KaiwaLens published corpora — licence and attribution
=====================================================

This directory holds prebuilt dictionary databases:

    kaiwa_corpus_<date>.db3.gz    the complete corpus, read by KaiwaLens
    kotoba_corpus_<date>.db3.gz   a trimmed copy, read by Kotoba Battle

Each is a DERIVATIVE WORK of the data listed below, compiled into SQLite.
It is not part of the KaiwaLens application, which is MIT-licensed and
distributed separately, and it carries none of that licence.


LICENCE
-------

These databases are made available under the Creative Commons
Attribution-ShareAlike 4.0 International licence (CC BY-SA 4.0):

    https://creativecommons.org/licenses/by-sa/4.0/

CC BY-SA 4.0 because the share-alike terms of JMdict, KANJIDIC2 and
KanjiVG reach any work derived from them, and a compiled corpus is such a
work. If you redistribute these files, or anything built from them, you
must do so under the same terms and keep the attribution below.


ATTRIBUTION
-----------

JMdict / JMnedict
    Copyright (C) Electronic Dictionary Research and Development Group
    (EDRDG). Licensed under CC BY-SA 4.0.
    https://www.edrdg.org/edrdg/licence.html

KANJIDIC2
    Copyright (C) Electronic Dictionary Research and Development Group
    (EDRDG). Licensed under CC BY-SA 4.0.
    https://www.edrdg.org/edrdg/licence.html

KanjiVG
    Copyright (C) Ulrich Apel. Licensed under CC BY-SA 3.0.
    https://kanjivg.tagaini.net/

Tatoeba example sentences
    Copyright (C) Tatoeba contributors. Licensed under CC BY 2.0 FR.
    https://tatoeba.org/eng/terms_of_use

UniDic
    Copyright (C) The UniDic Consortium, NINJAL. Tri-licensed GPL / LGPL /
    BSD-3-Clause; used here under the BSD-3-Clause arm. The data is read;
    no UniDic code is linked.
    https://clrd.ninjal.ac.jp/unidic/

Japanese WordNet
    Copyright (C) National Institute of Information and Communications
    Technology. See the upstream licence.
    https://bond-lab.github.io/wnja/

JLPT level lists
    Compiled by hand for this project and covered by KaiwaLens's own MIT
    licence, not by any third party's terms.

EDRDG asks that applications using JMdict and KANJIDIC acknowledge them
visibly. KaiwaLens does so on its first-run screen, where the user accepts
these terms before any of this data is downloaded, and again in Settings.


PROVENANCE
----------

Each corpus is compiled from the sources above by the KaiwaLens release
process. The application can compile its own from the same sources, and a
corpus built that way is identical in kind to the ones here — these exist
only so that a user need not download roughly 800 MB of archives and parse
them to arrive at the same database.

Full third-party notices, including the libraries the application links:

    https://github.com/Pugnator/KaiwaLens/blob/master/THIRD-PARTY-NOTICES.md