I have disabled the sick review section, it was full of automated spam.
Release 0.3.0 of the Breton-French dictionary prepared by Denis Arnaud
is ready for download.
The DICT db can be found at
This is to announce the availability of five new databases, with thanks
to Jim O'Regan, the Apertium project and its Google Code-in students.
The download links are below. Online query is available at http://freedict.org/dict
Piotr & Michael
As of today, all the databases bar two (eng-hun, hun-eng) have been converted to TEI P5. Work can now start on tightening the schemas and adding all kinds of fancy functionality.
The 2nd TEI-P5-encoded dictionary has been released. In contrast to the Swahili-English dictionary, this one is on the one hand merely a glossary (with extremely simple entry structure) and on the other a phrase book (with some phrases or even sentences as headwords). It contains over 20 thousand entries. It is an example of (nearly) the lowest level of TEI-P5 markup conformance for FreeDict dictionaries.
Version 0.4.2, 17 April 2009. Changes by Piotr Banski.
This release contains primarily modifications of the underlying markup, in order to find a balance between TEI-P5 markup and the features of the FreeDict TEI-to-DICT converter. This is the first FreeDict release of a TEI-P5 dictionary with an accompanying DICT database -- previously, FreeDict scripts were not able to handle TEI P5 XML.
See the header or the release notes for more details.
26 March 09: Kęstutis Biliūnas uploaded new versions of the Lithuanian-English (1050 new headwords) and the English-Lithuanian (270 new headwords) dictionaries.
Initial versions of gle<->pol dictionaries by Jimmy O'Regan are now available as TEI P4 and .dict. They have around 280 entries each; created on 2007-09-20.
Swahili-English xFried/FreeDict Dictionary
Version 0.4 by Piotr Bański and Beata Wójtowicz (15-feb-2009) has
more explicit XML structure, more entries, more equivalents, more
cross-entry references (e.g. between all singular-plural pairs).
It contains 2612 entries, 2629 headwords, 5027 tokens in definitions.
It has been described in a paper entitled "A repository of free
lexical resources for African languages: the project and the method"
by Piotr Bański and Beata Wójtowicz, accepted for the AfLaT-2009
(Language Technologies for African Languages) workshop, part of
the 12th meeting of the European Association for Computational
Linguistics (EACL), see http://www.aflat.org/ .... read more
We have recently done quite a few maintenance releases of FreeDict dictionaries:
* English - Portuguese (0.2.1)
* Serbo-Croat - English (0.1.1)
* Hungarian - English (0.2.1)
* Irish - English (0.1.1)
* English - Irish (0.3)
Additionally, an Irish - Polish dictionary has been submitted to the CVS and is waiting for little fixes before it gets released.
The FAQ has grown as well. See
The Arabeyes-derived Arabic-English dictionary, ver. 0.6.1, has been released. It contains ca. 53,000 words.
The following packages have been released:
* English-Arabic (0.6.1, ca. 87400 headwords)
* English-Lithuanian (0.4, 4200 headwords)
* Lithuanian-English (0.4, 3010 headwords)
* Freedict-Tools (0.3)
The Swahili-English dictionary has been upgraded to TEI P5, revised and extended. It contains only 2060 headwords for now, but that will change :-)
English-Lithuanian and Lithuanian-English dictionaries have been expanded, reaching 4200 and 3010 headwords, respectively.
FreeDict wiki is now operative at
The entire HOWTO has been moved there and is now waiting for annotations and updates. You need to have an SF account to edit it.
DictionaryForMIDs is a free Java application coming with several free dictionaries, some of which are from FreeDict. It runs on cell phones and PDAs that implement the MIDP (Mobile Information Device Profile). See:
Naturally, adding DictionaryForMIDs as a supported platform to FreeDict would be a nice thing. As always, helping hands are welcome. Whether you are skilled or unskilled doesn't matter too much. The required knowledge can be accumulated on the way. It is ready for transmission as ever. The requirement is only for an active motivated receiver.
I am very happy to be able to make this release today after a long time of mostly lone development, occasionally cheered up by a mail from Rafal Mantiuk.
The FreeDict databases have been converted to bedic format, renewing support for the zbedic application, running on the Sharp Zaurus PDA platform.
The toolchain for conversion into bedic format was completely rewritten. The Python script used formerly is replaced by a set of XSLT stylesheets. Newly supported features of the bedic format are examples, cross references and usage domain indications.... read more
Bug #920855 that was open for almost one year is finally fixed now! It involved writing a new import script to lure the data out of the DBase3 source database file.
The new packages for dictd and the source package can be downloaded from here:
Support for BEDic is underway.
Next bug to be squashed is actually a meta-bug: #933179. Any help appreciated!
* more transparent navigation
* available in three languages: English, German and Khasi
* includes the HOWTO that was formerly hidden in the SourceForge DocManager
* links to the FreeDict wiki
Find it at: http://freedict.org
Thanks to the fdicts.com project there is now a huge (150 000 entries) English-Czech dictionary available.
But the fdicts.com website has many more dictionaries that could be imported into FreeDict as well. Your help would be appreciated in this undertaking!
Thanks to Erdal Ronahi there is now a English <-> Turkish dictionary with 35 000 headwords. A reverse index is available, it contains many English words though.
The data is derived from the GPL program gtksozluk2 by Mehmet Ali Vardar <email@example.com>:
The data is at the same time imported at fdicts.com - let's see which database develops further. But I guess theirs because we lack editing capabilities.
Beata Wojtowicz undertook the task to merge the old FreeDict Swahili-English dictionary and the "Fried" Swahili-English dictionary from dict.org. This one has about 1500 entries and part of speech information. The reverse version to answer queries from English to Swahili is also available now.
Finally now we have the first completely non-ASCII index working with dictd! This dictionary employs Devanagari and not just transliteration.
Both preformatted dictd databases (eng->hin, hin->eng) are grouped under the English-Hindi package, because they were generated from the same source, which is an English-Hindi dictionary.
I'm happy being able to release the Croatian-English / English-Croatian dicionaries today. They are the result of the efforts of "dennisne" and the upstream project at http://www.taktikanova.hr/eh/
Also, these dictionaries underwent testing insofar as each headword was looked up and retrieved successfully.
The new english-romanian dictionary is added