Hi Sebastian,
FreeDict dictionaries are only partly in a Git, we need to find a
different
solution here. We've got a static API though from which you could
pull:
https://github.com/freedict/fd-dictionaries/wiki/FreeDict-API
I have not looked in detail how are the non-TEI dicts formatted. In
the
This is not how it works :). All dictionaries are in TEI, it is only
that a
few are not in the Git repository, because these are auto-imported.
We use Kosh for our Sanskrit dictionaries encoded in TEI:
https://cceh.github.io/c-salt_sanskrit_data/
What is the licence of this data?
I do not fully understand their encoding, but as far as I
understand,
you have
Sanskrit-German and Sanskrit-English there. Do you think there is
achance we
could import these dictionary into FreeDict?
These dictionaries are of public domain. So you can do whatever you
want with them. The idea of hosting these dictionaries ourselves is
to
have control of the corrections that are made to them as also not
to do
double work. They are all digitized dictionaries, and most of the
errors are typos. We have Sanskrit experts working with us, so the
idea
is to offer curated data. Users can propose modifications via
GitHub
and these are accepted/rejected by our team. Just like with the fd-
dictionaries GitHub repo. While they could be probably more visible
to
some people at the fd-dictionaries repo, I do not see any benefit
of
hosting them twice.
No, certainly not hosting twice. Instead, I would prefer if you could
provide
them in a way that matches our TEI P5 subset, this could be
automated. We
would then regularly import your dictionaries and offer them on our
web site
as .dict and .slob dictionaries, import them into Debian-based
GNU/Linux
distributions and make them available via our static API. Since these
dictionaries are already in TEI, would you mind writing a
transformation
script to the FreeDict subset? Otherwise I would open an issue and
hope for
time to implement it.
