2009/9/11 Graham Dumpleton <[email protected]>:
> 2009/9/11 Greg <[email protected]>:
>>
>> I did see those - the thing is that the html2text script isn't
>> actually printing anything, it's just obtaining a reference to the
>> sys.stdout.write method. I didn't expect that to throw an error.
>
> Yes, I know that. This is one of the corner cases. I have kept
> updating the dummy log object to cater for these but for mod_wsgi 3.0
> just decided to default to not bothering to check so had stopped
> putting in code to deal with these corner cases. I guess though since
> they are doing that, I will have to anyway for 2.6 and for 3.0 where
> someone wants to enable it to check for portability.

Actually, getting confused here. The work arounds I was thinking of
related to something else. The solution has always been for this type
of issue to disable the validation by way of configuration or
reassignment of sys.stdout.

Graham

> Graham
>
>> Cheers,
>> Greg
>>
>> On Sep 9, 2:41 pm, Graham Dumpleton <[email protected]>
>> wrote:
>>> 2009/9/9 Greg <[email protected]>:
>>>
>>>
>>>
>>>
>>>
>>> > Hi Graham,
>>>
>>> > I'm getting an
>>>
>>> >    IOError: sys.stdout access restricted by mod_wsgi
>>>
>>> > error from an unexpected source. The offending code is a class
>>> > definition in html2text.py (http://www.aaronsw.com/2002/html2text/)
>>>
>>> >    class _html2text(sgmllib.SGMLParser):
>>> >        def __init__(self, out=sys.stdout.write, baseurl=''):
>>> >    ...
>>>
>>> > Not sure whether this should be considered an issue with mod_wsgi or
>>> > not, but the script isn't actually attempting to print anything - the
>>> > class is not even initialised in my code.
>>>
>>> Read:
>>>
>>>  http://code.google.com/p/modwsgi/wiki/ApplicationIssues#Writing_To_St...
>>>  http://blog.dscpl.com.au/2009/04/wsgi-and-printing-to-standard-output...
>>>
>>> The intent was to stop people from writing to sys.stdout for debugging
>>> as doing so makes WSGI applications non portable.
>>>
>>> I have given up pushing this barrow because the majority, rather than
>>> fix there code, seem happier to disable the complaints. There are also
>>> some corner cases as you have found which are hard to deal with.
>>>
>>> Anyway, in mod_wsgi 3.0, the restriction is off by default and would
>>> have to be enabled. Ie., opposite of now.
>>>
>>> Graham
>> >>
>>
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"modwsgi" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/modwsgi?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to