mbox: remove remaining OFFSET usage in SQLite

We can use id_batch in the common case to speed up full mbox
retrievals.  Gigantic msets are still a problem, but will
be fixed in future commits.

diff --git a/lib/PublicInbox/Mbox.pm b/lib/PublicInbox/Mbox.pm
index 05de6be1a678f75ac258b24233c9f52d8c7a585f..0be1968584d9c68a7a76a72b98f82afea13bc564 100644 (file)
--- a/lib/PublicInbox/Mbox.pm
+++ b/lib/PublicInbox/Mbox.pm
@@ -138,8 +138,12 @@ sub thread_mbox {
        my ($ctx, $srch, $sfx) = @_;
        eval { require IO::Compress::Gzip };
        return sub { need_gzip(@_) } if $@;
-
-       my $cb = sub { $srch->get_thread($ctx->{mid}, @_) };
+       my $prev = 0;
+       my $cb = sub {
+               my $msgs = $srch->get_thread($ctx->{mid}, $prev);
+               $prev = $msgs->[-1]->{num} if scalar(@$msgs);
+               $msgs;
+       };
        PublicInbox::MboxGz->response($ctx, $cb);
 }
 
@@ -160,7 +164,25 @@ sub mbox_all {
 
        eval { require IO::Compress::Gzip };
        return sub { need_gzip(@_) } if $@;
-       my $cb = sub { $ctx->{srch}->query($query, @_) };
+       if ($query eq '') {
+               my $prev = 0;
+               my $msgs = [];
+               my $cb = sub {
+                       $ctx->{-inbox}->mm->id_batch($prev, sub {
+                               $msgs = $_[0];
+                       });
+                       $prev = $msgs->[-1] if @$msgs;
+                       $msgs;
+               };
+               return PublicInbox::MboxGz->response($ctx, $cb, 'all');
+       }
+       my $opts = { offset => 0 };
+       my $srch = $ctx->{srch};
+       my $cb = sub { # called by MboxGz->getline
+               my $msgs = $srch->query($query, $opts);
+               $opts->{offset} += scalar @$msgs;
+               $msgs;
+       };
        PublicInbox::MboxGz->response($ctx, $cb, 'results-'.$query);
 }
 
@@ -192,7 +214,6 @@ sub new {
                cb => $cb,
                ctx => $ctx,
                msgs => [],
-               opts => { offset => 0 },
        }, $class;
 }
 
@@ -223,6 +244,10 @@ sub getline {
        do {
                # work on existing result set
                while (defined(my $smsg = shift @$msgs)) {
+                       # id_batch may return integers
+                       ref($smsg) or
+                               $smsg = $ctx->{srch}->{over_ro}->get_art($smsg);
+
                        my $msg = eval { $ibx->msg_by_smsg($smsg) } or next;
                        $msg = Email::Simple->new($msg);
                        $gz->write(PublicInbox::Mbox::msg_str($ctx, $msg,
@@ -247,10 +272,10 @@ sub getline {
                }
 
                # refill result set
-               $msgs = $self->{msgs} = $self->{cb}->($self->{opts});
-               $self->{opts}->{offset} += scalar @$msgs;
+               $msgs = $self->{msgs} = $self->{cb}->();
        } while (@$msgs);
        $gz->close;
+       # signal that we're done and can return undef next call:
        delete $self->{ctx};
        ${delete $self->{buf}};
 }

diff --git a/lib/PublicInbox/Over.pm b/lib/PublicInbox/Over.pm
index b230d44a105791442f7262d7cb6b6be6686c45a0..0bd6008ba4adbfc706f742cdfbba08265616436b 100644 (file)
--- a/lib/PublicInbox/Over.pm
+++ b/lib/PublicInbox/Over.pm
@@ -50,9 +50,7 @@ sub do_get {
        my ($self, $sql, $opts, @args) = @_;
        my $dbh = $self->connect;
        my $lim = (($opts->{limit} || 0) + 0) || 1000;
-       my $off = (($opts->{offset} || 0) + 0) || 0;
        $sql .= "LIMIT $lim";
-       $sql .= " OFFSET $off" if $off > 0;
        my $msgs = $dbh->selectall_arrayref($sql, { Slice => {} }, @args);
        load_from_row($_) for @$msgs;
        $msgs
@@ -77,7 +75,7 @@ ORDER BY num ASC
 sub nothing () { wantarray ? (0, []) : [] };
 
 sub get_thread {
-       my ($self, $mid, $opts) = @_;
+       my ($self, $mid, $prev) = @_;
        my $dbh = $self->connect;
 
        my $id = $dbh->selectrow_array(<<'', undef, $mid);
@@ -96,13 +94,14 @@ SELECT tid,sid FROM over WHERE num = ? LIMIT 1
 
        defined $tid or return nothing; # $sid may be undef
 
-       my $cond = 'FROM over WHERE (tid = ? OR sid = ?) AND num > 0';
-       my $msgs = do_get($self, <<"", $opts, $tid, $sid);
-SELECT * $cond ORDER BY ts ASC
+       $prev ||= 0;
+       my $cond = 'FROM over WHERE (tid = ? OR sid = ?) AND num > ?';
+       my $msgs = do_get($self, <<"", {}, $tid, $sid, $prev);
+SELECT * $cond ORDER BY num ASC
 
        return $msgs unless wantarray;
 
-       my $nr = $dbh->selectrow_array(<<"", undef, $tid, $sid);
+       my $nr = $dbh->selectrow_array(<<"", undef, $tid, $sid, $prev);
 SELECT COUNT(num) $cond
 
        ($nr, $msgs);

diff --git a/lib/PublicInbox/Search.pm b/lib/PublicInbox/Search.pm
index f7fdf85478cfd387a728b935ecf5d212df35be3b..eca2b0f553da802821c25438b143f19e0bfab1bd 100644 (file)
--- a/lib/PublicInbox/Search.pm
+++ b/lib/PublicInbox/Search.pm
@@ -179,8 +179,8 @@ sub query {
 }
 
 sub get_thread {
-       my ($self, $mid, $opts) = @_;
-       $self->{over_ro}->get_thread($mid, $opts);
+       my ($self, $mid, $prev) = @_;
+       $self->{over_ro}->get_thread($mid, $prev);
 }
 
 sub retry_reopen {

diff --git a/t/psgi_v2.t b/t/psgi_v2.t
index 31c4178bd06aafb89243aba0f3d361d85f571ad6..aa3279cc6870743e43f62efb4b8f98141d4a7193 100644 (file)
--- a/t/psgi_v2.t
+++ b/t/psgi_v2.t
@@ -125,8 +125,28 @@ test_psgi(sub { $www->call(@_) }, sub {
                like($out, qr/^hello world$/m, 'got first in t.mbox.gz');
                like($out, qr/^hello world!$/m, 'got second in t.mbox.gz');
                like($out, qr/^hello ghosts$/m, 'got third in t.mbox.gz');
-               @from_ = ($raw =~ m/^From /mg);
+               @from_ = ($out =~ m/^From /mg);
                is(scalar(@from_), 3, 'three From_ lines in t.mbox.gz');
+
+               # search interface
+               $res = $cb->(POST('/v2test/?q=m:a-mid@b&x=m'));
+               $in = $res->content;
+               $status = IO::Uncompress::Gunzip::gunzip(\$in => \$out);
+               like($out, qr/^hello world$/m, 'got first in mbox POST');
+               like($out, qr/^hello world!$/m, 'got second in mbox POST');
+               like($out, qr/^hello ghosts$/m, 'got third in mbox POST');
+               @from_ = ($out =~ m/^From /mg);
+               is(scalar(@from_), 3, 'three From_ lines in mbox POST');
+
+               # all.mbox.gz interface
+               $res = $cb->(GET('/v2test/all.mbox.gz'));
+               $in = $res->content;
+               $status = IO::Uncompress::Gunzip::gunzip(\$in => \$out);
+               like($out, qr/^hello world$/m, 'got first in all.mbox');
+               like($out, qr/^hello world!$/m, 'got second in all.mbox');
+               like($out, qr/^hello ghosts$/m, 'got third in all.mbox');
+               @from_ = ($out =~ m/^From /mg);
+               is(scalar(@from_), 3, 'three From_ lines in all.mbox');
        };
 
        local $SIG{__WARN__} = 'DEFAULT';