Re: Replace text in file

[Yaerox]

Maybe another possible way for you could be using perl native function substr() instead of using regular expressions. Laurent_R already explained those above ;-)

[Laurent_R]

Yes, that could be used, but the problem I see with substr here is that we cannot be sure whether the version number has one or two or several digits, so you need further processing for determining that. To me, a regex substitution is simpler in this context.

[kcott]

While I probably wouldn't use substr for this, I don't really see the number of (before or after) digits as being a problem. Here's a (multi-line) one-liner to explain:

$ perl -Mstrict -Mwarnings -E '
    my $find = q{#define MYVERSION };
    my $flen = length $find;
    my @strings = (
        q{#define MYVERSION 11},
        q{#define MYVERSION 9},
        q{#define MYVERSION 99}
    );
    for (@strings) {
        say;
        substr($_, $flen) = substr($_, $flen) + 1;
        say;
    }
'
#define MYVERSION 11
#define MYVERSION 12
#define MYVERSION 9
#define MYVERSION 10
#define MYVERSION 99
#define MYVERSION 100

Anyway, that's all fairly academic: I'm pretty sure I would have reached for something like s/^($find)(\d+)/$1.($2+1)/e in the first instance.

[Yaerox]

"we cannot be sure whether the version number has one or two or several digits"

Totally agree with you on that point.

Regards