Wikipedia client for the command line

Don’t you want the power of Wikipedia at your fingertips when you are logged into your Linux server? With this tool you can! It is build with PHP/Curl and is based on the Wikipedia API.

Wikipedia (open)search API

Wikipedia allows us to program software that takes advantage of their huge database of articles and definitions. They do this by exposing an API. I implemented their “opensearch” API call to search for articles. We are not doing anything unintended. Wikipedia made the API to allow people to build useful (or less useful, you decide) tools like this one. They do ask people to behave nicely in their Wikipedia API Etiquette, so read them and behave!

HTML to text

Instead of making my own wikitext parser, I used the Wikipedia “parse” API call. This call transforms articles from the internal wikitext format to HTML. I use the command line application “elinks” (alternative to the “lynx” text-mode web browser) to convert the HTML to text so it can be displayed on the console. This browser software is configured not to use the network or write files to disk, so you do not have to worry about your privacy.

How does this work?

It is a PHP script that uses Curl. The whole wped script is currently only 70 lines of code. The opensearch call is really easy to implement as you can see below:

$url = 'http://en.wikipedia.org/w/api.php';
$user_agent = 'Wped/0.1 (http://github.com/mevdschee/wped) PHP-Curl';

$curl = curl_init();
curl_setopt($curl, CURLOPT_RETURNTRANSFER, true);
curl_setopt($curl, CURLOPT_USERAGENT, $user_agent);

$arguments = array(
  'action'=>'opensearch',
  'search'=>$search,
  'limit'=>3,
  'namespace'=>0,
  'format'=>'xml'
);

curl_setopt($curl, CURLOPT_URL, $url.'?'.http_build_query($arguments));
$results = simplexml_load_string(curl_exec($curl));

With this example it should be easy for you to start building your own Wikipedia tool, let me know if you did!

Quickstart

To install, just copy-paste the following lines in your console:

sudo apt-get install php5-cli php5-curl elinks
wget https://raw.githubusercontent.com/mevdschee/wped/master/wped.php -O wped
chmod 755 wped
sudo mv wped /usr/bin/wped

To uninstall use the following:

sudo rm /usr/bin/wped

Congratulations, now you can search for elephants using the “wped” command! And if you need the full article you can use the “-f” flag.

Yes! I need this, fast!

How can you have lived without it? Why would you ever use “wget”, when you can use “wped”? It adds so much fun to all your Wikipedia searches. The source and full instructions are available on my Github account: https://github.com/mevdschee/wped

The post Wikipedia client for the command line appeared first on LeaseWeb Labs.