wp_specialchars doesn't capture pre-encoded entities correctly

[westi]

Working on updating some of the wordpress-tests test cases for the formatting code to work with the new special chars and it looks like the capture of pre-encoded entities doesn't work correct.

Test Cases are here:
​http://svn.automattic.com/wordpress-tests/wp-testcase/jacob/TestFormatting.php

The one that fails is:

    function test_ignores_existing_entities() {
        $source = '& £  &';
        $res = '& £  &';
        $this->assertEquals($res, wp_specialchars($source));
    }

[westi]

Proposed fix to wp_specialchars so that it catches all existing entities

[westi]

CC: sambauers as this may affect bbPress as well.
Once committed will need to merge into BackPress trunk

Related ticket is #8767

[sambauers]

Totally true, well spotted...

I've made a minor change to the patch though, the regexp is case insensitive, so there is no need for the upper case chars.

[westi]

(In [11338]) Ensure that wp_specialchars captures all pre-encoded entities correctly when preventing double encoding. Fixes #9821.