Most of the documentation can be found in the wiki. This page is mostly an overview over the available resources. A good starting point is the section on frequently asked questions. If you only need help on installing a program to use our dictionaries, the downloads page should help you getting started. If you are stuck at any point, you might want to ask for help.
To develop an importer or to build a dictionary from source, it is necessary that you get a fresh copy of the FreeDict tools, e.g. with Git:
$ git clone https://github.com/freedict/tools
The README details the installation in more detail and the building process is detailled in the HOWTO. In summary you need Python >= 3.4 and (GNU) make and xsltproc or tei2slob (depending on the output format). If you are on windows, we do advise that you take a look at the Msys2 project.
Once the tools directory is in place, you should set the environment variable
FREEDICT_TOOLS to its absolute path.\
Note: Due to limitations in Make, we do not support paths with whitespace.
As son as all this has been set up, you can build a dictionary by changing to
its directory and type
Creating And Importing Dictionaries
A lengthy introduction into FreeDict internals, including documentation on how to format and process dictionaries can be found in the FreeDict HOWTO.
Most of the dictionaries have not been written by hand, but where important from other free sources. The import centre gives an overview about which dictionaries have been imported with which program and is a good starting point for new importers.
Static information about our dictionaries can be retrieved over a CORS-enabled API, both in XML and JSON format. It can be used to retrieve download links and other meta information useful for embedding them into a dictionary program. The URLs are http://freedict.org/freedict-database.xml http://freedict.org/freedict-database.json
The individual fields are documented here.
New Dictionary Output Formats
We are always trying to reach more users by exporting into more output formats and our limiting factor is solely the number of contributors. There is a list of clients and ideally we would support them all, one day.