Wikicene is a demo project I created while investigating the Apache Lucene search library. It is a web application that provides the ability to search through a dump of Wikipedia articles. Its goal is to showcase some of Lucene's features (e.g. different analyzers and query types) in an attractive form.
Before starting the main Wikicene application, you need to generate a dump of Wikipedia articles. To do that, simply run the python script:
$ python3 wikirandom.pyThis will begin downloading a random Wikipedia article from
this endpoint each couple of seconds and putting it into
a random-articles.dump text file (this is the file Wikicene will read on start-up).
You can stop the script with CTRL+C.
The download cadence can be increased, but we want to avoid spamming such a generous endpoint.
After you have your desired amount of articles in the dump (Wikicene can handle tens of thousands easily), you can start Wikicene with:
$ ./gradlew clean build runNote that the first time you run it, Wikicene will index all the articles in the dump several times, once per supported analyzer. A normal application would only use one analyzer when indexing, but this is a demo designed to showcase the search behavior with different analyzers. For around 25k articles, expect a first-run time of about 1 or 2 minutes before you can interact with the application, as well as about 0.5 GB of disk usage.
Then you can interact with Wikicene in one of two ways. First, you can make requests via a REST-like API by accessing the endpoint:
localhost:8080/?term=planet
This endpoint will return a JSON response with the results.
However, you'll likely want to interact with Wikicene using its humble frontend, available at:
localhost:8080/frontend/
- Python3
- Java 8
- Gradle
- Vert.x
- Apache Lucene
- Bootstrap
- bootstrap-select
- Brython
If you believe I lack any mention or attribution, feel free to contact me.