Readcube-Overleaf integration: Adds an "Update Library" button to Overleaf that allows you to import your Readcube library.

readleaf.user.js

// ==UserScript==
// @name         Readcube-Overleaf integration
// @namespace    https://tillahoffmann.github.io/
// @version      0.1
// @description  Adds an "Update Library" button to Overleaf that allows you to import your Readcube library.
// @author       Till Hoffmann
// @match        https://www.overleaf.com/*
// @connect      readcube.com
// @grant        GM_xmlhttpRequest
// ==/UserScript==

function formatPageNumbers(x) {
    // Remove spaces
    x = x.replaceAll(/\s/g, '');
    // Replace single dashes with double dashes
    x = x.replaceAll(/(?<=\d)-(?=\d)/g, '--');
    return x;
}

const formattingLookup = {
    'title': 'article/title',
    'journal': 'article/journal',
    'pages': {
        'path': 'article/pagination',
        'format': formatPageNumbers,
    },
    'volume': 'article/volume',
    'year': 'article/year',
    'doi': 'ext_ids/doi',
};

function formatItem(item, usedKeys) {
    var citekey = item.user_data.citekey;
    // Cite key is available
    if (citekey) {
        citekey = citekey.replaceAll("'", "");
    }
    // Generate a cite key based on the author
    else if (item.article.authors.length) {
        var author = item.article.authors[0].split(/\s+/);
        // Get the last name of the author (naively, anyway)
        citekey = author[author.length - 1];
        // Add the year if it's available
        if (item.article.year) {
            citekey = citekey + item.article.year;
        }
        // Add a suffix if it's not unique
        if (usedKeys[citekey]) {
            usedKeys[citekey] += 1;
            citekey += String.fromCharCode(95 + usedKeys[citekey]);
        } else {
            usedKeys[citekey] = 1;
        }
    }
    // Just generate something random
    else {
        var randomInt = Math.floor(Math.random() * 0xffffffff);
        citekey = randomInt.toString(16);
    }
    var lines = [
        '@' + item.item_type + '{' + citekey + ',',
        '    author = {' + (item.article.authors || []).join(' and ') + '},',
    ];
    for (var [key, value] of Object.entries(formattingLookup)) {
        if (typeof value === "string") {
            value = {'path': value};
        }
        var x = item;
        for (var subpath of value.path.split('/')) {
            x = x[subpath];
            if (x === undefined) {
                break;
            }
        }
        if (x) {
            if (value.format) {
                x = value.format(x);
            }
            lines.push('    ' + key + ' = {' + x + '},');
        }
    }
    lines.push('}');
    return lines.join('\n');
}

function fetchItems(config) {
    // Construct the url
    var url = config.baseUrl;
    if (config.scrollId) {
        url = url + '?scroll_id=' + config.scrollId;
    }

    GM_xmlhttpRequest({
        method: "GET",
        url: url,
        onload: function(response) {
            unsafeWindow.config = config;
            var data = JSON.parse(response.responseText);
            if (data.items.length > 0) {
                config.items = (config.items || []).concat(data.items);
                // Continue recursively
                fetchItems(Object.assign({}, config, {scrollId: data.scroll_id}));
            } else if (config.callback) {
                config.callback(config);
            }
        }
    });
}


function updateLibrary() {
    // Get the first editor and fetch the code
    var editor = null;
    for (let element of document.getElementsByClassName('ace_editor')) {
        editor = ace.edit(element);
        break;
    }
    var code = editor.getValue();

    // Try to match the URL pattern and load the items
    var pattern = /%% https?:\/\/app.readcube.com\/library\/([\w-]+)\/list\/([\w-]+)/gi;
    for (var match of code.matchAll(pattern)) {
        console.log('library: ', match[1]);
        console.log('list: ', match[2]);

        var url = 'https://sync.readcube.com/collections/' + match[1] + '/lists/' + match[2] + '/items';
        console.log('fetching ' + url);

        fetchItems({
            'baseUrl': url,
            'match': match,
            'editor': editor,
            'callback': function(config) {
                // Format the code and add it to the editor
                console.log(config);
                // Sort the items for consistent cite key generation
                config.items.sort(function(a, b) {
                    if (a.id < b.id) {
                        return -1;
                    }
                    return 1;
                });

                var lines = [config.match[0]];
                var usedKeys = {};
                for (var item of config.items) {
                    lines.push(formatItem(item, usedKeys));
                }
                editor.setValue(lines.join('\n\n'));
            },
        });

        // We only process one library (for now anyway)
        break;
    }
}

(function() {
    'use strict';

    // Inject the update button in the toolbar
    setInterval(function() {
        var parent = document.querySelector('.toolbar-editor .toolbar-right');
        if (parent.getAttribute('readcube')) {
            return;
        }
        var child = document.createElement('a');
        child.innerText = 'Update Library';
        child.onclick = function() {
            updateLibrary();
        }
        parent.insertBefore(child, parent.firstChild);
        parent.setAttribute('readcube', 'injected');
    }, 1000);
})();

@jaepillee0315 Wow thank you for the quick response! Unfortunately, I'm still confused. On the webapp, if I click nothing, this is all I get "https://app.readcube.com/" regardless of "Always show full URLs." If I click the "My papers" folder, I get "https://app.readcube.com/library/0a5458f1-92f7-4b8d-91cb-33723708d367/all" and if I click a specific paper, I get "https://app.readcube.com/library/0a5458f1-92f7-4b8d-91cb-33723708d367/all/(sidepanel:details)?item_id=430c97a7-5c3b-4322-a36f-36de89796df9&collection_id=0a5458f1-92f7-4b8d-91cb-33723708d367"

None of the links resemble the ones here. The first number is there, but the number after "list" isn't, and the word list itself too.

If I use any of those links, the file in Overleaf doesn't fill up with anything when I click the Update ReadCube button.

You need to create a list. Click the gear button next to "My Papers," click "Create List," and give a name to the list. Then, assign the papers you want to cite to the list (When you click a paper, you will see there are "TAGS", "LISTS", "CITEKEY", etc. Use "LIST" to add the paper to the list). Then, use the list's URL on Overleaf. Make sure you put double percent sign (%%) in front of the URL. I hope this is useful.

Hi, just to let you all know, I've put together a new version of the code. Now, the imported bib entries can be sorted according to the name of the first author. This is necessary in order to have the multi-citatations with natbib properly sorted in the PDF. I've added a toggle to use this option (set as default) or not which appears after hovering over the import button for 1.5 s. Besides, a minor bug stemming from the Overleaf new text editor behaviour, which prevented re-importing bib entries if scrolled down far enough, is bypassed by automatic scroll-to-the-top in the editor before the import itself.


The revision is available in my repository: https://gist.github.com/MichalPt/0d08085321d03a63aca07720785849c8

Thanks @MichalPt . Nice workaround with the editor.

Another thing I recently added to my version is to rewrite the ReadCube type to the Bibtex type. Unfortunately, they don't always match:

function formatType(x) {
    const types = {
        'conference_paper' : 'inproceedings',
        'poster': 'inproceedings',
        'webpage': 'misc'
    }

    if( x in types ) {
        return types[x]
    } else {
        return x
    }
}

And call formatType(item.item_type)

I've updated my revision again. Changes:

• Merged changes from @MichalPt (scrolling to top, sorting)
• Added the formatType() function from above
• Added a formatRaw() for URL and DOI to prevent from escaping special chars in those fields.