perl-selenium: like() or $sel->like()?

Perl-Selenium also provides the standard methods of Test::More such as ok(), like(), is(), etc , also as object methods, e.g. $sel->like().

($sel is the selenium perl object)

Should these object methods be used at all? Which one is preferred?

I am puzzled because like() and $sel->like seem to behave differently in places. Here I assume text is indeed found on the page.

like( $found, qr /$text/, "found '$text' on page" ) ; # WORKS FINE

 # DOES NOT WOR开发者_如何学运维K, ERROR:    "doesn't look much like a regex to me."
$sel->like( $found, qr/$text/ , "found '$text' on page")

So it seems the pure like() method is preferable to $sel->like()?

It looks like Test::WWW::Selenium does not itself provide like method, but rather many methods like title_like or text_like.

That gives us either

like( $found, qr /$text/, "found '$text' on page" );

or

$sel->text_like('id', qr/$text/, "found '$text' on page");

Edit: There is like method available in Test::WWW::Selenium, but when dumped, it looks like this:

sub {
    my $tb = 'Test::More'->builder;
    return $tb->like(@_);
}

it means that by calling as $sel->like(...) you are adding one extra parameter ($sel) to the method call. The method is remnant of Test::Builder inheritance and it likely not supposed to be used directly.