preg_match_all

[widowmaker]

hiya, I'm having this problem. I've got a log file which looks similar to this:

 

2010/06/28 22:54:59 administrator	13	-1
2010/06/06 22:54:59 administrator	1	2
2010/06/06 22:55:04 user	1	1

 

I want to go thru the whole logfile and split each row into variables using this pattern:

 

$reg[0]=2010

$reg[1]=06

$reg[2]=28

$reg[3]=22

$reg[4]=54

$reg[5]=59

$reg[6]=administrator

$reg[7]=13

$reg[8]=-1

 

I'm using this code but it doesn't work for some reason:

 

$row = '2010/06/28 22:54:59 administrator	13	-1';
preg_match_all('/^(\d{4})\/(\d{2})\/(\d{2})\s(\d{2})\d{2})\d{2})\s(\w+?)\t(\d+?)\t(\d+)$/i', $row, $regs);
var_dump($regs);

 

thanx for any help

[widowmaker]

I get an empty array instead of filled:

 

array(10) {
   0 => array(0) 
   1 => array(0) 
   2 => array(0) 
   3 => array(0) 
   4 => array(0) 
   5 => array(0) 
   6 => array(0) 
   7 => array(0) 
   8 => array(0) 
   9 => array(0) 
}

 

any help? thanx

[.josh]

first thing i notice is looks like you aren't checking for negative numbers.

[salathe]

This would be much easier without regular expressions, there are tools in PHP dedicated to parsing this kind of (tab-separated value) file, or more useful (and simple) string-parsing functions to do the job manually.

 

Do you ultimately want the array to be like that you posted (in the first post) or would you like the date and time values kept together?

[widowmaker]

ok I don't mind, the date and time could be together, the most important for me is this part:

 

$reg[6]=administrator
$reg[7]=13
$reg[8]=-1

 

including the fact that $reg[8] can be non-positive

[salathe]

One way would be to use sscanf like:

 

$row = '2010/06/28 22:54:59 administrator    13    -1';
$reg = sscanf($row, '%d/%d/%d %d:%d:%d %s %d %d');

 

And another would be to use fgetcsv and specify tab as the delimiter character. (The function is named CSV for comma-separated values, but is equally happy with tab-separated values like your file).