Re^2: Reading Regexes from File

OK, tried a slightly different approach. Still reading the regular expressions from file, but I'm using a :: to delimit the search pattern from the replacement text. Problem is, it's literally interpreting the replacement. For example this in the file:

(\d{4})(\d{2})(\d{2})::$2-$3-$1

When run this way in the code:

my ($search, $replace) = split "::", $rule; $value =~ s/$search/$replace/si;

Yields - $2-$3-$1

I wanted it to refactor the date into a MM-DD-YYYY format.

Comment on Re^2: Reading Regexes from File

Replies are listed 'Best First'.
Re^3: Reading Regexes from File by johngg (Canon) on Apr 12, 2007 at 22:57 UTC
I don't think that is ever going to work, although more experienced Monks may know better. The method I described is only good for replacing a pattern with simple text. You are getting the literal $2-$3-$1 because the perl interpreter sees the scalar `$replace` when it parses your code and interpolates it's contents as a literal string rather than seeing the "magical to regular expressions" `$1` etc. If you try to rectify things by `eval`'ing `$replace`, i.e `$value =~ s/$search/eval "$replace"/sie` then Perl interprets that as a sum and comes up with, for today's date, -2015. To do anything more fancy than a simple text replacement you will almost certainly have to take the `eval` approach suggested by duff and expanded on by ikegami. Cheers, JohnGG	[reply] [d/l] [select]

Replies are listed 'Best First'.

Re^3: Reading Regexes from File
by johngg (Canon) on Apr 12, 2007 at 22:57 UTC

$replace

$1

eval

$replace

$value =~ s/$search/eval "$replace"/sie

To do anything more fancy than a simple text replacement you will almost certainly have to take the eval approach suggested by duff and expanded on by ikegami.

Cheers,

JohnGG

[reply]
[d/l]
[select]