Case sensitive replace for case insensitive search with regex

Question

Ok so I doing the following regex replace:

 Input: ([Ll])ocation
Output: \1abel

That replaces Location with Label and location with label. However, if I want to replace location with geocode and Location with Geocode, can I do that with a single regex?

I'm doing a search and replace in notepad++ at the moment, but I've had this problem in multiple regex dialects.

Answer 1

GNU Emacs can do it

Answer 2

Here you are:

#!/usr/bin/perl -w

$str1 = 'Location Loc10';
$str2 = 'location Loc10';

$str1 =~ s/(L(?{ $g = 'G'; }) | l(?{ $g = 'g'; }) )ocation/${g}eocode/x;
print "$str1\n"; # prints Geocode Loc10

$str2 =~ s/(L(?{ $g = 'G'; }) | l(?{ $g = 'g'; }) )ocation/${g}eocode/x;
print "$str2\n"; # prints geocode Loc10

N.B. This piece of code uses experimental 'evaluate any Perl code' zero-width assertion. It is here only because you asked about it, I would not recommend using such expressions, it is much better to do several replaces. And I really doubt that Notepad++ supports it.

Answer 3

In perl:

echo Location | perl -pe 's/(l)ocation/$1^lc($1)^"geocode"/ei'
Geocode

Answer 4

EditPad Pro will do that if you select the "Adapt Case" option while replacing. It's not really a regex feature, just a feature that's supported by some tools that do search and replace. Said tools are usually powered by regexes, but that's not required; in fact, I don't think regexes would be much help in implementing such a feature.