Why repeated matches...

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

Hi again,

What's wrong with the code below? I get repeated values for each school matched.

my $file = 'schools.txt';
open (FH, "$file") or die $!;

my @schools;

while (my $line = <FH>) {
   $line =~ m/\s([\w ]+? PRIMARY SCHOOL)/;
   push (@schools, $1) if $1;
}

print "@schools\n";

# Sample data from schools.txt

SCH
CODE,
ZONE SCHOOL NAME &
ADDRESS PRINCIPAL &
VICE-PRINCIPAL TELEPHONE,
FAX NUMBERS &
EMAIL ADDRESS 

----------------------------------------------------------------------
+----------
 
1744
North ADMIRALTY PRIMARY SCHOOL 
???? 
11 WOODLANDS CIRCLE 
SINGAPORE 739807 P : MDM LIM SOH LIAN 
    ??? 
VP: MR TAN MENG HUI 
    ???  Tel: 63620598 
Fax: 63627512 
ADMIRALTY_PS@MOE.EDU.SG  

----------------------------------------------------------------------
+----------
 
1738
North AHMAD IBRAHIM PRIMARY SCHOOL 
?????? 
10 YISHUN STREET 11 
SINGAPORE 768643 P : MISS FOONG YIN WEI 
    ???  Tel: 67592906 
Fax: 67592927 
AIPS@MOE.EDU.SG
[download]

Comment on Why repeated matches... Download Code

Replies are listed 'Best First'.
•Re: Why repeated matches... by merlyn (Sage) on Dec 13, 2003 at 17:48 UTC
You get repeats because $1 is not reset on a failed match. Your code should be more like: `while (...) { if (some match) { do something with $1 } }` [download] -- Randal L. Schwartz, Perl hacker Be sure to read my standard disclaimer if this is a reply.	[reply] [d/l]
Re: Why repeated matches... by Anonymous Monk on Dec 13, 2003 at 17:52 UTC
Once your regex succeeds, $1 holds its value until the next time the regex succeeds. You want to conditionalize your use of $1 to only when the regex succeeds, not to when $1 is true: `while (my $line = <FH>) { next unless $line =~ m/\s([\w ]+? PRIMARY SCHOOL)/; push (@schools, $1); }` [download]	[reply] [d/l]
Re: Why repeated matches... by delirium (Chaplain) on Dec 14, 2003 at 00:34 UTC
Less reformatting of code: `while (my $line = <FH>) { $line =~ m/\s([\w ]+? PRIMARY SCHOOL)/ \|\| next; push @schools, $1; }` [download] Or you could get sneaky: `@schools = grep {/\s([\w ]+? PRIMARY SCHOOL)/; $_ = $1} <FH>;` [download]	[reply] [d/l] [select]
Re: Re: Why repeated matches... by bart (Canon) on Dec 15, 2003 at 02:30 UTC
Sneakier still: `@schools = map /\s([\w ]+? PRIMARY SCHOOL)/, <FH>;` [download] map creates a list of matches, here consisting only of $1, when it matches, and an empty list when it fails to match.	[reply] [d/l]
Re: Why repeated matches... by carric (Beadle) on Dec 14, 2003 at 00:25 UTC
Here is another possile way to write it: `#!/usr/bin/perl -w use strict; my $file=shift; # get the name of file you # you are passing to script open (FH,$file) \|\| die "Can't open - $!\n"; my @schools; while(<FH>) { if(//\s([\w ]+? PRIMARY SCHOOL)/) { push(@schools, $1); } } close FH;` [download]	[reply] [d/l]