Re: Property Overloading RFC

Hello Andi, Saturday, August 27, 2005, 1:45:43 AM, you wrote: > At 04:27 PM 8/26/2005, Marcus Boerger wrote: >>Hello Andi, >> >>Saturday, August 27, 2005, 12:52:38 AM, you wrote: >> >> > At 11:32 AM 8/26/2005, Marcus Boerger wrote: >> >> > I don't see why the __get/__set/__isset/__unset methods themselves >> >> > can't check if the property exists and throw an exception if it >> >> > doesn't. I always do that in all my examples... >> >> >> >>Interesting, How do you do that? There is no __isset or __unset. >> >> > Did you check HEAD? >> >>When did this get sneaked in and why? And why didn't we make it like >>ArrayAccess with interfaces? And was this even being discussed and i >>was sleeping? >> >>I see dmitry did it: >>$ cvs log -r1.288 Zend/zend.h >>date: 2005/07/07 16:07:08; author: dmitry; state: Exp; lines: +5 -1 >>Fixed bug #33512 (Add missing support for isset()/unset() >>overloading to complement the property get/set methods) > I asked Dmitry to fix this as it seemed kind of buggy to me (bug > #33512). Thought everyone saw it. > It's not in an interface because __get/__set were introduced before > interfaces existed.... Ah yes, makes quite sense then and we're consistent for properties and array acccess though we don't use the same technology for them. Best regards, Marcus