Mailing List Archive: svn commit: r230837 - /spamassassin/trunk/lib/Mail/SpamAssassin/ArchiveIterator.pm

Author: jm
Date: Mon Aug 8 11:20:44 2005
New Revision: 230837

URL: http://svn.apache.org/viewcvs?rev=230837&view=rev
Log:
revert back to what the documented API says; caller code can easily set to 0 if they want to avoid the performance hit, and that's what should be done, instead of messing with public APIs. be clearer in the doco though.

Modified:
spamassassin/trunk/lib/Mail/SpamAssassin/ArchiveIterator.pm

Modified: spamassassin/trunk/lib/Mail/SpamAssassin/ArchiveIterator.pm
URL: http://svn.apache.org/viewcvs/spamassassin/trunk/lib/Mail/SpamAssassin/ArchiveIterator.pm?rev=230837&r1=230836&r2=230837&view=diff
==============================================================================
--- spamassassin/trunk/lib/Mail/SpamAssassin/ArchiveIterator.pm (original)
+++ spamassassin/trunk/lib/Mail/SpamAssassin/ArchiveIterator.pm Mon Aug 8 11:20:44 2005
@@ -154,7 +154,9 @@
=item opt_want_date

Set to 1 (default) if you want the received date to be filled in
-in the C<wanted_sub> callback below.
+in the C<wanted_sub> callback below. Set this to 0 to avoid this;
+it's a good idea to set this to 0 if you can, as it imposes a performance
+hit.

=item wanted_sub

@@ -191,7 +193,7 @@

$self->{opt_head} = 0 unless (defined $self->{opt_head});
$self->{opt_tail} = 0 unless (defined $self->{opt_tail});
- $self->{opt_want_date} = 0 unless (defined $self->{opt_want_date});
+ $self->{opt_want_date} = 1 unless (defined $self->{opt_want_date});

# If any of these options are set, we need to figure out the message's
# receive date at scan time. opt_n == 0, opt_after, opt_before

jm@apache.org writes:

> revert back to what the documented API says; caller code can easily
> set to 0 if they want to avoid the performance hit, and that's what
> should be done, instead of messing with public APIs. be clearer in
> the doco though.

The date extraction should *only* be done if it is requested.
Otherwise, it is completely pointless. Changing this is harmless and
correct.

Also, you do realize mass-check doesn't set this. Only the -n (opt_n in
ArchiveIterator) flag in mass-check was rendered useless by your changes
to the ArchiveIterator.

Essentially, you added a new flag, opt_want_date, that changed the
behavior of opt_n. Who is messing with public APIs, again? ;-)

Daniel

--
Daniel Quinlan
http://www.pathname.com/~quinlan/