We are releasing Infoboxer
version 0.2 today – which is still young but already
highly useful MediaWiki (including Wikipedia) hi-level data client.
Please look at Infoboxer’s showcase
to have a feeling of what it can do for you.
Since first announce
in Aug 2015, Infoboxer already made a huge advance in stability and usability.
In new release:
new backend library, allowing to receive various lists of pages;
new shiny infoboxerinteractive console for experiments;
various cleanups and fixes.
Lists of pages
Previously, Infoboxer allowed you only to fetch one or several pages from
MediaWiki by their titles:
Number of pages you could fetch this way was limited to 50 (MediaWiki
API limitation) in version 0.1.
Currently, with version 0.2 and new backend library, you can fetch as
many pages as you like in one #get call. Internally, it will be translated
in several API requests. Overall process can take some time, but eventually
you’ll have them :)
Moreover, new version introduces #category, #search and #prefixsearch
methods for receiving whole lists of pages:
Warning: currently, there’s no way to limit number of pages fetches.
So, saying something like Infoboxer.wikipedia.search('ruby') may
require you to wait… a lot of time. Better lists fetching granularity
is on the way!
New backend
For Infoboxer’s needs, we started to develop new generic low-level
MediaWiki client library, called mediawiktory.
It is still really young, yet, unlike all the rivals, it targets all
MediaWiki API, which is pretty hard yet useful. It tries to provide
structure, chanable, Arel/Sequel-alike query interface for perfoming
any task MediaWiki allows. A first glance to MediaWiktory functionality:
> infoboxer -wikipedia
Default Wiki selected: https://en.wikipedia.org/w/api.php.
Now you can use `get('Pagename')`, `category('Categoryname')` and so on.
$ get('Argentina')
With any wiki:
> infoboxer -w http://mywiki/w/api.php
Default Wiki selected: http://mywiki/w/api.php
Now you can use `get('Pagename')`, `category('Categoryname')` and so on.
$ get('MyProduct')
# you can still use any other wiki:
$ wikipedia.get('Argentina')
Cleanups, fixes, shortcuts and so on
Internally, there were a lot of work done for cleanup of code, providing
and documenting new smart features, like Template#to_h allowing to
convert entire large infoboxes into hashes in a blink.
But, there’s a long way ahead of us! You can contribute to infoboxer
development by describing (or fixing!) issues
or just by giving it a try and provide us with your precious feedback.