user/dev discussion of public-inbox itself
 help / color / mirror / code / Atom feed
From: Eric Wong <e@80x24.org>
To: meta@public-inbox.org
Subject: [PATCH 05/12] nntp: NEWNEWS: speed up filtering
Date: Fri, 27 Nov 2020 09:52:47 +0000	[thread overview]
Message-ID: <20201127095254.21624-6-e@80x24.org> (raw)
In-Reply-To: <20201127095254.21624-1-e@80x24.org>

With 50K newsgroups, the filtering phase goes from ~2000 seconds
to ~90 MILLISECONDS by relying on the grep perlop.  This moves
->over checking out of the main dispatch and amortizes the cost
via long_response.  (Fairly scheduled) long_response time in
newnews_i now takes ~360 seconds as opposed to ~30 seconds
before this change, however; but the initial filtering speedup
eliminating 2000s is more than worth it.
---
 lib/PublicInbox/NNTP.pm | 49 ++++++++++++++++++++---------------------
 1 file changed, 24 insertions(+), 25 deletions(-)

diff --git a/lib/PublicInbox/NNTP.pm b/lib/PublicInbox/NNTP.pm
index 68222196..5cbf5a16 100644
--- a/lib/PublicInbox/NNTP.pm
+++ b/lib/PublicInbox/NNTP.pm
@@ -294,23 +294,28 @@ sub ngpat2re (;$) {
 }
 
 sub newnews_i {
-	my ($self, $overs, $ts, $prev) = @_;
-	my $over = $overs->[0];
-	my $msgs = $over->query_ts($ts, $$prev);
-	if (scalar @$msgs) {
-		more($self, '<' .
-			join(">\r\n<", map { $_->{mid} } @$msgs ).
-			'>');
-		$$prev = $msgs->[-1]->{num};
-	} else {
-		shift @$overs;
-		if (@$overs) { # continue onto next newsgroup
-			$$prev = 0;
-			return 1;
-		} else { # break out of the long response.
-			return;
+	my ($self, $names, $ts, $prev) = @_;
+	my $ngname = $names->[0];
+	if (my $ibx = $self->{nntpd}->{groups}->{$ngname}) {
+		if (my $over = $ibx->over) {
+			my $msgs = $over->query_ts($ts, $$prev);
+			if (scalar @$msgs) {
+				more($self, '<' .
+					join(">\r\n<",
+						map { $_->{mid} } @$msgs ) .
+					'>');
+				$$prev = $msgs->[-1]->{num};
+				return 1; # continue on current group
+			}
 		}
 	}
+	shift @$names;
+	if (@$names) { # continue onto next newsgroup
+		$$prev = 0;
+		1;
+	} else { # all done, break out of the long_response
+		undef;
+	}
 }
 
 sub cmd_newnews ($$$$;$$) {
@@ -321,17 +326,11 @@ sub cmd_newnews ($$$$;$$) {
 	my ($keep, $skip) = split('!', $newsgroups, 2);
 	ngpat2re($keep);
 	ngpat2re($skip);
-	my @overs;
-	foreach my $ng (@{$self->{nntpd}->{grouplist}}) {
-		$ng->{newsgroup} =~ $keep or next;
-		$ng->{newsgroup} =~ $skip and next;
-		my $over = $ng->over or next;
-		push @overs, $over;
-	};
-	return '.' unless @overs;
-
+	my @names = grep(!/$skip/, grep(/$keep/,
+				@{$self->{nntpd}->{groupnames}}));
+	return '.' unless scalar(@names);
 	my $prev = 0;
-	long_response($self, \&newnews_i, \@overs, $ts, \$prev);
+	long_response($self, \&newnews_i, \@names, $ts, \$prev);
 }
 
 sub cmd_group ($$) {

  parent reply	other threads:[~2020-11-27  9:52 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-27  9:52 [PATCH 00/12] some NNTP-related fixes + speedups Eric Wong
2020-11-27  9:52 ` [PATCH 01/12] nntp: use Inbox->uidvalidity instead of ->mm->created_at Eric Wong
2020-11-27  9:52 ` [PATCH 02/12] nntpd: share {groups} hash with {-by_newsgroup} in Config Eric Wong
2020-11-27  9:52 ` [PATCH 03/12] mm: min/max: return 0 instead of undef Eric Wong
2020-11-27  9:52 ` [PATCH 04/12] nntp: use grep operation for wildmat matching Eric Wong
2020-11-27  9:52 ` Eric Wong [this message]
2020-11-27  9:52 ` [PATCH 06/12] miscsearch: implement ->newsgroup_matches Eric Wong
2020-11-27  9:52 ` [PATCH 07/12] nntp: LIST ACTIVE.TIMES use angle brackets around address Eric Wong
2020-11-27  9:52 ` [PATCH 08/12] nntp: move LIST iterators to long_response Eric Wong
2020-11-27  9:52 ` [PATCH 09/12] t/extsearch: show a more realistic case Eric Wong
2020-11-27  9:52 ` [PATCH 10/12] nntp: some minor golfing Eric Wong
2020-11-27  9:52 ` [PATCH 11/12] nntp: xref: simplify sub signature Eric Wong
2020-11-27  9:52 ` [PATCH 12/12] nntp: xref: use ->ALL extindex if available Eric Wong
2020-11-30 19:42   ` xref3 + NNTP problems Eric Wong
2020-11-30 23:37     ` [PATCH] nntp: make ->ALL Xref generation more fuzzy Eric Wong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://public-inbox.org/README

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20201127095254.21624-6-e@80x24.org \
    --to=e@80x24.org \
    --cc=meta@public-inbox.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/public-inbox.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).