Perl正则表达式不起作用

Question

File mentioned below is used for providing access to code repository.

cat /home/user/file
@abc_dev_group = @rel_team ram_ji jack_daniel tom_alter
@DEF_release-25Apr_dev_group= @rel_team  @SDV_dev_group
@Ship_dev_group= udit_srinivasan himanshu_singhal  vishal_tripathi
@Ship_release-15Jan2010_group = @release_team
repo cust
    RW+ dev* = @And_dev_group
@And_dev_group = @rel_team  piyush_rastogi naveen_srivastava arpit_chandra

I need to search for user in lines that contain string starting with @ and ending with '_dev_group' For example, @Ship_dev_group

Such strings are followed by a space (optional), followed by an equal (=) sign and then user names. Once i search for a given user and find that that particular user is absent, i append that user's name in the line.

I wrote this code below but it doesn't seem to work. Actually it's working for some but failing for others. Tried many different variations but couldn't get it working for all.

use strict;
use warnings;

my ($FILE, $repo);
my $user_name = "john_doe";
my $input_file = "/home/user/file";
my @repository_list = ("abc", "DEF","Ship", "And");

open (FILE, "<", $input_file) || die "Could not open file [$input_file]: $!";
foreach my $repo (@repository_list) {
    if ( ! grep {/^\@${repo}_dev_group.*$user_name.*/} <FILE> ) {
        system ("sed -i \"s/^\\(\@${repo}_dev_group.*\\)/\\1 $user_name/\" $input_file");
    } else {
        printf "User entry already exists in file $input_file\n";
    }
}
close FILE || die "Could not close file [$input_file]: $!";

I also used print statement to see what's going on. It does show the correct (supposedly) regex and it also works on command line (Linux box)(though i know bash and Perl regex behave differently) but it's not working through this script.

UPDATE:
The script is now working. :) Please see the lines below where i have added comments:

open (FILE, "<", $input_file) || die "Could not open file [$input_file]: $!";
@file = <FILE>; # Read into array
foreach my $repo (@repository_list) {
    if ( ! grep {/^\@${repo}_dev_group.*$user_name.*/} @file ) { # grepped the array @file
        system ("sed -i \"s/^\\(\@${repo}_dev_group.*\\)/\\1 $user_name/\" $input_file");
    } else {
        printf "User entry already exists in file $input_file\n";
    }
}
close FILE || die "Could not close file [$input_file]: $!";

Why did it start working now? Why was the earlier approach not working? I'm really confused now.

Answer 1

I think I found the culprit: you have to use non-greedy quantifiers: so please use .*? instead of .* ... For example, in a string like @And_dev_group = @rel_team piyush_rastogi naveen_srivastava arpit_chandra , you have two and occurrences, so the greediness of the regex will match as much as possible, which is not what you want...

Answer 2

This will search for a given user name on each line and print the line containing it, and the group of user names associated with that line.

use strict;
use warnings;

my @file = ('@abc_dev_group = @rel_team ram_ji jack_daniel tom_alter',
'@DEF_release-25Apr_dev_group= @rel_team  @SDV_dev_group',
'@Ship_dev_group= udit_srinivasan himanshu_singhal  vishal_tripathi',
'@Ship_release-15Jan2010_group = @release_team',
'repo cust',
'    RW+ dev* = @And_dev_group',
'@And_dev_group = @rel_team  piyush_rastogi naveen_srivastava arpit_chandra');

my $user_name = 'jack_daniel';

my $capture;
foreach (@file){
    print "$_\n" if /\@.*_dev_group\s?=(.*$user_name.*)/g;
    ($capture) = $1;
}
print "$capture\n";

prints

Line containing username match:

@abc_dev_group = @rel_team ram_ji jack_daniel tom_alter

All the user names on that line:

 @rel_team ram_ji jack_daniel tom_alter

Answer 3

my @file; #load your file here

@file = ('@abc_dev_group = @rel_team ram_ji jack_daniel tom_alter',
'@DEF_release-25Apr_dev_group= @rel_team  @SDV_dev_group',
'@Ship_dev_group= udit_srinivasan himanshu_singhal  vishal_tripathi',
'@Ship_release-15Jan2010_group = @release_team',
'repo cust',
'    RW+ dev* = @And_dev_group',
'@And_dev_group = @rel_team  piyush_rastogi naveen_srivastava arpit_chandra');

my $user_name = "john_doe";

my @new_file = (); #rewrite lines into this

foreach my $line (@file) {
  if ($line =~ /^\@.+_dev_group\s?=\s?((.+\s?)+)$/) {
    my @users = ();
    push @users, split(/\s+/, $1);
    $line .= " $user_name" unless (grep {$_ eq $user_name} @users);    
  }

  push @new_file, $line;
}

.

foreach my $line (@new_file) {
  print "$line\n";
}

This last piece of code will print:

@abc_dev_group = @rel_team ram_ji jack_daniel tom_alter john_doe
@DEF_release-25Apr_dev_group= @rel_team  @SDV_dev_group john_doe
@Ship_dev_group= udit_srinivasan himanshu_singhal  vishal_tripathi john_doe
@Ship_release-15Jan2010_group = @release_team
repo cust
    RW+ dev* = @And_dev_group
@And_dev_group = @rel_team  piyush_rastogi naveen_srivastava arpit_chandra john_doe