This repository complements my blog post on my Mathematics PhD research workflow.

Pdf reference:

Url reference:

First install dependencies by running

sudo apt install xclip exiftool xdotool x11-utils zenity xsel

Then clone this repository into a directory.

cd directory
git clone https://github.com/gillescastel/instant-reference
cd instant-reference
npm install

Then add this directory to your PATH by adding the following line to the bottom of your ~/.profile file:

export PATH="/home/username/path/to/directory/instant-reference:$PATH"

You might need to log out and log in for this to work.

Install the phd protocol by creating a file located at ~/.local/share/applications/phd.desktop containing:

[Desktop Entry]
Name=Phd helper
Exec=phd-protocol-handler.js %u
Icon=emacs-icon
Type=Application
Terminal=false
MimeType=x-scheme-handler/phd;

Then run the following to register it

sudo update-desktop-database
xdg-mime default phd.desktop x-scheme-handler/phd

In config.js located in instant-reference, change the pdfViewer to the one you use (in Ubuntu, the default is evince).

Install a userscripts manager add-on for your browser, for example Tampermonkey (Chrome, Firefox). Then add the following userscript:

// ==UserScript==
// @name         ArXiv add to library
// @version      0.1
// @description  Add a button to ArXiv to add a paper to your library
// @match        https://arxiv.org/*

// ==/UserScript==

(function() {
    'use strict';
    const download = document.querySelector('.download-pdf').href
    const title = document.querySelector('.title').textContent.replace(/^Title:/, '')
    const authors = [...document.querySelectorAll('.authors a')].map(a => a.textContent)
    const query = new URLSearchParams({title, download, authors});
    const url = 'phd://download-arxiv/download?' + query
    const list = document.querySelector('.extra-services .full-text ul')
    const li = document.createElement('li')
    const a = document.createElement('a')
    a.target = '_blank'
    a.innerHTML = 'Add to library'
    a.href = url
    li.appendChild(a)
    list.appendChild(li)
})();

So this simply adds a link to ArXiv with url phd://download-arxiv/download?title=Title&authors=&Authors&download=pdfUrl. When you click on it the custom protocol handler takes over and downloads the file.

Change download location by editing config.js (located in the instant-reference directory) and changing the value of papersDirectory. Be sure to create the directory.

If you get errors using Firefox, try this. Go to about:config and add the following settings:

network.protocol-handler.app.phd	/home/username/path/to/directory/phd-protocol-handler.js	
network.protocol-handler.expose.phd	false	
network.protocol-handler.external.phd	true

Then once you click a phd:// link it will ask you to how to open these kind of links. Specify the protocol handler located at /home/username/path/to/directory/phd-protocol-handler.js and click on 'Don't ask next time'.

Using your preferred shortcut manager (E.g. Settings > Keyboard Shortcuts in Ubuntu), add a shortcut that executes /home/username/path/to/directory/copy-reference.js.

When you press this shortcut, the script will copy a bit of LaTeX code for you to paste in your editor. By adding the following definition of \pdfref to your preamble, the copied LaTeX snippet will transform in a clickable link, and upon clicking on it, the custom protocol handler will open the document at the correct page.

\usepackage{hyperref}
\hypersetup{hidelinks}
\usepackage{xifthen}
\usepackage{fontawesome}

\newcommand\urlref[2]{%
    \href{#1}{\raisebox{0.15ex}{\scriptsize \faLink}\:\textup{\textbf{#2}}}%
}

\newcommand\pdfref[3]{%
    \href{phd://open-paper?id=#1&page=#2}{%
    \textup{[\textbf{\ifthenelse{\isempty{#3}}{here}{#3}}]}}%
}

The first argument of \pdfref is the id of the document, the second the page number and the third the title, which is by default the title of the pdf, which you can change by running

exiftool -overwrite_original_in_place -Title="New title" document.pdf

Zathura is supported out of the box.

Evince is also supported, but beware of the following. First, you need to have gvfs installed and running (the filesystem metadata framework Evince uses to save current page number.) This is by default on Ubuntu.

The snap version of Ubuntu has some problems with talking to gvfs, so this doesn't work. To fix this, remove it and install a deb-based version:

sudo snap remove evince
sudo apt install evince

You'll probably also need to disable App Armor for Evince to allow it to talk to gvfs. Do this by running:

sudo apt install apparmor-utils
sudo aa-complain /usr/bin/evince

Interested in adding support for other pdf viewers? Feel free to add a pull request. Have a look at get-current-pdf-page.js to get started.

To debug the protocol handler, run the following and have a look at the output. This should download 'Ricci flow with surgery on three-manifolds' to your preferred download location.

./phd-protocol-handler.js "phd://download-arxiv/test?title=Ricci+flow+with+surgery+on+three-manifolds&download=https%3A%2F%2Farxiv.org%2Fpdf%2Fmath%2F0303109&authors=Grisha+Perelman"

For the reference shortcut, run

sleep 2; ./copy-reference.js

Within two second, move activate your pdf reader and have a look at the output.