You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Originally reported by: Eric Siegerman (BitBucket: eks, GitHub: eks)
Typically, the ExceptionInfo.value returned by pytest.raises() (when the exception is indeed raised) is an Exception instance. But under certain conditions, it isn't. It looks as though it contains the Exception's .args, rather than the Exception itself. This happens when all of the following are true:
Python is 2.6 (I've tested with 2.6.6 and 2.6.9)
The exception is one of Python's internal ones (I've tested with ZeroDivisionError, KeyError, and IOError; not sure whether this applies to others)
The exception was raised by the interpreter -- if I explicitly raise KeyError("msg"), for example, the test passes
The attached tests demonstrate the problem -- and the more useful behaviour when you use vanilla try/except instead of pytest.raises(). Under Python 2.6, the three pytest.raises() cases fail; under Python 2.7 and 3.4, they pass. (The try/except tests pass under all three Python versions.)
Originally reported by: Eric Siegerman (BitBucket: eks, GitHub: eks)
Typically, the ExceptionInfo.value returned by pytest.raises() (when the exception is indeed raised) is an Exception instance. But under certain conditions, it isn't. It looks as though it contains the Exception's .args, rather than the Exception itself. This happens when all of the following are true:
raise KeyError("msg")
, for example, the test passesThe attached tests demonstrate the problem -- and the more useful behaviour when you use vanilla try/except instead of pytest.raises(). Under Python 2.6, the three pytest.raises() cases fail; under Python 2.7 and 3.4, they pass. (The try/except tests pass under all three Python versions.)
The text was updated successfully, but these errors were encountered: