Sorry for both the lateness of my reply and the fact that I missed
that you were just trying to run the tests. Though my sqlite
application runs under Windows, I've only touched SQLObject itself on
my Mac, which doesn't have the c: problem.
I was hoping that os.path.normpath might correct for this, but it
doesn't. I'd rather be explicit about fixing this, on the off chance
that someone actually has a ":" in their filename...
if filename[0] =3D=3D "/" and filename[2] =3D =3D ":":
filename =3D filename[1:]
Even better would be a fix for this in the URI parser, which I think
is where the leading slash came from to begin with.
It looks like you can get debugging output by adding -S or -O
parameters to your py.test commandline. (Take a look at
sqlobject/tests/conftest.py).
Kevin
On 7/14/05, Ben Scherrey <scherrey@...> wrote:
> Kevin,
>=20
> I'm not entirely certain as I'm just trying to get the py.test
> environment running for SQLObject so its whatever the tests are using.
> Since I'm running into difficulties getting certain types of debug
> output I'm not exactly sure what mechanism its using to open the
> connections. You can read my initial email in this thread to get more
> details about my setup.
>=20
> -- Ben Scherrey
>=20
> Kevin Dangoor wrote:
>=20
> >Until fairly recently, I was using sqlite on Windows and it worked
> >just fine. I even just confirmed that the paths I was using were full
> >paths with the drive letter.
> >
> >However, I was instantiating sqliteconnections myself (not using a
> >URI). Are you using a URI to open the connection?
> >
> >
> >On 7/14/05, Benjamin Scherrey <scherrey@...> wrote:
> >
> >
> >>Per the earlier msg - this is the latest svn source as of less than a
> >>week ago. Any clues as to the other issues? I take it that the win32
> >>environment is the bastard child of target platforms for SQLObject? :)
> >>
> >> -- Ben Scherrey
> >>
> >>
>=20
>
|