Re: 5.4 regression: non-existent sub-sub keys now have values

On Fri, Nov 25, 2011 at 4:07 AM, Daniel Convissor < danielc@analysisandsolutions.com> wrote: > Hi Anthony: > > > isset($foo['bar'][1]['baz']) && > > is_array($foo['bar'][1]) > > > > You don't need to check each level. Only the one above the key you're > > looking at. > > Excellent thinking. One hitch... that goes fatal in 5.3 if $foo is a > string. > Here's how to write the test so it works the same way under 5.3 and 5.4: > yeah, and as you just noticed, one has to write less boilerplate code for the current 5.4 to guarantee that $array["foo"]["bar"]["baz"] is set and it is an element of an array. so I think we are just get to the point that everybody can agree that: 1, the string offset pseudo type and the chaining restriction sucks in 5.3, as it can fatal even an isset call 2, the int type juggling for string offsets is un-intuitive, and it can bite people. 3, because of 1 and 2, it writing variable checks which can handle type mismatch for complex datatypes requires much boilerplate code. The change introduced in 5.4 solves 1, doesn't touch 2, and makes 3 much easier to solve. So I think that we should keep the new behavior and focus on fixing 2, either via documentation, or changing the behavior (raise a notice, or don't accept non numeric values for string offsets at all). -- Ferenc Kovács @Tyr43l - http://tyrael.hu