"def rescue_action(e) raise e end" causing problems in integ

Rails
1

Functional tests all come with a handy line that gives more sensible
exception reporting, looking something like :

class FooController; def rescue_action(e) raise e end; end

This causes problems in autotest. AFAICT, autotest doesn’t reload the
controllers when moving between functional tests and integration tests,
so controllers still have the rewritten rescue_action method in
integration.

Normally, this doesn’t matter. When an exception is raised, you get an
HTTP status of 500, as expected, and you don’t notice that
rescue_action is behaving differently.
However, the status 500 is being set by dispatcher.rb:42, and so there
are differences. In particular, ActiveRecord::RecordNotFound is
supposed to return a status of 404. In autotest’s integration
testing, you might get 404, you might get 500, depending on whether or
not your functional tests got loaded first on that particular run.

Has anyone found a workaround for this? At the moment I’m having to
remove the “def rescue_action(e) raise e end” line from functional
tests, and it’s making functional test debugging a bit painful…

Jon

2

Jon wrote the following on 10.01.2007 15:54 :

Has anyone found a workaround for this? At the moment I’m having to
remove the “def rescue_action(e) raise e end” line from functional
tests, and it’s making functional test debugging a bit painful…

I’ve no solution for you, but I’ve found another problem with
integration/functionnal tests that gets in the way with autotest. For
nearly all tests, I’m using a method for login a user in with this kind
of code :
old_controller = @controller
@controller = LoginController.new
post :login, <whatever_params_needed>
@controller = old_controller

There’s a hack in Integration tests which prevents this (a cache returns
the same controller on each Controller.new call) as soon as a
controller has been used in an integration test.

A solution :

Make IntegrationTest classes revert their changes to

ActionController::Base (which prevent functional tests from

working properly)

module IntegrationTestPatch
def self.included(base)
base.class_eval do
def run_with_cleanup(*args, &block)
ActionController::Base.send(:include,

ActionController::Integration::ControllerCapture)
run_without_cleanup(*args, &block)
if
ActionController::Base.respond_to?(:clear_last_instantiation!)
ActionController::Base.send(:include,
RevertControllerCapture)
end
end
alias_method :run_without_cleanup, :run
alias_method :run, :run_with_cleanup
end
end
end

Used to patch ActionController::Base

module RevertControllerCapture
def self.included(base)
base.class_eval do
class << self
alias_method :new, :new_without_capture
end
end
end
end

I realize I could manipulate the session directly without using my
LoginController to the same effect and this problem would go away too,
but I’ll have to duplicate the LoginController behaviour, which isn’t
DRY…

Lionel.