Open Contact Book is a buiness-oriented, cross-platform, Python Qt application allowing to manage databases of contacts through CardDAV.

CardDAV is an open protocol for client/server access and sharing of contacts. It is based on .vcf files that are essentially text files.

It is supported by a large amount of server applications, to host your contact books on the internet and share them between devices, like Google Agenda, OwnCloud/NextCloud, or even CPanel.

It is also supported by a large amount of client applications, DAVx⁵ on Android, Blackberry OS 10, Apple iOS and Contacts, Thunderbird (through CardBook extension, and more recently natively) on Windows, Linux and Mac, Evolution on Linux, etc.

However, most CardDAV application are really limited when it comes to managing a large number of contacts, or just to merge duplicates. They also have nothing to backup changes and prevent file corruption, nor to manage geolocation. They became really slow as the number of entries increases.

The spreadsheet view is a read/write table of all the .vcf contacts files found in the address book repository. It allows fast access for bulk-editing and exposes some of the technicality of the Vcard format for the sake of fast manual access.

TODO: the changes made to the spreadsheet are not actually saved for now. This is just a view.

The map view uses Nominatim and OpenStreetMap databases to geolocate your contacts on a map, in a completely de-googled, privacy-compliant, way.

Where are your contacts located ? Turn your address book into knowledge to plan for efficient touring or clients meetings.

An advanced text parsing tries to find hints of the accurate location, using a spellcheck on the country names and various combinations of addresses parts until it finds a match. The geolocation data is cached on your disk and will run faster the next time.

TODO: display a sum-up of the contact info with preview/display modes.

TODO: fetch contacts whose tags and data match some rules.

TODO: merge duplicate contacts with rules, merge columns and refactor your custom Vcard field.

TODO: include git commit and versionning built-in, to track history and revert any change in case of issues or corrupted data.

All the data processing can run headless, without a GUI. The GUI Qt code and the data processing are actually fully separated.

The database of contacts is actually a usual pandas.DataFrame. This allows fast data handling because Python only acts as a binding to lower-level data routines, while retaining all the power and flexbility of Python objects. It also enables to use the usual data-mining and maching-learning libraries.

TODO: access the internal pandas.DataFrame directly from a terminal widget, to apply advanced regex parsing etc.

For now, only local repositories of .vcf files are supported. It is tested with Vcard 3.0 created with Thunderbird CardBook plugin. It is tested against a database of 9000 contacts.

Install the Python interpreter and PIP

Dowload it here.

Run the script install.sh at the root of the code directory:

$ sh install.sh

The code can't be installed system-wise yet, as this work is still in a very early stage.

Start it from the terminal:

$ python main.py