Using Wikipedia as brute forcing dictionary

The success and the time elapsed in a brute forcing attack depends by the number of discovered brute forcing points, the quality of the tool used (like THC-hydra, brutus or medusa) and the quality of the dictionary used.

Sometimes using a incremental dictionary is a waste of time, a good dictionary can be the success key to a fast brute forcing attack. Get a dictionary of common words and keep it updated is an hard work.

Wikipedia is a free multilingual encyclopedia, it currently contains 2,683,099 articles. This is a really good database to generate a dictionary of common words.

Wikipedia offers free copies of all available content to interested users. These databases can be used for mirroring, personal use, informal backups, or database queries. All text content is licensed under the GNU Free Documentation License (GFDL). Images and other files are available under different terms, as detailed on their description pages.

The Wikipedia database download page is available here: http://en.wikipedia.org/wiki/Wikipedia_database and the database dumps are available here: http://download.wikimedia.org/backup-index.html

A good dictionary must contains the most common terms used in a current language and also common words that can be used as password, an example is “foo”, “bar”, “1234”, “antani”, etc.

We can create two types of dictionary, a dictionary containing all the words inside wikipedia, a dictionary containing all article titles, a dictionary containing all the words in the article titles.

After downloading a bunch of gigs we get the wikipedia database dump in XML, the fields that we need to create our dictionary are <title> and <text>.

Now you can create all the types of dictionary that you need: words, titles, case sensitive or case insensitive.

To achieve better performances I used simple bash scripting for parsing because using a DOM or SAX parser is too slow with these very big XMLs.

This dictionary contains all the article titles, so you can guess password like names, cities, etc.

To create it you can use the following or you can edit it to fit your needs, it’s not beautiful but works:

grep -E '<title>(.*?)</title>' itwiki-20081206-pages-meta-current.xml | \
cut -d '>' -f2| cut -d '<' -f1 | grep -v : | sed s/\(.*\)//g| sort | uniq

Word dictionary contains all the words in the wikipedia articles, you can create it with a command similar to the above, I left it for your homework ;)

Happy brute forcing!