-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Emiliano.
- --On 27.03.2002 14:38:09 +0100 Emiliano wrote: >> mgd_snippet? mgd_include? I'm beginning to suspect that I've >> discovered only the tip of an iceberg. As long as the >> eval(mgd_preparse()) Construct stays the offical replacement for the >> eval() thing, I'll be fine... *g* > > I will unless there's a pressing reason to change it. But nost users > won't see this in action, it'll just be something midgard does so it > can do its job. Right and wrong. Many more complex websites rely on a mechanism that can dynamically evaluate Midgard Constructs. All of the sites we built here use the eval() thing for example. MidCOM will rely heavily on it. It is no concept for a John-Doe-Standard-User, but most certainly one for a more advanced Midgard Developer. >> >> [Documentation...] >> >> >> > Yes. They would. >> >> The old problem. If you could give me some description about all >> these missing things (including a short overview over the deprecated >> ones and how they should be replaced) I'd be glad to write some docs >> about it. (Not that I have the slightest Idea how to incorporate >> this in the Main Documentation...) > > [... Some things about this whole parser problem] > Thanks, that's some start for me. I'll think about something and let you know. >> > Not possible without re-implementing the Zend eval (which is what >> > mgd_eval previously did). eval works in the scope of the code that >> > calls it. A function, _any_ function, works in its local scope >> > unless you explicitly name the variables you want to be globals, >> > and even then it only works in the global scope, not the calling >> > scope. >> >> Which is definitly too much work... > > And trust me to try. Ow. Parsing bits of string to see what could be a > variable and then declaring it global. It was a total mess (as you > probably guessed), and even if you could get it perfect (and you can't > unless you can hook into the Zend parser, trust me) it would still > _only_ work in the global scope, so it would still not be a proper > replacement for eval. That sound really scary. Definitly not a thing we want to do. > The reason why I'm doing all this is that I want to strongly simplify > Midgard to make it more robust and maintainable, and avoid, unless > utterly, utterly necesary, to reimplement (or worse, alter) > functionality that is present in Zend. Which is a very important part. Midgard suffered long enough from a poorly maintainable Architecture. If the codebase can really stabilize, this would be a large step forward. Live long and prosper! Torben Nehmer - -- Torben Nehmer, Munich, Germany http://www.nathan-syntronics.de, mailto:[EMAIL PROTECTED] PGP Public Key ID on wwwkeys.(de.)pgp.net: 0x7E9DE456 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Weitere Infos: siehe http://www.gnupg.org iD8DBQE8odqbJPh4Kn6d5FYRArGWAJ9MgddtiR9TBSzTUAltJV0dFMCqggCdENGQ xq84pareG8FV8KTyjmso9po= =e0uQ -----END PGP SIGNATURE----- --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
