Does anybody know...

[Ondra]

Hello,
does anybody know what is this:

my $tag
$line =~ /<(\S+)[^>]*?>/;
$tag = $1 || ""; /*****exactly this***/

 

[Peter Hickman]

Hello,
does anybody know what is this:

my $tag
$line =~ /<(\S+)[^>]*?>/;
$tag = $1 || ""; /*****exactly this***/

Looks like a regex to match htlp tags and extract the body of the tag.

 

[Peter Hickman]

Hello,
does anybody know what is this:
my $tag
$line =~ /<(\S+)[^>]*?>/;
$tag = $1 || ""; /*****exactly this***/

Looks like a regex to match htlp tags and extract the body of the tag.

Or even HTML! My fingers are cold this morning

 

[Tad McClellan]

Subject: Does anybody know...

Please put the subject of your article in the Subject of your article.

does anybody know what is this:

my $tag
$line =~ /<(\S+)[^>]*?>/;
$tag = $1 || ""; /*****exactly this***/

I know what it is *not*. It is not Perl code, it has syntax errors.

It is a bug, you should never use the dollar-digit variables unless
you have first ensured that the match *succeeded*.

It is a bug since it tries to parse HTML with pattern matching
rather than with a real HTML parser. There are a bunch of examples
in the Perl FAQ of legal HTML that will break it...

The non-greediness serves no purpose, it is not needed.

It could be replaced with this:

my $tag = '';
$tag = $1 if $line =~ /<(\S+)[^>]*>/;


The

$tag = $1 || "";

is a common idiom for selecting a default value.

Note that it will do the wrong thing if the data contains "<0>"...

 

[Ondra]

Hello,
does anybody know what is this:

my $tag
$line =~ /<(\S+)[^>]*?>/;
$tag = $1 || ""; /*****exactly this***/

Looks like a regex to match htlp tags and extract the body of the tag.

That doesn't answer the OP's question.


$tag = $1 || "";


means "set $tag to the value of $1 if the value of $1 is true, otherwise
set $tag to the empty string".

Yes that's it. Thank you

 

[David Oswald]

Hello,
does anybody know what is this:

my $tag
$line =~ /<(\S+)[^>]*?>/;
$tag = $1 || ""; /*****exactly this***/

Honestly, it's a mistake.
You should never rely on $1 in any way unless you first ensure that the
match succeeded. It is entirely possible that $1 may still contain an
artifact of a previous pattern match, even if the most recent match failed.
That being the case, you cannot assume that $1 will be 'undef' or 'false' if
the most recent pattern match failed, for, if there were previously
successful pattern matches with capturing parens, $1 will still contain the
results of that successful capture.

You really should be checking the success of the match before fiddling with
$1, et. al.