Hi,
I'd like to share a small project I've done for last couple of week.
I've building last couple of my projects's backend api in GraphQL (
http://graphql.org/ ), which is a new query language specification started
by Facebook It solved many problem in the modern web application
developing, read their website for more if you're interested.
In the mean time, I'm really interested in using wikidata's structured
data for some of my scripts and toy projects. It seems to me somehow we
could comebine these two to provide an easy to use API for application
users.
So here's the simple demo using GraphQL to querying the wikidata, it's just
a prove of concept certainly not ready for any serious use. The underlying
fetching it's done by using the rest API at *https://www.wikidata.org/w/api.php
<https://www.wikidata.org/w/api.php>*
Demo time: you could write some query to
1. *Get the beatles's all members(has_part) 's award_received and
place_of_birth and twitter_username*
(permalink for this query at
http://bit.ly/2eXTmbG )
{
entity(id: "Q1299") {
label(lang: "en")
has_part {
mainsnak {
label(lang: "en")
award_received {
mainsnak {
label(lang: "en")
}
}
place_of_birth {
mainsnak {
label(lang: "en")
}
}
twitter_username {
mainsnak
}
}
}
}
}
*2. Get China's all past head_of_government and where they educated_at and
what other position_held before*
(permalink for this query at
http://bit.ly/2xmxaT1 )
{
entity(id: "Q148") {
label(lang: "en")
head_of_government {
mainsnak {
label(lang: "en")
educated_at {
mainsnak {
id
label(lang: "en")
}
}
position_held {
mainsnak {
label(lang: "en")
}
}
}
}
}
}
You could visit here to use the api explorer to write some query to try
this one
http://wikidata.notimportant.org/graphql
Certainly there is so much to do in order to get most of wikidata:
1. need add reference and qualifiers
2. support search entity
3. other datatype like Time and Media, Geographic and etcs to the graphql
schema
4. and maybe many things to help make reverse lookup possible (like get
one artist's all albums)
Also I realized this certainly can't do a lot of things that SPARQL does,
but in a way it provide a more intuitive way to get many useful data needed
out of wikidata. And it's really expressive for deep nested info compare to
the what we have. It's still an early project, ideas and comments are
welcome.
Source code available:
https://github.com/seansay/wikidata-graphql
--
Best,
李松, Sean Lee
https://notimportant.org