A good maxim of war is "know your enemy," and the same advice applies to business. You need to know your customers or, on a web site, your visitors. Everything you can know about them is in the Environment variables (discussed in Chapter 16) that Apache gets from the incoming request. Apache's logging directives, which are explained in this chapter, extract whichever elements of this data you want and write them to log files.
However, this is often not very useful data in itself. For instance, you may well want to track the repeated visits of individual customers as revealed by their cookie trail. This means writing rather tricky CGI scripts to read in great slabs of log file, break them into huge, multilevel arrays, and search the arrays to track the data you want.
If your site uses a database manager, you could sidestep this cumbersome procedure by writing scripts on the fly to log everything you want to know about your visitors, reading data about them from the environment variables, and recording their choices as they work through the site. Depending on your needs, it can be much easier to log the data directly than to mine it out of the log files. For instance, one of the authors (PL) has a medical encyclopedia web site (www.Medic-Planet.com). Simple Perl scripts write database records to keep track of the following:
How often each article has been read
How visitors got to it
How often search engine spiders visit and who they are
How often visitors click through the many links on the site and where they go
Having stored this useful information in the database manager, it is then not hard to write a script, accessed via an SSL connection (see Chapter 11), which can only be accessed by the site management to generate HTML reports with totals and statistics that illuminate marketing problems.
Copyright © 2003 O'Reilly & Associates. All rights reserved.