Keeping logs and analyzing the data gathered for web servers is one of the key aspects of search engine optimization. Below is a function coded in php & mysql which is capable to store basicly any information avaliable from a client requesting data from your web server. To have your own web traffic log analyzer you need to have few things avaliable with your hosting package such as it should allow you to run server side scripts and you should have access to any kind of database. Right now i m posting a code that works with php & mysql. The method used below requires certain space on your hosting but it gives you full control over your logs

//Here is the mysql syntax to create a database where we will store traffic 
information for our web pages. The fields listed below are if defined by the browser 
requested url,request date and time, parameters from Get request, parameters from POST
request, request method to web server, client language, client user-agent,client cookie,
referer and other fields if defined bytcp/ip clients ip address, clients 
connetion port. After creating the table we will code a function that will store the
data we need to analyze afterwards. Function uses global variables defined in php
itself as parameters , parses them and then stores them in database for generating
reports later on. 


sql commdand line
CREATE TABLE logstats
(
	recid int(11) NOT NULL auto_increment, 
	logrequestlanguage varchar(255) default NULL,
	logrequestuseragent varchar(255) default NULL, 
	logrequestcookie varchar(255) default NULL, 
	logrequestremoteaddr varchar(255) default NULL, 
	logrequestremoteport varchar(255) default NULL,
	logrequestmethod varchar(255) default NULL,
	logrequesturl varchar(255) default NULL,  
	logrequesttime varchar(255) default NULL, 
	logrequestgetdata text,  
	logrequestpostdata text, 
	logtime datetime default NULL,  
	logreferer varchar(255) default NULL, 
	PRIMARY KEY  (recid)
) ENGINE=InnoDB default charset=utf8;




//And here is the php syntax 
to create a function to parse the data received and store it into the database 
table we ve just created. 


library.php
function logstats($SERVERVARIABLES,$GETDATA,$POSTDATA)
{


	global $db;
//print_r($SERVERVARIABLES);
$ulang=$SERVERVARIABLES['HTTP_ACCEPT_LANGUAGE']; 
$uagent= $SERVERVARIABLES['HTTP_USER_AGENT']; 
$ucookie=(isset($SERVERVARIABLES['HTTP_COOKIE'])?$SERVERVARIABLES['HTTP_COOKIE']:""); 
$uaddr=$SERVERVARIABLES['REMOTE_ADDR']; 
$uport=$SERVERVARIABLES['REMOTE_PORT']; 
$umethod=$SERVERVARIABLES['REQUEST_METHOD']; 
$uurl=$SERVERVARIABLES['REQUEST_URI']; 
$utime=$SERVERVARIABLES['REQUEST_TIME'];
$ureferer=(isset($SERVERVARIABLES['HTTP_REFERER'])?$SERVERVARIABLES['HTTP_REFERER']:"");

$uget="";

foreach($GETDATA as $key=>$val)
$uget.=$key.":".$val."||";

$upost="";
foreach($POSTDATA as $key=>$val)
$upost.=$key.":".$val."||";



$utime=getdate($utime);
$utimetmp=$utime['year']."-".$utime['mon']."-".$utime['mday']." ".$utime['hours']."
:".$utime['minutes'].":".$utime['seconds'];
$utime=$utimetmp;



$q="insert into logstats (recid,logrequestlanguage,logrequestuseragent,logrequestcookie ,logrequestremoteaddr,logrequestremoteport ,logrequestmethod ,logrequesturl,logrequesttime,logrequestgetdata,logrequestpostdata,logtime,logreferer) values(null,";


$q.="'".$ulang."',";  $q.="'".$uagent."',"; $q.="'".$ucookie."',";
$q.="'".$uaddr."',";  $q.="'".$uport."',";  $q.="'".$umethod."',";
$q.="'".$uurl."',";   $q.="'".$utime."',";  $q.="'".$uget."',";
$q.="'".$upost."',";  $q.="now(),";			$q.="'".$ureferer."')"; 

$r=mysql_query($q);
if(!$r)
{
	return 0;
}


if(!mysql_affected_rows())
{
	return 0;
	
}
return 1;

}



//Right now we need to call the function above
from a php file also it is a good habit to store your functions in a diffrent file
in php to create a library so i ll assume the function above is saved ina a php file
called library.php


test.php

  include('library.php');
  echo "welcome to my site. I am logging everything
  for security and search engine optimization reasons.";

  logstats($_SERVER,$_GET,$_POST);



//So now we have all the information 
we need in our database. Whats left is that we need to processthe data so
it will be something meaningfull for us. Like generating reports for
daily visits.Or anything else we just need to add new functions to our
table. I dont want to get in too deep into programming so i ll create functions and 
sql 1 by 1




select logrequestremoteaddr, count(*) from logstats where date_sub(now(),interval
1 DAY) > logtime group by (logrequestremoteaddr) ;

//The query above will return ip adresses 
of visitors and the number of requests they have made to your web site today.
What we have to do is to create new functions to parse the data. 
(Note: since this project is under development i ll will make additional 
updates on the script till it is a standalone web application.)





library.php
function gettodaysoverview()
{
	
	global $db;
	$retarr=NULL;
	$q="select logrequestremoteaddr as ipaddr, count(*) as numberofvisits from logstats
	where date_sub(now(),interval 1 DAY) > logtime group by (logrequestremoteaddr)";
	$r=mysql_query($q);
	if(!$r)
	{
		return NULL;
	}
	if(!mysql_num_rows($r))
	return 0;

	while($rows=mysql_fetch_array($r))
	{
	 $retval[]=$rows;

	}
	return $retarr;

}



function getnumberofUniquevisitstoday()
{
	
	global $db;
 $retval="";
 $q="select count(*) as total from (select logrequestremoteaddr, count(*) from logstats where date_sub(now(),interval 1 day) < logtime group by logrequestremoteaddr) drvtable ";

//echo $q;

 $r=mysql_query($q);

	if(!$r)
	{
		return NULL;
	}
	
	if(!mysql_num_rows($r))
	return 0;

	while($rows=mysql_fetch_array($r))
	{
		$retval=$rows['total'];

	}
	return $retval;
	
}


function getnumberofUniquevisitsmonth()
{
	
global $db;
 $retval="";
 $q="select count(*) as total from (select logrequestremoteaddr, count(*) from logstats where date_sub(now(),interval 1 month) < logtime group by logrequestremoteaddr) drvtable ";

//echo $q;

 $r=mysql_query($q);

	if(!$r)
	{
		return NULL;
	}
	
	if(!mysql_num_rows($r))
	return 0;

	while($rows=mysql_fetch_array($r))
	{
		$retval=$rows['total'];

	}
	return $retval;
	
}


function getnumberofUniquevisitsweek()
{
	
	global $db;
 $retval="";
 $q="select count(*) as total from (select logrequestremoteaddr, count(*) from logstats where date_sub(now(),interval 1 week) < logtime group by logrequestremoteaddr) drvtable ";

//echo $q;

 $r=mysql_query($q);

	if(!$r)
	{
		return NULL;
	}
	
	if(!mysql_num_rows($r))
	return 0;

	while($rows=mysql_fetch_array($r))
	{
		$retval=$rows['total'];

	}
	return $retval;
	
}


function getnumberofUniquevisitsyear()
{
	
	global $db;
 $retval="";
 $q="select count(*) as total from (select logrequestremoteaddr, count(*) from logstats where date_sub(now(),interval 1 year) < logtime group by logrequestremoteaddr) drvtable ";

//echo $q;

 $r=mysql_query($q);

	if(!$r)
	{
		return NULL;
	}
	
	if(!mysql_num_rows($r))
	return 0;

	while($rows=mysql_fetch_array($r))
	{
		$retval=$rows['total'];

	}
	return $retval;
	
}



//Now that we have some functions we can
use them to parse the data and have some stats finnaly :). The code 
below will return unique ips on periodic bases on dynamic ip system 
it would only be assumption to categorize ips as unique so a timestamp
should be used as base standart to differ unique ips and unique users or it
can be grouped by daily bases. I will decide on that while developing this
application. Same problem also occurs on every log and traffic analyzers 
even google analytics can not solve that problem. Even if you dont take ip
as only unique index even user-agent variables or any header sent by any 
browser can  fool analyzers. Only if you monitor tcp/ip packets and compare 
mac adresses you may get an accurate result but thats also relative to the 
time being analyzed. 1000 users behind a proxy could visit your site and you
would see only 1 unique visitor or 1 user can visit your site 1000 times in
2 hours. The structure of your web site and content you provide can be used 
to determine some parameters to define what is unique visit and what is not.



analytics.php

	include("dbo.php");
	include("library.php");


	echo "test<br />";
	 logstats($_SERVER,$_GET,$_POST);


	
	echo "<br /> Number of unique visits Today:".getnumberofUniquevisitstoday()."<br />";
	echo "<br /> Number of unique visits this week:".getnumberofUniquevisitsweek()."<br />";
	echo "<br /> Number of unique visits this month:".getnumberofUniquevisitsmonth()."<br />";
	echo "<br /> Number of unique visits this year:".getnumberofUniquevisitsyear()."<br />";





// Ok so we will define unique visits based on daily hits from an ip since it 
seems like the most logical approach to get an accurate result. After all 
Web pages are ranked by their daily visit rate. So we have to modify our 
sql syntaxes in our functions and add a column to group by syntax



select count(*) as total from (select logrequestremoteaddr, count(*) from logstats where date_sub(now(),interval 1 year) < logtime group by logrequestremoteaddr,dayofyear(logtime)) drvtable



// Now we have a working date based counter. But knowing how many unique
visitors our site has doesnt help us enough on our efforts to increase our
search engine visibility rate. But hey we have all the data we need in our 
database so we should get into details  we should be able to see who is
visiting our web site and for that we have to add new functions to our 
library.