A hit is basically a line from your log file, it is a file requested from your website. In general hits have the following information associated with them (gained from the log):

• IP address of the visitor - this is the IP address that the visitor used to access the file.
• Request URL - the url that was used by the visitor to access the file.
• Date/Time of request - the date/time the request was made.
• HTTP Status Code - the status code the server returned to the visitor.
• Size of file returned - the size of the file returned (if appropriate) to the visitor.

The above is generally the basic information stored in a log file about each file request. Some web servers will also, optionally, log the following information that polliwog uses:

• Referer - the referring url for the file access.
• User Agent - information about the application the visitor used to access the file.

(Note: log files can contain lots of other information but polliwog does not use it. For those who are interested, it is relatively simple to extend polliwog to make use of extra information present in the log but how to do that is not covered in this manual.)

To convert log file hit information into objects that polliwog can make use of the format of the log file needs to be specified. This is done via the use of a log entry format. In general this is defined in a XML file. polliwog has a number of these files pre-configured for various log formats. See the log entry formats page for more details of the formats and how to create one.

For each line in the log file(s) polliwog will use the relevant log entry format and use it to create an instance of the class specified by property:
By default this is set to: org.polliwog.data.Hit.

For each hit, if there is a page resolver present, then polliwog will try and determine whether the hit is a page and thus should be added to a visit. If no resolver is present then polliwog considers all hits to be pages.

In general, in any log, you are not interested in all the hits. You are generally interested in a particular subset, for example, hits within a certain chunk of time. polliwog provides the ability to filter hits based on criteria in the Hit object. Filtered hits are not counted in the statistics generated by polliwog and are not resolved into pages. See the hit filtering page for more details.

It is possible to create custom collections of hits (filtered and non-filtered) that you can produce your own statistics pages with. See the hit collections page for more details.