Re: archive by months and years

1999-11-04 12:47:39
At 12:23 PM 11/4/99 -0500, Nathaniel Irons wrote:
On 11/4/99 at 9:50 AM, lnp3(_at_)columbia(_dot_)edu (Louis Proyect) wrote:

Furthermore, the script does not address itself to date anomalies
which are crucial for proper archiving. My mailing list has some
messages that, for example, are dated Oct 31, but are filtered into
the November archive--only because they arrived after midnight, Nov.1.

Why does this strike you as crucial?  

If I have a link to November archives and the first messages to appear are
dated "October 31," that's confusing.

As I just a moment ago mentioned to Andrzej, if you want to get clever
about interpreting date headers in your mailbox-assignment scripts,
you're inheriting the responsibility to reject those date headers which
are wildly wrong.  That's a much more complex script; I'm not sure how
it should work.

I am not interested in rejecting anything, only making sure that it gets
spooled to the right directory. I really had no problem doing this right,
since the test "if @msg[$cntr] =~ 'Date: ' will reliably identify the date
line. From there, I extract the month and year in order to either spool the
message to an existing directory or create a new one. My problem, however,
seems localized to getting the author/Otherindex resource file to work
correctly. My author index is screwed up, even when I run Mhonarc directly
from .procmail rather than a perl script as a wrapper. The only way I can
get it to work apparently is directly from the command line.

If I were bothered by this slightly-wrong-month phenomenon, I'd probably
write a script to examine consecutive mailboxes, which would make it far
easier to correctly identify and relocate those few messages which may
be just on the wrong side of the fence.  

I am correcting things manually right now, but would prefer to let a
computer program handle such mundane tasks.

Incidentally, if you're not using the MailTools modules for your mbox
manipulation scripts, by all means look into it.  It takes care of the
identification and manipulation of individual messages for you, loading
a whole mailbox file neatly into an array.  It can also replace your
code for extraction of dates, or any other headers you care to specify.



Thanks. I'll check it out.

<Prev in Thread] Current Thread [Next in Thread>