Skip to content

Latest commit

 

History

History
80 lines (59 loc) · 3.65 KB

README.md

File metadata and controls

80 lines (59 loc) · 3.65 KB

pdf-scribe.nvim

A Neovim plugin for importing annotations and metadata from PDFs

Dependencies

You will need LuaJIT (if you're using this with Neovim, it's baked in), poppler-glib, GObject, and GLib installed on your machine.

If you're using Neovim, it needs to be v0.4.3+ (I think - might need to be later).

Installation

Once you have the dependencies installed, use your preferred Neovim package manager. With vim-packager, this looks like:

call packager#add('wbthomason/pdf-scribe.nvim')

Usage

pdf-scribe is (a) a (mostly) Neovim-agnostic LuaJIT library for extracting information (title, author, keywords, external links) and (highlight, underline, and pop-up) annotations from PDFs, and (b) a Neovim plugin serving as an example use of the library/a hopefully good out-of-the-box experience for general users.

For (a), the best documentation is the code. The short version is that the module pdfscribe provides a function load_pdf(pdf_file_path) which returns a PDF object for a given PDF file path. The PDF object has methods to extract the aforementioned information and annotations from its PDF file. There are also convenience functions provided by the pdfscribe module to get just annotations from a PDF (get_annotations(pdf_file_path)) and to get all info from a file (get_all_info(pdf_file_path)).

You can use the pdfscribe library to write your own PDF manipulation plugins.

(b) is designed to cover the most common imagined use case for pdfscribe: creating and updating plain-text notes on PDF files. See the docs for complete information. The short version is that there are two commands provided: :PdfScribeInit to create a notes file and :PdfScribeUpdateNotes to re-extract annotations and update the file. Both commands use configurable templates with primitive conditional value substitution to generate files in whatever format you prefer.

Example Configuration

The only values you must configure are g:pdfscribe_pdf_dir and g:pdfscribe_notes_dir, which set the directory in which pdfscribe looks for your PDF files and for your notes files, respectively. See the docs for more complete information.

Contributing

PRs and issues (bugs and feature requests) are welcome!

Notes

  • The core PDF library only depends on Neovim for checking if files exist. The rest of it is portably reusable in any LuaJIT environment.
  • As the LuaJIT module is a wrapper around a C library, bugs may cause segfaults. Bug reports for segfault bugs are particularly welcomed.
  • This plugin may eat your files, lunch, and/or car.

Known issues

  • Poppler doesn't always agree with the highlight rectangles created by other PDF engines; this can lead to slightly off text selection extraction
  • Highlights including mathematical symbols/other non-ASCII Unicode usually work, but sometimes get chopped up weirdly due to how Poppler determines glyph inclusion in selection regions/maps symbols to UTF-8 codepoints.

TODO

Help is welcome with any of the below! Otherwise, I'll get to them as I have time

  • Reimplement template engine in Lua, make easier to use
  • Reimplement note initialization/update in Lua
  • Refactor pdfscribe module to separate interface and implementation
  • Allow easier customization of note sorting
  • Allow easier customization of note grouping
  • Allow multiple PDF directories/search paths
  • Allow multiple note directories/search paths
  • Add support for ink annotations
  • Add more formatting options: line wrapping, indent matching, etc.