perl regex help needed

Shell Programming and Scripting
1

Hi,

I want to validate strings in perl, the string may contains characters from a-zA-Z0-9 and symbols +-_.:/\

To validate such a string I computed a regex

if ($string =~ m/^[a-zA-Z0-9+-_.:\/\\]/) {
   print "valid";
} else {
   print "invalid";
}

but this regex also validates strings that contain other characters like @%^&*?.

I want to disallow strings which contain all other characters from above valid set.

plz help me

thanks,
zing_foru

2 
if ($string =~ m{\A[-+.:/\\a-zA-Z0-9_]+\z}) {
print "valid";
} else {
print "invalid";
}

and if you have Perl 5.14+, you could replace the first line with:

if ($string =~ m{\A[-+.:/\\\w]+\z}a) {
3

If an empty string is allowed, you could also negate the bracketed expression, rendering anchors unnecessary.

Regards,
Alister

4

thanks elixir_sinari, but this regex even not validating correct strings

e.g. "e:\test1\test2"
      "/usr/bin/test"
      "/var/log-test"
      "/usr/test bin/goak"

invalid strings e.g.

"e:\test\test?newdata"  #contains ? invalid symbol
"goak&<>"  #contains & <> invalid symbol
5

Try this Reg Exp

if($string =~ m/^[\w\s+-.:\/\\]*$/)
{
print "Valid";
}else {
print "Invalid"
}
6

thanks msabhi.. worked :slight_smile:

7

No, it did not. You just did not notice the error. A comma is still allowed even though it's not on your list of allowed characters.

The portions highlighted in red represent ranges of characters, not the three characters which appear literally in the regular expression.

Neither of those is correct. msabhi's example is simply less wrong. zing_foru's unintended range expression spans the entirety of the following list. msabhi's attempt spans the highlighted section.

From POSIX - Locale - collation sequence

<plus-sign>
<comma>
<hyphen>
<period>
<slash>
<zero>
<one>
<two>
<three>
<four>
<five>
<six>
<seven>
<eight>
<nine>
<colon>
<semicolon>
<less-than-sign>
<equals-sign>
<greater-than-sign>
<question-mark>
<commercial-at>
<A>
<B>
<C>
<D>
<E>
<F>
<G>
<H>
<I>
<J>
<K>
<L>
<M>
<N>
<O>
<P>
<Q>
<R>
<S>
<T>
<U>
<V>
<W>
<X>
<Y>
<Z>
<left-square-bracket>
<backslash>
<right-square-bracket>
<circumflex>
<underscore>

To fix your errors, refer to the perlre man page @ perldoc.perl.org

Regards,
Alister

1 Like
8

Rightly pointed out at Alister. I just overlooked "-" or somehow din't stress much while giving solution.."-" got a special meaning inside the character class just like "\","]" and "^"(only if used in the beginning)...Now here i believe escaping "-" should work...

if($string =~ m/^[\w\s+\-.:\/\\]*$/) 
{ print "Valid";
 }else { 
print "Invalid" 
}

Correct me if am wrong...

9

Have you checked my post again? I've edited some things :).

---------- Post updated at 10:34 AM ---------- Previous update was at 10:28 AM ----------

\w and \s, in Unicode context, would match a lot many unintended characters if these occur in the input.