Re: private, protected, readonly, public

Hello Marcus, I've put together a simple test framework and a 18 test cases to start with. It's a stand-alone system that should work correctly on PHP 5 installs. Both CLI mode and Web Server mode work fine. NOTE: The tests I wrote are only testing standard functionality. We need to add tests to check out the new readable patch. -- View the tests online: http://open.izrm.com/~jason/php-readable -- Download a .tgz - just unpack it and it's ready to go: http://open.izrm.com/~jason/download/php-readable.tgz -- Get it with SVN: svn checkout svn://open.izrm.com/php-readable -- If anyone is interested in contributing test cases, please send them over. Thanks! -- Best regards, Jason mailto:jason@ionzoft.com private-outside-read.php private-outside-write.php private-subclass-read.php private-subclass-write.php private-thisclass-read.php private-thisclass-write.php protected-outside-read.php protected-outside-write.php protected-subclass-read.php protected-subclass-write.php protected-thisclass-read.php protected-thisclass-write.php public-outside-read.php public-outside-write.php public-subclass-read.php public-subclass-write.php public-thisclass-read.php public-thisclass-write.php testing-system-test.php Tuesday, May 16, 2006, 6:11:23 PM, you wrote: MB> Hello Jason, MB> write testcase, write in a way that they capture any usage you can MB> think of. Parameter parsing, reference handling returning, inheritance, MB> trying to circumvent the whole stuff, checking all error messages are MB> in place, overloading internal objects that have special handlers just MB> to hwat happens with them (dom, simplexml for instance). How do the MB> magic methods __get, __set, __unset, __isset interfere if at all.... MB> best regards MB> marcus MB> Tuesday, May 16, 2006, 11:59:08 PM, you wrote: >> Hello Andi, >> Your request for edge condition research is an excellent one. We've >> just been through a hellish couple weeks of QA failures (at my >> company) which just *underscore* your point. The last thing any of >> us needs is a broken PHP. >> That being said, is there anything I can do to help us find this >> stuff out? >> -- >> Best regards, >> Jason mailto:jason@ionzoft.com >> Tuesday, May 16, 2006, 5:41:31 PM, you wrote: AG>>> Where would readable be enforced? Would it try and prevent getting AG>>> references to it? Are there any internal functions/classes which need AG>>> fixing to honor readable? AG>>> I think these answers would really be helpful. AG>>> Thanks. AG>>> Andi AG>>> At 02:37 PM 5/16/2006, Marcus Boerger wrote: >>>>Hello Andi, >>>> >>>> that is why most here already switched to "public readable". >>>> >>>>best regards >>>>marcus >>>> >>>>Tuesday, May 16, 2006, 11:31:14 PM, you wrote: >>>> >>>> > I can't quite explain it but for me the ability to work-around >>>> > private with methods which are able to access the private variable, >>>> > is different than marking a property as read-only but it not being >>>> > read-only in all semantics. Probably because private variables do >>>> > often have getters and setters, whereas something which is marked as >>>> > read-only (like a harddrive) tends to be read-only always. >>>> >>>> > Andi >>>> >>>> > At 02:08 PM 5/16/2006, Zeev Suraski wrote: >>>> >>>However, the reason i write this mail is that you said there could be >>>> >>>problems. Well this is deply integrated in the handlers and they don't >>>> >>>let you out. In other words if this stuff is not working then the whole >>>> >>>PHP 5+ object model is broken. Or in other words, if this is broken alot >>>> >>>of other stuff regarding object handling is already broken. >>>> >> >>>> >>You're probably right about this one. You can already return a >>>> >>reference to a private variable today and change it. Andi - did you >>>> >>mean something else? >>>> >>>> >>>> >>>> >>>>Best regards, >>>> Marcus MB> Best regards, MB> Marcus