Boolean operators in PERL regexp? 1

[lenrobert]

Boolean operators in PERL regexp?

I am aware of OR ( | ), but is there logical NOT in the PERL regex syntax?

The task would be the following: to extract the relative links (i.e.. the href property of the "a" element) from an HTML file, even if it is not enclosed in quotation marks. This means I don't want to retrieve hyperlinks beginning with /, or # or javascript:

I would express the following string, and capture (or extract) the content of the second parenthesis.

( <a href=" OR <a href=) THEN NOT(/ OR # OR javascript: OR \s OR " ) THEN ( \s OR " )

The best regexp I could do is this, but it does not handle the case of / # javascript: etc.

/(<a href="|<a href=)([^"]*?)(\s|")/gi)

Does anyone know the answer, and help me? Thanks in advance,

Robert

 

[mikevh]

Please post an example of your input and your expected output.

 

[mlibeson]

if (! (0 == 1)) { print "True\n"; }


Michael Libeson

 

[ishnid]

You can read up on `negative lookaheads' in perlretut (google for it or type `perldoc perlretut' on your command line). Here's a simplified example of its use:

my $string = '<a href="blah.html">';
if ( $string =~ /<a href="(?!#|\/|javascript|\s)([^"]+)">/ ) {
   print "Matched $1\n";
}
else {
   print "No match\n";
}

 

[lenrobert]

Ishnid, thank you for your answer it works, but it doesn't handle the case, when there are no quotation marks. On Google pages for example it is often the case. I have read about lookahead and lookbehind assertions, the only very-very big problem is that they are non-capturing.

Mikevh here is a code. The expected output specified in the comments.

The case of / # is OK, but if I would like to use the elegance of regexps, I should sacrifice either javascript: or the case when there are no quotation marks.

my @strings;

# These should not match
$strings[0] = '<a href="/slash.htm">';
$strings[1] = '<a href="#name">';
$strings[2] = '<a href="javascript:func()">';

# These should match
$strings[3] = '<a href=noquotemark.htm>';
$strings[4] = '<a href="quotemark.htm">';

my $i = 0;

# This is just a title row for the output
printf "%-6s %-13s %-20s %s", "ROW", "Y/N", "\$2", "\$3\n\n"; 

for my $element (@strings)
  {
    $element =~ /(<a href="|<a href=)([^"\/#]+)(\s|"|>)/gi;
    if ( $2 ne '' ) 
      {
       printf "%-20s %-20s %-20s %s", "Row #$i Match", $2, $3, "\n";
      }
    else 
      {
        print "Row #$i No match\n";
      }
  $i++;
  }

 

[KevinADC]

I am sure this coul dbe done with a single regexp, but with two this is pretty easy to get accomplished:

# These should not match
$strings[0] = '<a href="/slash.htm">';
$strings[1] = '<a href="#name">';
$strings[2] = '<a href="javascript:func()">';

# These should match
$strings[3] = '<a href=noquotemark.htm>';
$strings[4] = '<a href="quotemark.htm">';
$strings[5] = '<a href="">';# <--added an unusual string for testing

my $i = 0;

# This is just a title row for the output
printf "%-6s %-13s %-20s %s", "ROW", "Y/N", "\$2", "\$3\n\n";

for my $element (@strings)
  {
    $element =~ m/<a href="*([^"].+?)"*>/gi;
    if ($1 && $1 !~ m!^#|/|javascript!i)
      {
       printf "%-20s %-20s %-20s %s", "Row #$i Match", $1, "\n";
      }
    else
      {
        print "Row #$i No match\n";
      }
  $i++;
  }

You may never get this to work 100% without further checking though.

 

[ishnid]

The regexp I posted was only to demonstrate the negative lookahead - it wasn't intended to do everything you're trying to do, as I felt that would over-complicate the example.

WRT capturing, because lookaheads are zero-width, there's technically nothing to capture. They're effectively an assertion that a particular condition exists at a certain point. In the example I posted previously, the lookahead is just checking that it's not followed by `#', `/', `javascript' or a space. Once that condition is true, the regexp continues matching and captures everything between the quotes into $1. If you were to change that to a positive lookahead (i.e. change ?! to ?=), again it's just checking to make sure it *is* followed by `#', `javascript' etc. If that's true, the entire string between the quotes will still be captured in $1, including whatever you're checking for.

I hope that all made sense.

This type of problem is why I normally recommend using a proper HMTL parser such as HTML::TokeParser::Simple
A proper parser will extract the href="whatever" bit for you (taking into account the type/lack of quotes used). Then it's just a case of testing what's in the attribute to see if it's what you're looking for.

 

[lenrobert]

KevinADC, Ishnid, thank you for your useful responses.