|
Graham Dumpleton
grahamd at dscpl.com.au
Thu Dec 8 17:43:53 EST 2005
Radek =?utf-8?q?Barto=C5=88?= wrote .. > > If using mod_python.publisher, as: > > > > index = MyClass > > > > you should I believe get NotFound as it will not automatically create > > instances > > of classes for you if a class type is referenced. I know Nicolas hand > > it in mind > > to perhaps support this, but personally I don't think it is a good > > idea as it could > > introduce other problems where people think methods are not > > accessible but > > may actually. The authorisation schemes as written also would not > > work for the > > instance created. Ie., authorisation could only be done at file scope. > > With file test.py: > > class MyClass: > def __init__(self): > pass # some initialization > def __call__(self, req): > return "whole class" > def method1(self, req): > return "method1" > def method2(self, req): > return "method2" > > index = MyClass() > > and URLs http://locahost/test/ http://localhost/test/index/method1 and > http://localhost/test/index/method2 I didn't get any Not Found error. That is different to what you originally had and not what I was talking about. It may have been a typo, but I believe you had at one point: index = MyClass not: index = MyClass() There is a big difference. In one case you are publishing the class itself and in the other and instance of the class. My description above was talking about publication of a class verses publication of an instance of the class. > And by the way it does all I want. Here is some example of using this > approach: > > from mod_python import Session,util > > class MyClass: > def __init__(self): > pass # some initialization > def __call__(self, req): > session = Session.Session(req) > # lazy evaluation > if session.has_key('logged') and session['logged']: > return """ > <HTML> > <HEAD> > </HEAD> > <BODY> > Logged > </BODY> > </HTML>""" > else: > return """ > <HTML> > <HEAD> > </HEAD> > <BODY> > <FORM method="POST" action="/test/index/submit"> > <INPUT test="text" name="user" size="10"><BR> > <INPUT TYPE="password" name="password" size="10"><BR> > <INPUT TYPE="submit" name="submint" value="Login"> > </FORM> > </BODY> > </HTML>""" > def submit(self, req): > form_user = req.form['user'].value > form_password = req.form['password'].value > session = Session.Session(req) > if form_user == 'test' and form_password == 'test': > session['logged'] = True > session.save() > util.redirect(req, '/test') > > index = MyClass() So in other words you didn't actually need access to the request object in the constructor after all. What you were missing was that __call__() could be used in place of __str__() with __call__() allowing you to still access the request object and form parameters if need be. Graham
|