Replace character between slashes

http://photos-c.ak.fbcdn.net/hphotos-ak-as开发者_如何学编程h4/####_####_####_###_####_a.jpg
http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/wefawf/vawvaw/a####_####_####_###_####.jpg

I'm trying to create a Regex that replaces the 'a' in between the last '/' and '.jpg' with an 'n'.

I'm completely lost right now. I'm using Perl and from what I understand it should be along these lines:

s/^.*\/.*$/n/g

... except I'm not sure where to put the 'a'.

If there's guaranteed only one 'a', then Paul's answer will suffice, with a couple of small changes:

$foo =~ s|a([^/]*\.jpg)\Z|n$1|;

First, I have no '/' preceding the 'a'. It's sufficient to specify that there are no '/'s between 'a' and '.jpg', and that allows for both of your examples (with the '/' present it only matched the second of your examples). I also removed the '/g' modifier and added '\Z' to anchor the regex to the end of the string. We only need the last match.

If there may be multiple 'a's between the final '/' and '.jpg', things get only slightly trickier:

while ($foo =~ s|(/[^/]*)a([^/]*\.jpg\Z)|$1n$2|) { }

Basically, while there are still 'a's between the final '/' and '.jpg', keep swapping 'a's for 'n's one at a time.

$foo =~ s?/a([^/]*.jpg)?/n$1?g;

This replaces the first 'a' between the last '/' and '.jpg' with an 'n' (tested):

$ cat test.pl 
#!/usr/bin/env perl

use strict;
use warnings;

sub a_with_n {
    shift;
    s!(.*/.*)a(.*\.jpg)!$1n$2!;
    return $_;
}

my @test_strings = (
    'http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/####_####_####_###_####_a.jpg',
    'http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/wefawf/vawvaw/a####_####_####_###_####.jpg'
);

foreach (@test_strings) {
    printf("%s\n", a_with_n($_));
}


$ ./test.pl 
http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/####_####_####_###_####_n.jpg
http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/wefawf/vawvaw/n####_####_####_###_####.jpg

Another way of doing this replacement (matching multiple occurrences of "a"):

use strict;
use warnings;

while (<DATA>) {
    my $last = rindex ($_, "/");    # get index of last "/"
    substr ($_,$last) =~ s/a/n/g;   # change all "a" to "n" starting from last "/"
    print;
  }

__DATA__
http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/####_####_####_###_###a#_aa.jpg
http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/wefawf/vawvaw/a####_####_####_###_####.jpg

Output:

http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/####_####_####_###_###n#_nn.jpg
http://photos-c.ak.fbcdn.net/hphotos-ak-ash4/wefawf/vawvaw/n####_####_####_###_####.jpg