Finding a safe char with tr///d

You have a string, probably code and want to find an unused character to use as a delimiter for a quote-ish operator.

For me this arose using Filter::Simple on arbitrary code.

The snippet uses tr/// backwards in a sense, the eval is required as tr's char sets are determined at compile time

TIMTOWTDI but it think this is nice, though perhaps not best.

$possibly_safe = "^!|~ŚĄŁ\0";
$mystery_str = '$a !~ /qr^some|thing^';
$mystery_str =~ tr!/!!d; # make eval interpolation safe
eval "(\$safe = \$possibly_safe) =~ tr/$mystery_str//d";
print $safe; # "ŚĄŁ\0"
[download]

Update: There's a bit of chicken-egg action here. Can't believe I didn't notice it for 6 months.
Oh, the shame.

The old code:

$possibly_safe = "^!|~ŚĄŁ\0";
$mystery_str = '$a !~ qr^some|thing^';
eval "(\$safe = \$possibly_safe) =~ tr/$mystery_str//d";
print $safe; # "ŚĄŁ\0"
[download]

Comment on Finding a safe char with tr///d Select or Download Code

Replies are listed 'Best First'.
Re: Finding a safe char with tr///d by BrowserUk (Patriarch) on Jan 29, 2003 at 00:51 UTC
Nice++. I could have used this recently in Re: Cleaning Data Between Specified Columns (the original version at the bottom). If you used `$possible_safe = join'', map chr, 0 .. 255;` your pretty much guarenteed to find a safe char to use. (You might need to temporarially disable utf-8 if your dealing with unicode stuff I guess.) Examine what is said, not who speaks. The 7th Rule of perl club is -- pearl clubs are easily damaged. Use a diamond club instead.	[reply] [d/l]
Re^2: Finding a safe char with tr///d by Aristotle (Chancellor) on Jan 31, 2003 at 04:21 UTC
Except you're in for some funny looks when what happens to be a "safe character" is an open paren, bracket, etc. `:-)` Makeshifts last the longest.	[reply]
Re: Re^2: Finding a safe char with tr///d by BrowserUk (Patriarch) on Jan 31, 2003 at 04:37 UTC
Ain't that what \Q and \E are for:)... I know, I know... you'd have to monkey with the regex as well. Silly me. On a related note, it good to know I'm not the only one who uses control chars as temporary placeholders in strings. I agree that 0x7f was a bad choice in a utf-8/unicode world. Using a BOM like "\xEF\xBB\xBF" might make some sense though. Examine what is said, not who speaks. The 7th Rule of perl club is -- pearl clubs are easily damaged. Use a diamond club instead.	[reply]