public void SchedulerParentContextPreserved()
{
// This test was added due to a lot of paranoia around how the subcontext call stack is created.
// It's taking the existing call stack and just shoving on change_a. The interpreter then runs
// through everything as normal. What I expect to happen then is that after change_a finishes,
// it attempts to run the parent context. This is a no-no! This test will show if this happens
// based on the result of the variable 'b'. If it's greater than 1, then it got run more than
// once.
//
// We use an external entity to track this because its reference is independent of different
// contexts.
var intHaver = new GotAnInt();
runBasicTest(
"import sys\n" +
"\n" +
"a = 0\n" +
"def change_a():\n" +
" global a\n" +
" a = 1\n" +
"\n" +
"b.TheInt += 1\n" +
"sys.scheduler.schedule(change_a)\n",
new Dictionary <string, object>()
{
{ "b", intHaver }
}, new VariableMultimap(new TupleList <string, object>
{
{ "a", PyInteger.Create(1) },
}), 2);
Assert.That(intHaver.TheInt, Is.EqualTo(1));
}
public void SchedulerPassOneArg()
{
var intHaver = new GotAnInt();
runBasicTest(
"import sys\n" +
"\n" +
"def change_a(got_an_int):\n" +
" got_an_int.TheInt = 1\n" +
"\n" +
"sys.scheduler.schedule(change_a, container)\n",
new Dictionary <string, object>()
{
{ "container", intHaver }
}, new VariableMultimap(new TupleList <string, object>
{
}), 2);
Assert.That(intHaver.TheInt, Is.EqualTo(1));
}