How to create my own HTTP 200 result class?

Question

In CherryPy, I can raise cherrypy.HTTPError(code, message) when something goes wrong, or I return string when it all works out.

What I am looking for now, is how can I create a class, so that I can do this:

class MyOwnResultClass:
   foo = None

   def __init__(self, foo):
      self.foo = foo

   def __str__(self):
      return "Result: {f}".format(f=self.foo)

class cherrypystuff(object):
    @cherrypy.expose
    def index(self):
        return MyOwnResultClass("f")

This fails with the error

TypeError: iteration over non-sequence

And I have to use return str(MyOwnResultClass("f").

Answer 1

I dived a little more into the error message, and apparently, CherryPy wants an iterable object as return type, if not a string.

So, the solution is quite simple: make the MyOwnResultClass iterable:

class MyOwnResultClass:
   foo = None

   def __init__(self, foo):
      self.foo = foo

   def __iter__(self):
      yield "Result: {f}".format(f=self.foo)

Update after seeing the talk suggested by @webKnjaZ, I refactored to this:

def MyOwnResult(foo):
    yield "Result: {f}".format(f=foo)