On Tue, 2005-02-01 at 17:15 -0500, Chris Lumens wrote:
> Attached is a patch for converting anaconda to using urlgrabber. It
> does not include the grabber.py or keepalive.py from urlgrabber. You'll
> have to go look at those on your own. It also is against the latest CVS
> (if that matters or not for these files).
Looks pretty good in general. Note that you'll want to s/import grabber/import
urlgrabber.grabber/. We can pull in urlgrabber out of the yum package
until I separate it out. Just add yum to PACKAGES in
scripts/upd-instroot and then add
usr/$LIBDIR/python?.?/site-packages/urlgrabber to the FILES list.
Some comments in-line
> Index: comps.py
> + try:
> + file = grabber.urlopen (filename)
> + except grabber.URLGrabError, e:
> + log ("URLGrabError: %s occurred getting %s", e.strerror,
> filename)
If there's an error getting the comps file here, we're going to hit a
traceback. There were infinite retries before. It's probably better to
raise something and have a reasonable error message pop up.
> Index: hdrlist.py
> + try:
> + extra_args = { "retry": 5 }
> + file = grabber.urlopen (filename, **extra_args)
Why not
file = grabber.urlopen(filename, retry = 5)
?
> + try:
> + url = grabber.urlopen (hdurl, **extra_args)
Same question here
With those changed, looks fine to go ahead and commit
Jeremy
