On 06/23/2014 07:06 PM, Andy Grover wrote:
The cache fix was unrelated.
<nod>, saw that now.
I agree, even though removing the set_attribute() resolved the issue,
the bug report raised questions about the exception handling. The syntax
<abc>
try:
<defgh>
except:
<abc cleanup>
raise
should be fine, and should not cause that error.
The only way this happens is if <defgh> somehow raises None. I imagine
<abc cleanup> could also manipulate the last exception traceback, but I
doubt this happens. I'll dig into it on my next cycle.
Regards,
--
Jerome