort of "label to the right or above" would help with that I think. Probably over fit this one particular example, but we'll see what the other downsides are pretty quickly probably.

The PDF with no fields:

The PDF with added fields:

Also added a bunch of options for other fields that we can make: choice, listboxes, radio, checkbox, etc. Radio and checkbox are likely to be auto-detected, but choices / listboxes need semantic knowledge of things.

Probably ready for a closer look @nonprofittechy, and I'll merge after this. Sorry, should have marked as draft yesterday when it was still a WIP.