On Jan 30, 2:20 am, Graham Dumpleton <[email protected]>
wrote:
> 2009/1/30 gert <[email protected]>:
> > On Jan 29, 1:58 am, Graham Dumpleton <[email protected]>
> > wrote:
> >> 2009/1/29 gert <[email protected]>:
> >> > On Jan 28, 3:00 am, gert <[email protected]> wrote:
> >> >> On Jan 28, 2:41 am, Graham Dumpleton <[email protected]>
> >> >> > This is for two reasons. The first is that a wrong configuration of
> >> >> > Apache will mean that people can download your source code, including
> >> >> > database logins and passwords if kept in those files. Second is that
> >> >> > if using .py extensions for WSGI script files and Python path being
> >> >> > set to be directory where WSGI script files are located, you can
> >> >> > double import WSGI script files under file name. The latter occurs
> >> >> > because mod_wsgi doesn't use standard Python module importer and so
> >> >> > using 'import' for name of WSGI script file will import it a second
> >> >> > time.
>
> >> > Ok so let say we have
> >> > /usr/httpd/htdocs/wsgi/double.py
> >> > WSGIPythonPath is set too /usr/httpd/htdocs/wsgi/
> >> > document root /usr/httpd/htdocs/
> >> > and apache is running embedded mode AddHandler wsgi-script .py
>
> >> > Does it do a double import of double.py when opening browser with
> >> >http://..../wsgi/double.py?
>
> >> We have had this discussion before some time back from memory. Last
> >> time I pointed you at:
>
> >> http://code.google.com/p/modwsgi/wiki/ReloadingSourceCode
>
> >> If you have two .py WSGI script files in that directory, eg. single.py
> >> and double.py, and both were used as WSGI script files, and in
> >> single.py it did 'import double', then you will end up with two copies
> >> of double.py in memory.
>
> >> That imported using 'import' from single.py will have __name__ set to
> >> 'double'. That loaded by mod_wsgi as WSGI script file will have
> >> __name__ set differently and will be stored separately in sys.modules.
>
> >> Thus, it isn't the act of sticking them in the directory, it is doing
> >> a separate 'import' of the WSGI script file from another WSGI script
> >> file or from some other Python module on sys.path.
>
> > Also if double.py does not contain def application(environ,
> > response): ?
>
> > /htdocs/wsgi/single.py
> > import double <------ MEMORY DUPLICATE IMPORT
> > def application(environ, response):
> > h = double.Hello()
>
> > /htdocs/wsgi/double.py
> > class Hello(object):
> > def __init__():
>
> If someone accesses a URL which maps to double.py and it doesn't have
> a 'application' entry point, it is still loaded, but would result in
> 404 response. That a duplicate occurred here would not be an issue
> unless the module was doing things at global scope which were
> cumulative. For example, if it added directories to sys.path, because
> it would be loaded twice, that directory would be added to sys.path
> twice.
ok what about
/htdocs/wsgi/single.py
import double <------ MEMORY DUPLICATE IMPORT
def application(environ, response):
h = double.Hello()
/htdocs/wsgi/double.py
import triple
class Hello(object):
def __init__():
where triple would be some big sucker like sqlite or something ?
Can you not make a WSGIPythonPath that mod_wsgi doesn't load
please :-)
--~--~---------~--~----~------------~-------~--~----~
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
-~----------~----~----~----~------~----~------~--~---
