The Webalizer - A web server log file analysis thingie
Copyright 1997, 1998 by Bradford L. Barrett (brad@mrunix.net)

Distributed under the GNU GPL.  See the files "COPYING" and
"Copyright" supplied with the distribution for additional info.


What is The Webalizer?
----------------------

The Webalizer is a web server log file analysis program which produces
usage statistics in HTML format for viewing with a browser.  The results
are presented in both columnar and graphical format, which facilitates
interpretation.  Yearly, monthly, daily and hourly usage statistics are
presented, along with the ability to display usage by site, URL, referrer,
user agent (browser) and country (user agent and referrer are only
available if your web server procduces Combined log format files).

The Webalizer supports CLF (common log format) log files, as well as
Combined log formats as defined by NCSA and others, and variations
of these which it attempts to handle intelligently.

This documentation applies to The Webalizer Version 1.2x

Running the Webalizer
---------------------

The Webalizer was designed to be run from a Unix command line prompt or
as a cron job.  There are several command line options which will modify
the results it produces, and configuration files can be used as well.
The format of the command line is:

webalizer [options ...] [log-file]

Where 'options' can be one or more of the supported command line
switches described below.  'log-file' is the name of the log file
to process (see below for more detailed information).


Once executed, the general flow of the program follows:

o A default configuration file is scanned for.  A file named
  'webalizer.conf' is searched for in the current directory, and if
  found, it's configuration data is parsed.  If the file is not
  present in the current directory,  the file '/etc/webalizer.conf'
  is searched for and, if found, is used instead.

o Any command line arguments given to the program are parsed.  This
  may include the specification of a configuration file, which is
  processed at the time it is encountered.

o If a log file was specified, it is opened and made ready for
  processing.  If no log file was given, STDIN is used for input.

o If an output directory was specified, the program does a 'chdir' to
  that directory in prepration for generating output.  If no output
  directory was given, the current directory is used.

o If no hostname was given, the program attempts to get the hostname
  using a uname system call.  If that fails, 'localhost' is used.

o A history file is searched for in the current directory (output
  directory) and read if found.  This file keeps totals for previous
  months, which is used in the main 'index.html' HTML document.

o If incremental processing was specified, a data file is searched for
  and loaded if found, containing the 'internal state' data of the
  program at the end of a previous run.

o Main processing begins on the log file.  If the log spans multiple
  months, a seperate HTML document is created for each month.

o After main processing, the main 'index.html' page is created, which
  has totals by month and links to each months HTML document.

o A new history file is saved to disk, which includes totals generated
  by The Webalizer during the current run.

o If incremental processing was specified, a data file is written that
  contains the 'internal state' data at the end of this run.


Incremental Processing
----------------------

Version 1.2x of The Webalizer adds incremental run capability.  Simply
put, this allows processing large log files by breaking them up into
smaller pieces, and processing these pieces instead.  What this means
in real terms is that you can now rotate your log files as often as you
want, and still be able to produce monthly usage statistics without the
loss of any detail.  Basically, The Webalizer saves and restores all
internal data in a file named 'webalizer.current'.  This allows the
program to 'start where it left off' so to speak, and allows the
preservation of detail from one run to the next.  The data file is
placed in the current output directory, and is a plain ascii text
file that can be viewed with any standard text editor.

Some special precautions need to be taken when using the incremental
run capability of The Webalizer.  Configuration options should not be
changed between runs, as that could cause corruption of the internal
data stored.  For example, changing the MangleAgents level will cause
different representations of user agents to be stored, producing invalid
results in the user agents section of the report.  If you need to change
configuration options, do it at the end of the month after normal
processing of the previous month and before processing the current month.
You may also want to delete the 'webalizer.current' file as well.

The Webalizer also attempts to prevent data duplication by keeping
track of the timestamp of the last record processed.  This timestamp
is then compared to current records being processed, and any records
that were logged previous to that timestamp are ignored.  This, in
theory, should allow you to re-process logs that have already been
processed, or process logs that contain a mix of processed/not yet
processed records, and not produce duplication of statistics.  The
only time this may break is if you have duplicate timestamps in two
seperate log files... any records in the second log file that do have
the same timestamp as the last record in the previous log file processed,
will be discarded as if they had already been processed.  There are
lots of ways to prevent this however, for example, stopping the web
server before rotating logs will prevent this situation.  This setup
also necessitates that you always process logs in chronological order,
otherwise data loss will occur as a result of the timestamp compare.


Command Line Options
--------------------

The Webalizer supports many different configuration options that will
alter the way the program behaves and generates output.  Most of these
can be specified on the command line, while some can only be specified
in a configuration file. The command line options are listed below,
with references to the corresponding configuration file keywords.

--------------------------------------------------------------------------

General Options
---------------

-h        Display all available command line options and exit program.

-v        Display program version and exit program.

-d        Display additional 'debugging' information for errors and
          warnings produced during processing.  This normally would
          not be used except to determine why you are getting all those
          errors and wanted to see the actual data.  Normally The
          Webalizer will just tell you it found an error, not the
          actual data.  This option will display the data as well.
          Config file keyword: Debug

-g        Use GMT (UTC) time instead of local timezone.  Normally,
          The Webalizer will display the time generated in the timezone
          of the local machine, such as EDT or PST.  This switch will
          cause the times to be displayed in GMT (UTC) time instead.
          Config file keyword: GMTTime

-i        Ignore history file.  USE WITH CAUTION.  This causes The
          Webalizer to ignore any existing history file produced from
          previous runs and generate it's output from scratch.  The
          effect will be as if The Webalizer is being run for the
          first time and any previous statistics will be lost (although
          the HTML documents, if any, will not be deleted) on the main
          index.html (yearly) web page.
          Config file keyword: IgnoreHist

-p        Preserve state (incremental processing).  This allows the
          processing of partial logs in increments.  At the end of
          the program, all relevant internal data is saved, so that
          it may be restored the next time the program is run.  This
          allows sites that must rotate their logs more than once a
          month to still be able to use The Webalizer, and not worry
          about having to gather and feed an entire months logs to
          the program at the end of the month.  See the section on
          "Incremental Processing" below for additional information.
          The default is to not perform incremental processing.  Use
          this command line option to enable the feature.
          Config file keyword: Incremental

-q        Quiet mode.  Normally, The Webalizer will produce various
          messages while it runs letting you know what it's doing.
          This option will supress those messages.  It should be
          noted that this WILL NOT supress errors and warnings, which
          are output to STDERR.
          Config file keyword: Quiet

-Q        ReallyQuiet mode.  This allows supression of _all_ messages
          generated by The Webalizer, including warnings and errors.
          Useful when The Webalizer is run as a cron job.
          Config file keyword: ReallyQuiet

-T        Display timing information.  The Webalizer keeps track of the
          time it begins and ends processing, and normally displays the
          total processing time at the end of each run.  If quiet mode
          (-q or 'Quiet yes' in configuration file) is specified, this
          information is not displayed.  This option forces the display
          of timing totals if quiet mode has been specified, otherwise
          it is redundant and will have no effect.
          Config file keyword: TimeMe

-c file   This option specifies a configuration file to use.  Configuration
          files allow greater control over how The Webalizer behaves, and
          there are several ways to use them.  As of version 0.98, The
          Webalizer searches for a default configuration file in the
          current directory named "webalizer.conf", and if not found,
          will search in the /etc/ directory for a file of the same name.
          In addition, you may specify a configuration file to use with
          this command line option.

-n name   This option specifies the hostname for the reports generated.
          The hostname is used in the title of all reports, and is also
          prepended to URL's in the reports.  This allows The Webalizer
          to be run on log files for 'virtual' web servers or web servers
          that are different than the machine the reports are located on,
          and still allows clicking on the URL's to go to the proper
          location.  If a hostname is not specified, either on the
          command line or in a configuration file, The Webalizer attempts
          to determine the hostname using a 'uname' system call.  If this
          fails, "localhost" will be used as the hostname.
          Config file keyword: HostName

-o dir    This options specifies the output directory for the reports.
          If not specified here or in a configuration file, the current
          default directory will be used for output.
          Config file keyword: OutputDir

-t name   This option specifies the title string for all reports.  This
          string is used, in conjunction with the hostname (if not blank)
          to produce the actual title.  If not specified, the default of
          "Usage Statistics for" will be used.
          Config file keyword: ReportTitle

-G        Supress hourly graph.  Normally, The Webalizer produces
          hourly statistics in both Graph and Columnar forms.  This
          option will supress the Hourly Graph only from being generated.
          Config file keyword: HourlyGraph

-H        Supress Hourly statistics.  Normally, The Webalizer produces
          hourly statistics in both Graph and Columnar forms.  This
          option will supress the Hourly Statistics table only from
          being generated.
          Config file keyword: HourlyStats

-M num    Mangle user agent names.  Normally, The Webalizer will keep
          track of the user agent field verbatim.  Unfortunately, there are
          a ton of different names that user agents go by, and the field
          also reports other items such as machine type and OS used. For
          Example, Netscape 4.03 running on Windows 95 will report a
          different string than Netscape 4.03 running on Windows NT, so even
          though they are the same browser type, they will be considered
          as two totally different browsers by The Webalizer.  For that
          matter, Netscape 4.0 running on Windows NT will report different
          names if one is run on an Alpha and the other on an Intel
          processor!  Internet Exploder is even worse, as it reports itself
          as if it were Netscape and you have to search the given string a
          little deeper to discover that it is really MSIE!  In order to
          consolidate generic browser types, this option will cause The
          Webalizer to 'mangle' the user agent field, attempting to
          consolidate generic browser types.  This currently only works
          for MSIE and Netscape browsers, and all others will be processed
          without any modification.  There are 6 levels that can be
          specified, each producing different levels of detail.  Level 5
          displays only the browser name (MSIE or Mozilla) and the major
          version number.  Level 4 will also display the minor version
          number (single decimal place).  Level 3 will display the minor
          version number to two decimal places.  Level 2 will add any
          sub-level designation (such as Mozilla/3.01Gold or MSIE 3.0b).
          Level 1 will also attempt to add the system type.  The default
          Level 0 will disable name mangling and leave the user agent
          field unmodified, producing the greatest amount of detail.
          Configuration file keyword: MangleAgents


Hide Options
------------

The following options take a string argument to use as a comparison
for matching.  Except for the IndexAlias option, the string argument
can be plain text, or plain text that either starts or ends with the
wildcard character '*'.

For Example:

Given the string "yourmama/was/here", the arguments "was", "*here" and
"your*" will all produce a match.


-a name   This option allows hiding of user agents (browsers) from the
          "Top User Agents" table in the report.  This option really
          isn't too useful as there are a zillion different names that
          current browsers go by, depending where they were obtained,
          however you might have some particular user agents that hit
          your site a lot that you would like to exclude from the list.
          You must have a web server that includes user agents in it's
          log files for this option to be of any use.  In addition, it
          is also useless if you disable the user agent table in the
          report (see the -A command line option or "TopAgents"
          configuration file keyword). You can specify as many of these
          as you want on the command line.  The wildcard character '*'
          can be used either in front of or at the end of the string.
          (ie: Mozilla/4.0* would match anything that starts with the
          string "Mozilla/4.0").
          Config file keyword: HideAgent

-r name   This option allows hiding of referrers from the "Top Referrer"
          table in the report.  Referrers are URL's, either on your own
          local site or a remote site, that refered the user to a URL
          on your web server.  This option is normally used to hide
          your own server from the table, as your own pages are usually
          the top referrers to your own pages (well, you get the idea).
          You must have a web server that includes referrer information
          in the log files for this option to be of any use.  In addition,
          it is also useless if you disable the referrers table in the
          report (see the -R command line option or "TopReferrers"
          configuration file keyword).  You can specify as many of these
          as you like on the command line.
          Config file keyword: HideReferrer

-s name   This option allows hiding of sites from the "Top Sites" table
          in the report.  Normally, you will only want to hide your own
          domain name from the report, as it usually is one of the top
          sites to visit your web server.  This option is of no use if
          you disable the top sites table in the report (see the -S
          command line option or "TopSites" configuration file option).
          Config file keyword: HideSite

-u name   This option allows hiding of URL's from the "Top URL's" table
          in the report.  Normally, this option is used to hide images,
          audio files and other objects your web server dishes out that
          would otherwise clutter up the table.  This option is of no
          use if you diable the top URL's table in the report (see the
          -U command line option or "TopURLs" configuration file keyword).
          Config file keyword: HideURL

-I name   This option allows you to specify additional index.html aliases.
          The Webalizer usually strips the string 'index.' from URL's
          before processing, which has the effect of turning a URL such
          as /somedir/index.html into just /somedir/ which is really the
          same URL and should be treated as such.  This option allows you
          to specify _additional_ strings that are to be treated the same
          way.  Use with care, improper use could cause unexpected resluts.
          For example, if you specify the alias string of 'home', a URL
          such as /somedir/homepages/brad/home.html would be converted
          into just /somedir/ which probably isn't what was intended.
          This option is useful if your web server uses a different default
          index page other than the standard 'index.html' or 'index.htm',
          such as 'home.html' or 'homepage.html'.  The string specified
          is searched for _anywhere_ in the URL, so "home.htm" would
          turn both "/somedir/home.htm" and "/somedir/home.html" into
          just "/somedir/".  Go easy on this one, each string specified
          will be scanned for in EVERY log record, so if you specify a
          bunch of these, you will notice degraded performance.  Wildcards
          are not allowed on this one.
          Config file keyword: IndexAlias

Table Size Options
------------------

-A num    This option specifies the number of entries to display in the
          "Top User Agents" table.  To disable the table, use a value of
          zero (0).
          Config file keyword: TopAgents

-C num    This option specifies the number of entries to display in the
          "Top Countries" table.  To disable the table, use a value of
          zero (0).
          Config file keyword: TopCountries

-R num    This option specifies the number of entries to display in the
          "Top Referrers" table.  To disable the table, use a value of
          zero (0).
          Config file keyword: TopReferrers

-S num    This option specifies the number of entries to display in the
          "Top Sites" table.  To disable the table, use a value of
          zero (0).
          Config file keyword: TopSites

-U num    This option specifies the number of entries to display in the
          "Top URL's" table.  To disable the table, use a value of
          zero (0).
          Config file keyword: TopURLs

--------------------------------------------------------------------------


CONFIGURATION FILES
-------------------

The Webalizer allows configuration files to be used in order to simplify
life for all.  There are several ways that configuration files are accessed
by the Webalizer.  When The Webalizer first executes, it looks for a
default configuration file named "webalizer.conf" in the current directory,
and if not found there, will look for "/etc/webalizer.conf".  In addition,
configuration files may be specified on the command line with the '-c'
option.  There are lots of different ways you can combine the use of
configuration files and command line options to produce various results.
The Webalizer always looks for and reads configuration options from a
default configuration file before doing anything else.  Because of this,
you can override options found in the default file by use of additional
configuration files specified on the command line or command line options
themselves.  If you specify a configuration file on the command line, you
can override options in it by additional command line options which follow.
For example, most users will most likely want to create the default file
/etc/webalizer.conf and place options in it to specify the hostname, log
file, table options, etc...  At the end of the month when a different log
file is to be used (the end of month log), you can run TheWebalizer as
usual, but put the different filename on the end of the command line, which
will override the log file specified in the configuration file.  It should
be noted that you cannot override some configuration file options by the
use of command line arguments.  For example, if you specify "Quiet yes" in
a configuration file, you cannot override this with a command line argument,
as the command line option only _enables_ the feature (-q option).

The configuration files are standard ASCII text files that may be created
or edited using any standard editor.  Blank lines and lines that begin
with a pound sign ('#') are ignored.  Any other lines are considered to
be configurgation lines, and have the form "Keyword Value", where the
'Keyword' is one of the currently available configuration keywords defined
below, and 'Value' is the value to assign to that particular option.  Any
text found after the keyword up to the end of the line is considered the
keyword's value, so you should not include anything after the actual value
on the line that is not actually part of the value being assigned.  The
file "sample.conf" provided with the distribution contains lots of useful
documentation and examples as well.

--------------------------------------------------------------------------

General Configuration Keywords
------------------------------

LogFile       This defines the logfile to use.  It should be a fully qualified
              name (ie: contain the path), but relative names will work as
              well.  If not specified, the logfile defaults to STDIN. 

OutputDir     This defines the output directory to use for the reports.  If
              it is not specified, the current directory is used.
              Command line argument: -o

ReportTitle   This specifies the title to use for the generated reports.
              It is used in conjunction with the hostname (unless blank)
              to produce the final report titles.  If not defined, the
              default of "Usage Statistics for" is used.
              Command line argument: -t

HostName      This defines the hostname.  The hostname is used in the
              report title as well as being prepended to URL's in the
              "Top URL's" table.  This allows The Webalizer to be run
              on "virtual" web servers, or servers that do not reside
              on the local machine, and allows clicking on the URL to
              go to the right place.  If not specified, The Webalizer
              attempts to get the hostname via a 'uname' system call,
              and if that fails, will default to "localhost".
              Command line argument: -n

Quiet         This allows you to enable or disable informational messages
              while The Webalizer is running.  The value for this keyword              while it is running.  The values for this keyword can be
              either 'yes' or 'no'.  Using "Quiet yes" will supress these
              messages, while "Quiet no" will enable them.  The default
              is 'no' if not specified, which will allow The Webalizer
              to display informational messages.  It should be noted that
              this option has no effect on Warning or Error messages that
              may be generated, as they go to STDERR.
              Command line argument: -q

TimeMe        This allows you to display timing information regardless of
              any "quiet mode" specified.  Useful only if you did in fact
              tell the webalizer to be quiet either by using the -q command
              line option or the "Quiet" keyword, otherwise timing stats
              are normally displayed anyway.  Values may be either 'yes'
              or 'no', with the default being 'no'.
              Command line argument: -T

GMTTime       This keyword allows timestamps to be displayed in GMT (UTC)
              time instead of local time.  Normally The Webalizer will
              display timestamps in the timezone of the local machine
              (ie: PST or EDT).  This keyword allows you to specify the
              display of timestamps in GMT (UTC) time instead.  Values
              may be either 'yes' or 'no'.  Default is 'no'.
              Command line argument: -g

Debug         This tells The Webalizer to display additional information
              when it encounters Warnings or Errors.  Normally, The
              Webalizer will just tell you it found a bad record or
              field.  This option will enable the display of the actual
              data that produced the Warning or Error as well.  Useful
              only if you start getting lots of Warnings or Errors and
              want to determine the cause.  Values may be either 'yes'
              or 'no', with the default being 'no'.
              Command line argument: -d

IgnoreHist    This supresses the reading of a history file.  USE WITH
              EXTREME CAUTION as the history file is how The Webalizer
              keeps track of previous months.  The effect of this option
              is as if The Webalizer was being run for the very first
              time, and any previous data is discarded.  Values may be
              either 'yes' or 'no', with the default being 'no'.
              Command line argument: -i

HourlyGraph   This keyword is used to either enable or disable the creation
              and display of the Hourly Usage graph.  Values may be either
              'yes' or 'no', with the default being 'yes'.
              Command line argument: -G

HourlyStats   This keyword is used to either enable or disable the creation
              and display of the Hourly Usage statistics table.  Values may
              be either 'yes' or 'no', with the default being 'yes'.
              Command line argument: -H

IndexAlias    This allows additional 'index.html' aliases to be defined.
              Normally, The Webalizer scans for and strips the string
              "index." from URL's before processing them.  This turns a
              URL such as /somedir/index.html into just /somedir/ which
              is really the same URL.  This keyword allows _additional_
              names to be treated in the same fashion for sites that use
              different default names, such as "home.html".  The string
              is scanned for anywhere in the URL, so care should be used
              if and when you define additional aliases.  For example,
              if you were to use an alias such as 'home', the URL
              /somedir/homepages/brad/home.html would be turned into just
              /somedir/ which probably isn't the intended result.  Instead,
              you should have specified 'home.htm' which would correctly
              turn the URL into /somedir/homepages/brad/ like intended.
              It should also be noted that specified aliases are scanned
              for in EVERY log record... A bunch of aliases will noticably
              degrade performance as each record has to be scanned for
              every alias defined.  You don't have to specify 'index.' as
              it is always the default.
              Command line argument: -I 

MangleAgents  The MangleAgents keyword specifies the level of user agent
              name mangling, if any.  Name mangling only works for MSIE
              and Netscape (Mozilla) browsers, and all others will be
              left unmodified.  There are 6 levels that may be specified,
              each producing a different level of detail displayed.  Level 5
              displays only the browser name (MSIE or Mozilla) and the major
              version number.  Level 4 adds the minor version (single
              decimal place).  Level 3 adds the minor version to two decimal
              places.  Level 2 will also add any sub-level designation
              (such as Mozilla/3.01Gold or MSIE 3.0b).  Level 1 will also
              attempt to add the system type.  The default level 0 will
              leave the user agent field unmodified and produces the
              greatest amount of detail.
              Command line argument: -M

Incremental   This allows incremental processing to be enabled or disabled.
              Incremental processing allows processing partial logs without
              the loss of detail data from previous runs in the same month.
              This feature saves the 'internal state' of the program so that
              it may be restored in following runs.  See the section below
              titled "Incremental Processing" for additional information.
              The value may be 'yes' or 'no', with the default being 'no'.
              Command line argument: -p

Top Table Keywords
------------------

TopAgents     This allows you to specify how many "Top" user agents are
              displayed in the "Top User Agents" table.  The default
              is 15.  If you do not want to display user agent statistics,
              specify a value of zero (0).  The display of user agents
              will only work if your web server includes this information
              in its log file (ie: a combined log format file).
              Command line argument: -A 

TopCountries  This allows you to specify how many "Top" countries are
              displayed in the "Top Countries" table.  The default is
              50.  If you want to disable the countries table, specify
              a value of zero (0).
              Command line argument: -C 

TopReferrers  This allows you to specify how many "Top" referrers are
              displayed in the "Top Referrers" table.  The default is
              30.  If you want to disable the referrers table, specify
              a value of zero (0).  The display of referrer information
              will only work if your web server includes this information
              in its log file (ie: a combined log format file).
              Command line argument: -R 

TopSites      This allows you to specify how many "Top" sites are
              displayed in the "Top Sites" table.  The default is 30.
              If you want to disable the sites table, specify a value
              of zero (0).
              Command line argument: -S 

TopURLs       This allows you to specify how many "Top" URL's are
              displayed in the "Top URL's" table.  The default is 30.
              If you want to disable the URL's table, specify a value
              of zero (0).
              Command line argument: -U 


Hide Object Keywords
--------------------

These keywords allow you to hide user agents, referrers, sites and
URL's from the various "Top" tables.  The value for these keywords
are the same as those used in their command line counterparts.  You
can specify as many of these as you want without limit.  Refer to the
section above on "Command Line Options" for a description of the string
formatting used as the value.  Values cannot exceed 80 characters in
length.

HideAgent     This allows specified user agents to be hidden from the
              "Top User Agents" table.  Not very useful, since there
              a zillion different names by which browsers go by today,
              but could be useful if there is a particular user agent
              (ie: robots, spiders, realaudio, etc..) that hits your
              site frequently enough to make it into the top user agent
              listing.  This keyword is useless if 1) your log file does
              not provide user agent information or 2) you disable the
              user agent table.
              Command line argument: -a 

HideReferrer  This allows you to hide specfied referrers from the
              "Top Referrers" table.  Normally, you would only specify
              your own web server to be hidden, as it is usually the
              top generator of references to your own pages.  Of course,
              this keyword is useless if 1) your log file does not include
              referrer information or 2) you disable the top referrers
              table.
              Command line argument: -r 

HideSite      This allows you to hide specified sites from the "Top
              Sites" table.  Normally, you would only specify your own
              web server or other local machines to be hidden, as they
              are usually the highest hitters of your web site, especially
              if you have their browsers home page pointing to it.
              Command line argument: -s 

HideURL       This allows you to hide URL's from the "Top URL's" table.
              Normally, this is used to hide items such as graphic files,
              audio files or other 'non-html' files that are transferred
              to the visiting user.
              Command line argument: -u 

Group Object Keywords
---------------------

The Group* keywords allow object grouping based on Site, URL, Referrer
and User Agent.  Combined with the Hide* keywords, you can customize
exactly what will be displayed in the 'Top' tables.  For example, to
only display totals for a particular directory, use a GroupURL and HideURL
with the same value (ie: '/help/*').  Group processing is only done after
the individual record has been fully processed, so name mangling and
site total updates have already been peformed.  Because of thie, groups
are not counted in the main site total (as that would cause duplication).
Groups can be displayed in bold and shaded as well.  There are no
command line switches for these keywords.

GroupReferrer Allows grouping Referrers.  Can be handy for some of the
              major search engines that have multiple host names a
              referral could come from.

GroupURL      This keyword allows grouping URL's. Useful for grouping
              complete directory trees.

GroupSite     This keywords allows grouping Sites.  Most used for
              grouping top level domains and unresolved IP address
              for local dial-ups, etc...

GroupAgent    Groups User Agents.  A handy example of how you could use
              this one is to use "Mozilla" and "MSIE" as the values for
              GroupAgent and HideAgent keywords.  Make sure you put the
              "MSIE" one first.

GroupShading  Allows shading of table rows for groups.  Value can be
              'yes' or 'no', with the default being 'yes'.

GroupHighlight Allows bolding of table rows for groups.  Value can be
               'yes' or 'no', with the default being 'yes'.


Ignore Object Keywords
----------------------

These keywords allow you to completely ignore log records when generating
statistics.  Records can be ignored based on site, URL, user agent and
referrer.  Be aware that by choosing to ignore records, the accuracy of
the generated statistics become skewed, making it impossible to produce
an accurate representation of load on the web server.  These keywords
behave identical to the Hide* keywords above, where the value can have
a leading or trailing wildcard '*'.  These keywords, like the Hide* ones,
have an absolute limit of 80 characters for their values.  These keywords
do not have any command line switch counterparts, so they may only be
specified in a configuration file.

IgnoreSite    This allows specified sites to be completely ignored from
              the generated statistics.

IgnoreURL     This allows specified URL's to be completely ignored from
              the generated statistics.  One use for this keyword would
              be to ignore all hits to a 'temporary' directory where
              development work is being done, but is not accessable to
              the outside world.

IgnoreReferrer This allows records to be ignored based on the referrer
               field.

IgnoreAgent   This allows specified User Agent records to be completely
              ignored from the statistics.  Maybe useful if you really
              don't want to see all those hits from MSIE :)


HTML Generation Keywords
------------------------

These keywords allow you to customize the HTML code that The Webalizer
produces, such as adding a corporate logo or links to other web pages.
You can specify as many of these keywords as you like, and they will be
used in the order that they are found in the file.  Values cannot exceed
80 characters in length, so you may have to break long lines up into two
or more lines.  There are no command line counterparts to these keywords.


HTMLHead       This keyword defines HTML code to be placed immediately
               after the <HEAD> section of the report, just before the
               title and "summary period/generated on" lines.  If used,
               the first HTMLHead line MUST include a <BODY> tag.  Put
               whatever else you want in subsequent lines, but keep in
               mind the placement of this code in relation to the title
               and other aspcects of the web page.  Some typical uses
               are to change the page colors and posibly add a corporate
               logo (graphic) in the top right.  If not specified, a
               default <BODY> tag is used that defines page color, text
               color and link colors (see "sample.conf" file for example).

HTMLPost       This keyword defines HTML code that is placed after the
               title and "summary period/generated on" lines, just before
               the initial horizontal rule <HR> tag.  Normally this keyword
               isn't needed, but is provided in case you included a large
               graphic or some other weird formatting tag in the HTMLHead
               section that needs to be cleaned up or terminated before the
               main report section.

HTMLTail       This keyword defines HTML code that is placed at the bottom
               right side of the report.  It is inserted in a <TABLE> section
               between table data <TD>..</TD> tags, and is top and right
               aligned within the table.  Normally this keyword is used to
               provide a link back to your home page or insert a small
               graphic at the bottom right of the page.

--------------------------------------------------------------------------


Notes on Log Files
------------------

The Webalizer supports CLF log formats, which should work for just
about everyone.  If you want User Agent or Referrer information, you
need to make sure your web server supplies this information in it's
log file, and in a format that the Webalizer can understand.  While
The Webalizer will try to handle many of the subtle variations in
log formats, some will not work at all.   Most web servers output
CLF format logs by default.  For Apache, in order to produce the
proper log format, add the following to the httpd.conf file:

LogFormat "%h %l %u %t \"%r\" %s %b \"%{Referer}i\" \"%{User-agent}i\""

This instructs the Apache web server to produce a 'combined' log
that includes the referrer and user agent information on the end of
each record, enclosed in quotes (This is the standard recommended
by both Apache and NCSA).   Netscape and other web servers have
similar capabilities to alter their log formats.


Notes on Referrers
------------------

Referrers are weird critters... They take many shapes and forms, which makes
it much harder to analyze than a typical URL, which at least has some
standardization.  What is contained in the referrer field of your log
files varies depending on many factors, such as what site did the referral,
what type of system it comes from and how the actual referal was generated.
Why is this?  Well, because a user can get to your site in many ways... They
may have your site bookmarked in their browser, they may simply type your
sites URL field in their browser, they could have clicked on a link on some
remote web page or they may have found your site from one of the many search
engines and site indexes found on the web.  The Webalizer attempts to deal
with all this variation in an intelligent way by doing certain things to
the referrer string which makes it easier to analyze.  Of course, if your
web server doesn't provide referrer information, you probably don't really
care and are asking yourself why you are reading this section...

Most referrer's will take the form of "http://somesite.com/somepage.html",
which is what you will get if the user clicks on a link somewhere on the
web in order to get to your site.  Some will be a variation of this, and
look something like "file:/some/such/sillyname", which is a reference from
a HTML document on the users local machine.  Several variations of this can
be used, depending on what type of system the user has, if he/she is on
a local network, the type of network, etc...  To complicate things even
more, dynamic HTML documents and HTML documents that are generated by
cgi scripts or external programs produce lots of extra information which
is tacked on to the end of the referrer string in an almost infinate number
of ways.  If the user just typed your URL into their browser or clicked on
a bookmark, there won't be any information in the referrer field and will
take the form "-".

In order to handle all these variations, The Webalizer parses the referrer
field in a certain way.  First, if the referrer string begins with "http",
it assumes it is a normal referral and converts the "http://" and following
hostname to lowercase in order to simplify hiding if desired.  For example,
the referrer "HTTP://WWW.MyHost.Com/This/Is/A/HTML/Document.html" will become
"http://www.myhost.com/This/Is/A/HTML/Document.html".  Notice that only the
"http://" and hostname are converted to lower case... The rest of the
referrer field is left alone.  This follows standard convention, as the
actuall method (HTTP) and hostname are always case insensitive, while the
document name portion is case sensitive.

Referrers that came from search engines, dynamic HTML documents, cgi
scripts and other external programs usually tack on additional information
that it used to create the page.  A common example of this can be found
in referrals that come from search engines and site indexes common on the
web.  Sometimes, these referrers URL's can be several hundred characters
long and include all the information that the user typed in to search for
your site.  The Webalizer deals with this type of referrer by stripping
off all the query information, which starts with a question mark '?'.
The Referrer "http://search.yahoo.com/search?p=usa%26global%26link" will
be converted to just "http://search.yahoo.com/search".

When a user comes to your site by using one of their bookmarks or by
typing in your URL directly into their browser, the referrer field is
blank, and looks like "-".  Most sites will get more of these referrals
than any other type.  The Webalizer converts this type of referral into
the string "- (Direct Request)".  This is done in order to make it easier
to hide via a command line option or configuration file option.  This is
because the character "-" is a valid character elsewhere in a referrer
field, and if not turned into something unique, could not be hidden without
possibly hiding other referrers that shouldn't be.


Notes on Character Escaping
---------------------------

The HTTP protocol defines certain ways that URL's can look and behave.  To
some extent, referrer fields follow most of the same conventions.  Character
escaping is a technique by which non-printable or other non-ASCII (and even
some ASCII) characters can be used in a URL.  This is done by placing the
Hexdecimal value of the character in the URL, preceed by a percent sign '%'.
Since Hex values are made up of ASCII characters, any character can be
escaped to ensure only printable ASCII characters are present in the URL.
Some systems take this concept to the extreme and escape all sorts of stuff,
even characters that don't need to be escaped.  To deal with this, The
Webalizer will un-escape URL's and referrers before being processed. For
Example, the URL "/www.mrunix.net/%7Ebrad/resume.html" is the same URL as
"/www.mrunix.net/~brad/resume.html", a very common form of a URL to access
users web pages.  If the URL's were not un-escaped, they would be treated as
two seperate documents, even though they are really one and the same.


Random thoughts on DNS lookups
------------------------------

A lot of people have asked for the ability to do DNS lookups within
The Webalizer, as their servers don't do the reverse mapping.  My
usual response is that there are far too many such programs and
scripts that will process a log file doing just that.  One such
program is 'logresolve' that comes with the Apache distribution
(in the /support directory of the source distribution).  On my
machines, there are several programs that process log files at
the end of each month, and it is much easier to do the reverse
DNS lookups one time only, instead of multiple times in seperate
places.  Reverse DNS lookups may be included in a future release,
but I don't guarantee it.


Log files and The Webalizer
---------------------------

Most sites will choose to have The Webalizer run from cron at specified
intervals.  Care should be taken to ensure that data is not lost as a
result of log file rotations.  A suggested practice is to rotate your
web server logs at the end of each month as close to midnight as possible,
then have The Webalizer process the 'end of month' log file before running
statistics on the new, current log.  On our systems, a shell script called
'rotate_logs' is run at midnight, the end of each month.  This script file
looks like:

------------------------- file: rotate_logs ------------------------------
#!/bin/sh

# halt the server
kill `cat /var/lib/httpd/logs/httpd.pid`

# define backup names
OLD_ACCESS_LOG=/var/lib/httpd/logs/old/access_log.`date +%y%m%d-%H%M%S`
OLD_ERROR_LOG=/var/lib/httpd/logs/old/error_log.`date +%y%m%d-%H%M%S`

# make end of month copy for analyzer
cp /var/lib/httpd/logs/access_log /var/lib/httpd/logs/access_log.backup

# move files to archive directory
mv /var/lib/httpd/logs/access_log `echo $OLD_ACCESS_LOG`
mv /var/lib/httpd/logs/error_log  `echo $OLD_ERROR_LOG`

# restart web server
/usr/sbin/httpd

# compress the archived files
/bin/gzip $OLD_ACCESS_LOG
/bin/gzip $OLD_ERROR_LOG
------------------------- end of file ------------------------------------
 
This script first stops the web server using a 'kill' command.  Apache
keeps the PID of the server in the file httpd.pid, so we use it as the
argument for the kill.  Next, it defines some names for the backup files,
which are basically the name of the files with the date and time appended
to the end of them.  It then makes a copy of the log file, appended with
'.backup' in the log directory, moves the current log files to an archive
directory (/var/lib/httpd/logs/old) and restarts the server.  This setup 
allows the web server to be down for the minimum amount of time needed,
which is important for busy sites.  If you don't want to stop the server,
you can remove the inital 'kill' command, and replace the '/usr/sbin/httpd'
line with "kill -1 `cat /var/lib/httpd/logs/httpd.pid`" command instead,
On most web servers, this will cause a restart of the server and create
the new log files in the process...

At this point, we have made copies of the previous months logs,  the web
server is going about it's business as usual, and we have all the time in
the world to do any other additional processing we want.  The last two
lines of the script compress the archived logs using the GNU zip program
(gzip).  Remember, we still have a copy of the log which we can now run
The Webalizer on without having to do any further processing.

Next, we define two crontab entries.  The first runs the above 'rotate_logs'
script at midnight at the end of the month.  The second runs The Webalizer
on the '.backup' log file created above at 5 minutes after midnight.  This
gives other end of month processing jobs a chance to run so we don't bog
the system down too much.  If you have lots of end of month stuff going on,
you can change the timing to suit your needs.  The crontab entries look
something like:

------------------------- crontab entries --------------------------------
# Rotate web server logs and run monthly analysis
0 0 1 * *       /usr/local/adm/rotate_logs
5 0 1 * *       /usr/bin/webalizer -Q /var/lib/httpd/logs/access_log.backup
------------------------- end of crontab ---------------------------------

As you can see, the log rotations occur at midnight, and the analysis
is done at 5 minutes after.  Once you verify that The Webalizer ran
successfully, the access_log.backup file can be deleted as it isn't
needed any more.  If you need to re-run the analysis, you still have
the compressed archive copy that the shell script created.  In order
for the above analysis to work properly, you should have already
created an /etc/webalizer.conf configuration file sutable for your
site, or otherwise specify configuration options or a configuration
file on the crontab command line above.

If you want The Webalizer to be run more often than once a month, you
can specify additional crontab entries to do this as well.  Care should
be taken however to ensure that The Webalizer is not running when the
end of month processing above occurs, or unpredictable results may
happen (such as an inability to rotate the logs due to a file lock). 
The easiest way is to run it on the half hour with a crontab entry like:

30 * * * *      /usr/bin/webalizer


Language Support
----------------

Version 1.0x of The Webalizer added language support.  This
support is only provided at compile time in the form of an
include file containing all the strings used by The Webalizer.
The source distribution contains all language files that were
available at the time, with English being the default as
that is the only human language I speek fluently, and me
Espanol es muy malo.  Several people have already indicated
the desire to do translations into various languages, and as
I receive the language files, will make them available via
ftp at ftp://ftp.mrunix.net/pub/webalizer/lang.  Unless there
happens to be a binary distribution in the language you need,
you will need to grab the source distribution and compile the
program yourself. See the file INSTALL that comes in the source
distribution for information on how to use a language other than
English.

It should also be noted that the GD graphics library, used to
produce the inline graphics in the output HTML,  doesn't
support extended character sets, so if you are translating
the language file, you will no doubt encounter this problem.


Known Issues
------------

 o Memory Usage.  The Webalizer makes liberal use of memory for internal
    data structures during analysis.  Lack of real physical memory will
    noticably degrade performance by doing lots of swapping between memory
    and disk.  One user who had a rather large log file noticed that The
    Webalizer took over 7 hours to run with only 16 Meg of memory.  Once
    memory was increased, the time was reduced to a few minutes.

 o Proxy servers.  The Webalizer does not know how to deal with proxy
    records, which can cause some rather unexpected results.  If you
    do run the stats on a proxy servers log file, you get to see where
    your users have been visiting, as well as who has visited your
    server.  If someone knows of a reliable way to differentiate
    between normal log records and proxy ones, please let me know.

 o Performance.  The Hide*, Group*, Ignore* and IndexAlias configuration
    options can cause a performance decrease if lots of them are used.
    The reason for this is that every log record must be scanned for
    each item in each list.  For example, if you are Hiding 20 objects,
    Grouping 20 more, and Ignoring 5,  each record is scanned 46 times
    (20+20+5 + an IndexAlias scan).  On really large log files, this
    can have a profound impact.


Final Notes
-----------

A lot of time and effort went into making The Webalizer, and to ensure that
the results are as accurate as possible.  If you find any abnormalities or
inconsistant results, bugs, errors, ommisions or anything else that doesn't
look right, please let me know so I can investigate the problem or correct
the error.  This goes for the minimal documentation as well.  Suggestions
for future versions are also welcome and apperciated.
