|
Re: Remove $
by FunkyMonk (Chancellor) on Aug 27, 2007 at 15:05 UTC
|
What makes you think it doesn't work?
$_ = 'db1$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_ClientID.sql'
+;
s*\$**g;
print;
#Output:
#db1/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_ClientID.sql
You're not using $ inside a double-quoted string are you? $/ is a valid Perl identifier (the input record separator).
update: double quotes stuff
update^2: gah. $/ --> input record separator
| [reply]
[d/l]
[select] |
|
|
I guess I am inside a double quoted string. I'm receiving the string from an Excel spreadsheet like this:
my $cellval = $Sheet->Cells($rownum,"B")->{'Value'};
chomp ($cellval);
As noted before it comes out like this:
db1$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_ClientID.sql
So how do I then convert that string to literal, remove the $, and flip back to interpreted? (that doesn't read right but I think you get the drift)
| [reply] |
|
|
I think there's something else going on. What output do you get if you place a print either side of the substitute? ie
my $cellval = $Sheet->Cells($rownum,"B")->{'Value'};
print "$cellval\n";
$cellval =~ s*\$**g;
print "$cellval\n";
| [reply]
[d/l]
[select] |
|
|
|
|
|
Re: Remove $
by SFLEX (Chaplain) on Aug 27, 2007 at 15:23 UTC
|
my $text2 = 'db1$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_Client
+ID.sql';
my $hmm = chr(36);
$text2 =~ s{\Q$hmm}{}gso;
That should do it...
Good Luck ^^ | [reply]
[d/l] |
|
|
That looks fabulous but I have no idea what it's actually doing. Can you explain it? Please type slowly. ;-)
| [reply] |
|
|
#!/usr/bin/perl
my $text2 = 'db1$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_Client
+ID.sql';
# chr(36) = $
# http://perldoc.perl.org/functions/chr.html
# It's another way to represent $
# my $hmm = chr(36);
my $hmm = '$';
# \Q Disable pattern metacharacters until \E
# http://perldoc.perl.org/perlreref.html#ESCAPE-SEQUENCES
# \Q makes it work =D
$text2 =~ s{\Q$hmm}{}g;
print "Content-type: text/html\n\n";
print "<html><h1>$text2</h1></html>\n";
same thing as
#!/usr/bin/perl
my $text2 = 'db1$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_Client
+ID.sql';
$text2 =~ s{\Q$}{}g;
print "Content-type: text/html\n\n";
print "<html><h1>$text2</h1></html>\n";
Is that good enough?
*Edited | [reply]
[d/l]
[select] |
Re: Remove $
by dogz007 (Scribe) on Aug 27, 2007 at 15:18 UTC
|
If there is no way to stop the interpolation of the $/ in your input, then perhaps you could localize it just within that context. For instance:
{
local $/ = '/';
$text = "db1$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_ClientID
+.sql";
print "$text\n";
}
Prints:
db1/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_ClientID.sql
Make sure that this exception doesn't flow over into the rest of your code. You probably don't want a screwed-up new-line character wandering around your script. | [reply]
[d/l]
[select] |
|
|
But there is a way... backslashing the $:
$text = "db1\$/SQL/GEMINI/StoredProcs/pboValidateClientRefNo_ClientID
If he can change the code to add { local $/ = '/'; ... }, he can change the code to add $.
I consider your advice very dangerous.
| [reply]
[d/l]
[select] |
|
|
I understand the danger. That's why I gave the warning at the end. However, notice the context of the string in which he's interested. It's a filepath, which means that it might not have been input by hand, but by some directory walker. In that case he may not be able to insert a '\' or catch the '$' before it interpolates. In that extreme case, he may choose to change the new-line character, but should only do so temporarily. Of course it's dangerous, but it's the bad-boy solutions that perl allows that make it such a wonderful language to use. Document it well, and be smart about it, and it should work fine.
| [reply] |
|
|
|
|
|