Programming Web Graphics with Perl and GNU Softwar

Creating Data Sets with Data::Xtab

When you are working with real-world data from normalized tables or from database engines, you will find that the data will often require a good deal of manipulation to get it into a well-formed GIFgraph data set format. The Data::Xtab module, available on CPAN, was written by Brian Jepson as a tool for making this sort of raw data easier to graph. It will “cross-tabulate” data from one form into another that is easy for GIFgraph to digest.

Take as an example the case of creating a graph based on the contents of an access log for a proxy server. Let’s say that there are three computers on a network that all share the same proxy server for access to the Internet. For each HTTP request, an entry is made in a log file called proxy-log. The format for each entry in the log in this example is the same format as that of the httpd-cern proxy server:

host - - [day/month/year:hour:minute:second zone]
"GET url version" status bytes

Or, looking at a few lines of the actual log file:

               burroughs.as220.org - - [07/Feb/1998:23:35:06 +0500] 
    "GET http://flatzilla.as220.org/images/spacer.gif HTTP/1.0" 200 61
               snuggles.as220.org - - [07/Feb/1998:00:45:38 +0500] 
    "GET http://flatzilla.as220.org/images/image.gif HTTP/1.0" 200 1624
               snuggles.as220.org - - [08/Feb/1998:00:45:39 +0500] 
    "GET http://flatzilla.as220.org/images/single.gif HTTP/1.0" 200 1499
               gumbo.as220.org - - [08/Feb/1998:01:12:12 +0500] 
    "GET http://flatzilla.as220.org/images/beatniks.gif HTTP/1.0" 200 1744 ...