A solution to the question, how long you have to travel between wiki articles.
Recursivly clone this repository. This is required, because we are using casablanca as http library.
cmake
is our build tool; Required system packages are:
- lib-xerces
- OpenCL
Get your copy of your favorite wikipedia (https://dumps.wikimedia.org), a file including pages and articles in xml is what you are looking for. This file is propably compressed. Untar/bz2 it.
call vorbereitung to generate a index of pages and once again for an dictonary of connections between articles.
./vorbereitung <wiki-dump.xml> wiki.articles
./vorbereitung <wiki-dump.xml> wiki.articles wiki.links
The generated files are between 100 MB for index and 280 MB for articles big (german wiki). They are not compressed.
To start run ./suchen wiki.articles wiki.links
. Caution: Spaces are written as "_". have fun!
To start run ./server wiki.articles wiki.links
. You will probably want to use wikipedia-topfschlagen-web, look there for further instructions.
- Redirects!
- Docs!
Feel free to make a PR.