johnnywang has asked for the wisdom of the Perl Monks concerning the following question:

Hi, I'm confused/puzzled by how LWP handles URI escapes vs. URI::Escape. Say I have a form parameter named "data" with value "12 34+56".
use strict; use URI::Escape: use LWP:Simple; my $res = get("".uri_escape("12 34+56")); #sen +t as: data=12%2034%2B12, both space and "+" are escaped. my $res2 = get(" 34+56")); # sent as: date=1 +2%2034+56, i.e. space is escaped, but "+" is not.
I was expecting get() to also escape "+" for me. After stepping through the code, I see that LWP::Simple::get() does call something like uri_escape:
# in URI::_init(C:/Perl/site/lib/ $str =~ s/([^$uric\#])/$URI::Escape::escapes{$1}/go; # where $uric is: # \;\/\?\:\@\&\=\+\$\,\[\]A-Za-z0-9\-_\.\!\~\*\'\(\)%
where you can see "+" is not escaped, plus a few others. URI::Escape::uri_escape is implemented as:
$text =~ s/([^A-Za-z0-9\-_.!~*'()])/$escapes{$1}/g;
which would escape "+". My question is why URI::Escape is escaping more than what get() does (via URI::_init)? or why get() doesn't do the "appropriate" escaping? is the best practice to escape everything before sending to get() rather than relying on its escaping? how about UserAgent::post? will it escape its values completely? or am I missing something? Thanks.