Robin Rosenberg <robin.rosenberg.lists@xxxxxxxxxx> wrote:
> fredagen den 5 september 2008 03.37.19 skrev Shawn O. Pearce:
> > Applications may wish to have this value, much as they
> > can get the raw seconds-since-epoch and reformat it
> > for their own uses.
>
> Necessary (performance) or just nice? We already provide the
> timezone in java.util format.
It isn't necessary from a performance perspective, but is required
to read the commit data exactly, but in parsed form.
What I'm doing at the application level is taking a RevCommit and
converting it into a different data structure (more specifically
a Google Protocol Buffer) that has all of the parts broken out.
It looks like this:
message GitPerson {
required string name = 1;
required string email = 2;
required int32 when = 3;
required int32 tz = 4;
}
message GitCommit {
required string id = 1;
repeated string parent_id = 2;
required GitPerson author = 3;
optional GitPerson committer = 4;
optional string message = 5;
}
I'm later processing the GitPerson in Python and need to use the tz
to offset when to format a time for display. If the author has a
"funny" offset in their commit our existing getTimeZone() method
can return null, which means I have to then lie and send 0 to the
Python code. I'd rather send the exact offset.
--
Shawn.
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
