h1. Twitalytic

Twitter data crawler, replies archiver, and statistics generator.


h2. WARNING: Pre-alpha code, PROCEED AT YOUR OWN RISK!

This is not production code. This is a pre-alpha web application. The intended audience is server administrators with experience installing and troubleshooting PHP/MySQL hosted web applications. While Twitaltyic uses OAuth to access Twitter, the webapp's login is not as secure as it should be. Right now this code is for experimentation and tinkering only. Do not run on a public, production server. You have been warned.


h2. SYSTEM REQUIREMENTS

* "PHP 5":http://php.net with cURL extension enabled
* "Smarty":http://smarty.net
* "MySQL 5":http://mysql.com/


h2. INSTALL

In future versions, this will get easier.

h3. Set up database

* Create a database and select it, i.e., CREATE DATABASE twitalytic
* Build tables with <code>build-db.sql</code>

h3. Register your app with Twitter (for OAuth support)

* Log into Twitter and "register your Twitalytic instance":http://twitter.com/oauth_clients/. Set the callback URL to <code>http://yourserver.com/path-to-twitalytic/account/oauth.php</code>. Write down the items labeled "Consumer key" and "Consumer secret."

h3. Configure the app

* *Shared libraries*: Edit <code>twitalytic/common/config.inc.php</code>. Add your database credentials, file and webapp paths, and the key/secret you got from Twitter in the previous step.
* *Crawler*: Edit <code>twitalytic/crawler/config.crawler.inc.php</code>, set include path to the absolute location of <code>twitalytic/common</code>
* *Webapp*: Edit <code>twitalytic/webapp/config.webapp.inc.php</code>, set include path to the absolute location of <code>twitalytic/common</code>
* *Webserver (optional)*: Set <code>httpd.conf</code> DocumentRoot to <code>twitalytic/webapp/</code>


h2. RUN

Visit the web application on your server, register/login, and set up your Twitter accounts.

Then, run the crawler to load your Twitter data:

<code>$ php crawler/crawl.php</code>

To view what's going on with the crawler, use this command:

<code>$ tail -f crawler/logs/crawler.log</code>

Cron the crawler's run command to go at least once an hour.