On Mon, Dec 08, 2008 at 10:09:16PM -0500, Cole Robinson wrote: > > for i in lst: > > - mode = os.stat(i)[stat.ST_MODE] > > - if stat.S_ISBLK(mode): > > - size.append(util.blkdev_size(i)) > > - typ.append(False) > > - elif stat.S_ISREG(mode): > > - size.append(os.path.getsize(i)) > > - typ.append(True) > > + (t, sz) = util.stat_disk(i) > > + typ.append(t) > > + size.append(sz) > > This, and all the other stat_disk code, seems like a separate cleanup. > Could we get this in a separate patch? It's not exactly an exciting change without the vdisk part, but OK. > > + elif (self.type == VirtualDisk.TYPE_FILE and self.path is not None > > + and os.path.exists(self.path) and util.is_vdisk(self.path)): > > + self._driverName = self.DRIVER_TAP > > + self._driverType = self.DRIVER_TAP_VDISK > > > > This could use a debug statement stating we are forcing the vdisk > driver. I'll add a debug statement, but we're not forcing it: if we detect a vdisk, then we know we will have to use vdisk to read it. > > +def vdisk_create(path, size, kind, sparse = True): > > +def vdisk_clone(path, clone): > > These two can be moved to VirtualDisk and CloneManager respectively > as non class functions. Please prefix them with an underscore to > mark them as private. OK. regards john _______________________________________________ et-mgmt-tools mailing list et-mgmt-tools@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/et-mgmt-tools
