Regex and the characters case
Okay, I got a rather simple one (at least seems simple). I have a multi lined string and I am just playing around with replacing different words with something else. Let me show you...
#!/usr/bin/perl -w
use strict;
$_ = "That is my coat.\nCoats are very expensive.";
s/coat/Hat/igm;
print;
The output would be
That is my Hat
Hats are very expensive...
The "hat" on the first line shouldn't be capitalized. Are there any tricks that can make the casing开发者_开发百科 compliant with how english is written? Thanks :)
see how-to-replace-string-and-preserve-its-uppercase-lowercase
For more detail go to How do I substitute case insensitively on the LHS while preserving case on the RHS?
You can use the e
modifier to s///
to do the trick:
s/(coat)/ucfirst($1) eq $1 ? 'Hat' : 'hat'/igme;
For one, you should use \b
(word boundary) to match only the whole word. For example s/hat/coat/
would change That
to Tcoat
without leading \b
. Now for your question. With the flag /e
you can use Perl code in the replacement part of the regex. So you can write a Perl function that checks the case of the match and then set the case of the replacement properly:
my $s = "That is my coat.\nCoats are very expensive.";
$s =~ s/(\bcoat)/&same_case($1, "hat")/igme;
print $s, "\n";
sub same_case {
my ($match, $replacement) = @_;
# if match starts with uppercase character, apply ucfirst to replacement
if($match =~ /^[A-Z]/) {
return ucfirst($replacement);
}
else {
return $replacement;
}
}
Prints:
That is my hat.
Hats are very expensive.
This may solve your problem:
#!/usr/bin/perl -w
use strict;
sub smartSubstitute {
my $target = shift;
my $pattern = shift;
my $replacement = shift;
$pattern = ucfirst $pattern;
$replacement = ucfirst $replacement;
$target =~ s/$pattern/$replacement/gm;
$pattern = lcfirst $pattern;
$replacement = lcfirst $replacement;
$target =~ s/$pattern/$replacement/gm;
return $target;
}
my $x = "That is my coat.\nCoats are very expansive.";
my $y = smartSubstitute($x, "coat", "Hat");
print $y, "\n";
精彩评论