| Directory | |
|---|---|
| User space: Home | Talk (archives) | Sandboxes: General 1 · General 2 | Smart questions · Cluebat | |
| Software: Test account | Wiki.java | Servlets | |
| Links: WikiProject Spam · Spam blacklist: local · global · XLinkBot | Copyvios | Contributor copyright | |
Finally, something that can search for spam across all 700 or so Wikimedia projects. It is licensed under the GNU General Public License version 3. A copy of the license is available at http://www.gnu.org/licenses/gpl-3.0.txt or in the JAR itself.
At the moment, 16 wikis are searched every 8.5 seconds. Hence, a one-site spamsearch takes about 6.5 minutes (in comparison, the toolserver spamsearch searches 57 wikis in about 2.5 minutes). This is a hard-coded limit - it tends to kill itself if the internet connection is overloaded - but you can tweak the values by editing the source code and recompiling.
- System requirements
- A properly configured
Java Runtime EnvironmentJava Development Kit, version 1.5 or greater. Go here to download one. - A reasonably fast internet connection (tested on 256/64k, ping ~350ms).
- Installation instructions
I need someone to host the JAR file. In the meantime, you can download the source (see below).
- Running instructions
Open up a command line interface and change directory to the directory you saved the above JAR file. Enter the following command:
java Spamsearch example.com
... where example.com the sites spammed. Beware of case sensitivity on non-Windows filesystems. The results of the spamsearch, once complete, will be in the same directory with filename results.txt.
- Source code and internals
Spamsearch uses Wiki.java as the wiki interface. The other half is here.
- Bugs and problems
As I have given you the code, {{sofixit}}, then edit the source files on-wiki. If you can't (perhaps because of a lack of knowledge of Java, a JDK or effort), then file bug reports at User talk:MER-C.