[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [rdiff-backup-users] More newby Qs: list of changed files, server-si
From: |
Bud P . Bruegger |
Subject: |
Re: [rdiff-backup-users] More newby Qs: list of changed files, server-side diffs |
Date: |
Mon, 16 Dec 2002 10:59:35 +0100 |
Ben, many thanks for the help. I think it will slip in the new year before I
can work on it...
best cheers
--bud
On Fri, 13 Dec 2002 17:16:58 -0800
Ben Escoto <address@hidden> wrote:
> >>>>> "BB" == Bud P Bruegger <address@hidden>
> >>>>> wrote the following on Tue, 3 Dec 2002 19:52:18 +0100
>
> BB> I tried to modify it such that it shows the type of change:
> BB> (new, modified, deleted) -- but there is still some problem that
> BB> I can't figure out.
>
> BB> My modified version of ListChangedSince in Main.py are attached
> BB> below.
>
> BB> The problem is that I can't recognize deleted files. I would
> BB> have expected get_rids_recursive(...).inc_list to include
> BB> increment files of type "snapshot". But they don't seem to be
> BB> listed. What am I doing wrong?
>
> Hi, sorry for the delay. Your patch looks good; I committed it to CVS
> with some modifications (new version below). However, it is
> impossible from just looking at the increments, to tell whether a file
> was deleted. Suppose a regular file is deleted, or suppose it gets
> replaced by a symlink. Or suppose the permissions on a device file
> get changed. All of these result in a snapshot, so we can't tell what
> happened by noting the fact that there is a snapshot. If you want
> this information, you could try using RestoreCombinedData objects
> instead of RestoreIncrementData.
>
> BB> Also, something that I plastered over with an if statement but
> BB> don't understand is that for some rids in
> BB> get_rids_recursive(...), rid.index seems empty.
>
> rid.index can be the empty tuple () when processing the root, but it
> should never be None.
>
> BB> PS. Beware, the version below has spaces instead of tabs for
> BB> indenting (the standard setting of my editor)...
>
> Maybe spaces are more standard. This all started when I couldn't
> figure out how to tell XEmacs's python mode to indent with 4 spaces.
> Maybe I should try to figure that out again and then switch. Tabs
> still seem more logical to me though.
>
>
> --
> Ben Escoto
>
... rest deleted..