habel_k has asked for the wisdom of the Perl Monks concerning the following question:

i am getting this error: search pattern not terminated...what could be the problem. I am trying to parse the apache access log:

eg of an entry: 

10.25.95.100 ab - [05/Aug/2003:12:00:30 -0700] "GET /creative/2|212727
+4-1;iframe? HTTP/1.1" 200 791

[download]
My code: 
#!/usr/bin/perl -w 
use strict; 


my $s = '10.25.95.100 ab - [05/Aug/2003:12:00:30 -0700] "GET /creative
+/2|2127274-1;iframe? HTTP/1.1" 200 791 "-" "Mozilla/4.0(compatible; M
+SIE 5.5; Windows 98)""TC1=1067055_5565+067055+0+30fc+3fb5+20742d+1; P
+WH=m2w/./o0.Nyvo0/I..d/I./I.; Xc294&2127030&1=1059866137&13384&18337;
+ Xc15&2127102&2=1060109088&13025&17516; Xc200&2127519&2=1060109135&13
+301&18215" "Imp: 294&18421&2127274&1&13384&1060110030" -'; 


my $LOG_PATTERN = q{(.*) \- \[(.*)\] \"(.*) (.*)\?(.*) HTTP\/(.*)\" ([
+0-9]*) ([0-9]*)}; 


print "\$s=$s\n"; 


my $var10=''; 


$s =~ /(.*) \- \[(.*)\] \"(.*) (.*)\?(.*) HTTP\/(.*)\" ([0-9]*) ([0-9]
+*) \"(.*)\" \"(.*)\"/; 


print "Trick 1: \$1=$1, \$2=$2, \$3=$3, \$4=$4, \$5=$5, \$6=$6, \$7=$7
+, \$8=$8, \$9=$9, \n", 


" \$\`=",$`," \$\'=",$',"\n"; 



my $len = @ARGV; 

if ($len <1){ 

print "Usage : perl script_name logfile\n"; 

exit(); 

} 


if (!(-e $ARGV[0])){ 

print $ARGV[0]," : not exists.\n"; 

exit(); 

}
 

open (SEM, "< $ARGV[0]") or die "Cannot open file $ARGV[0]\n";
while () 

{
 
if (my($ip,$date,$method,$url,$query,$protocol,$retcode,$byte)=($_ =~ 
+m/$LOG_PATTERN)) 
{ print "IP :",$ip,"\n";} 

}
 
close(SEM);

[download]
when I try to add a trailing /, I get the warning: Use of uninitialized value in pattern match (m//) at learn1.plx line 42

Replies are listed 'Best First'.
Re: search pattern not terminated
by GrandFather (Saint) on Jan 21, 2008 at 23:59 UTC

    It may be the missing '/' in:

            ($_ =~ m/$LOG_PATTERN))

    [download]
    Perl is environmentally friendly - it saves trees
[reply]
[d/l]
      thanks for your assistance. when i palce a trailing / I get the warning: Use of uninitialized value in pattern match (m//) at learn1.plx line 42
      Where should i be placing the trailing/ ?
[reply]

        Where did the code come from? I presume (given the level of your questions) that this is not your own code. There are some rather unusual aspects to the code and if you are not a reasonably competent Perl programmer I suspect you will continue to have trouble with it.

        For a start while () {...} is an unterminated loop, which is unlikely to be correct given that there is code following it.

        The match giving you your immediate trouble is against the default variable ($_), but nothing assigns a value to it, hence it is uninitialized (the immediate error). But it is not at all clear what value should be assigned to it.

        Perl is environmentally friendly - it saves trees
[reply]
[d/l]
        if (my($ip,$date,$method,$url,$query,$protocol,$retcode,$byte)=($s_=~ 
+m/$LOG_PATTERN/))
#                                                                     
+              ^

        [download]
[reply]
[d/l]
Re: search pattern not terminated
by ww (Archbishop) on Jan 22, 2008 at 02:32 UTC

    This does not capture all you wanted but perhaps you can build on this: 

    #!/usr/bin/perl
use strict;
use warnings;

my $log=q<10.25.95.100 ab - [05/Aug/2003:12:00:30 -0700] "GET /creativ
+e/2|2127274-1>;
                                             # note 1 (re above)
# print "\$log is $log \n";                  # note 2  

if ( $log =~ / 
    (\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}) # capture IP digits
    \s[a-z]{2,2}\s-\s\[                  # match but don't capture spa
+ce, dash, space
    (\d+\/[A-Z][a-z]{2,2}\/\d{4,4})      # capture the date
    :                                    # don't capture the perlimina
+ry colon
    (\d+:\d+:\d+\s-\d{4,4})              # capture time with colons &a
+mp; GMT offset
    (.*)                                 # dot-slash deprecated, but c
+apture to end
    /x  ) {                              # extended notation, end cond
+it, print
   print "\$1=$1, \$2=$2, \$3=$3, \$4 = $4 \n";  # note 3
} else {
   print "no matches\n";
}                                            # note 4

    [download]
    ... balance of code (Update: ...as corrected in other nodes and restoring the full original string to $log/$s)

    1 "$s" may be less than clear, a few hundred lines later; use a meaningful name for the $var
    2 I read this as OP's simple test to make sure the assignment above worked.
    3 It would be better to make this
     print "IP = $1, DATE = $2, TIME= $3"and so on, if you're merely testing your regex
    4 added the conditional: ALWAYS test captures before trying to use them

[reply]
[d/l]
[select]
Re: search pattern not terminated
by ww (Archbishop) on Jan 21, 2008 at 23:55 UTC

    It helps if you format your post properly (see Writeup Formatting Tips ) (Update OP updated without notice) and if you quote the entire output...

    Unmatched ) in regex; marked by <-- HERE in m/(.*) \- \(.*) <-- HERE \ "(.*) (.*)\?(.*) HTTP/(.*)" (0-9*) (0-9*) "(.*)" "(.*)"/ at (yourcode).pl line 7. Now, I admit, counting paren_matches and so forth, it's not immediately obvious to me, either, why that occurs, but wiser heads would have had more to work on had you investigated the norms here ( How do I post a question effectively? ) before posting.
[reply]
[d/l]
Re: search pattern not terminated
by poolpi (Hermit) on Jan 22, 2008 at 08:11 UTC
[reply]

Back to Seekers of Perl Wisdom