-# Copyright (C) 2018-2020 all contributors <meta@public-inbox.org>
+# Copyright (C) 2018-2021 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
# for XOVER, OVER in NNTP, and feeds/homepage/threads in PSGI
# are denoted by a negative NNTP article number.
package PublicInbox::OverIdx;
use strict;
-use warnings;
-use base qw(PublicInbox::Over);
+use v5.10.1;
+use parent qw(PublicInbox::Over);
use IO::Handle;
use DBI qw(:sql_types); # SQL_BLOB
use PublicInbox::MID qw/id_compress mids_for_index references/;
use PublicInbox::Smsg qw(subject_normalized);
use Compress::Zlib qw(compress);
-use PublicInbox::Search;
use Carp qw(croak);
sub dbh_new {
my ($self) = @_;
my $dbh = $self->SUPER::dbh_new($self->{-no_fsync} ? 2 : 1);
- # TRUNCATE reduces I/O compared to the default (DELETE)
- # We do not use WAL since we're optimized for read-only ops,
- # (and read-only requires SQLite 3.22.0 (2018-01-22)).
- $dbh->do('PRAGMA journal_mode = TRUNCATE');
-
# 80000 pages (80MiB on SQLite <3.12.0, 320MiB on 3.12.0+)
# was found to be good in 2018 during the large LKML import
# at the time. This ought to be configurable based on HW
}
}
+sub ibx_id {
+ my ($self, $eidx_key) = @_;
+ id_for($self, 'inboxes', 'ibx_id', eidx_key => $eidx_key);
+}
+
sub sid {
my ($self, $path) = @_;
return unless defined $path && $path ne '';
$$tid = $cur_tid;
} else { # rethreading, queue up dead ghosts
$$tid = next_tid($self);
- my $num = $smsg->{num};
- push(@{$self->{-ghosts_to_delete}}, $num) if $num < 0;
+ my $n = $smsg->{num};
+ if ($n > 0) {
+ $self->{dbh}->prepare_cached(<<'')->execute($$tid, $n);
+UPDATE over SET tid = ? WHERE num = ?
+
+ } elsif ($n < 0) {
+ push(@{$self->{-ghosts_to_delete}}, $n);
+ }
}
1;
}
if (my $del = delete $self->{-ghosts_to_delete}) {
delete_by_num($self, $_) for @$del;
}
- $tid // create_ghost($self, $mid);
-}
-
-sub create_ghost {
- my ($self, $mid) = @_;
- my $id = mid2id($self, $mid);
- my $num = next_ghost_num($self);
- $num < 0 or die "ghost num is non-negative: $num\n";
- my $tid = next_tid($self);
- my $dbh = $self->{dbh};
- $dbh->prepare_cached(<<'')->execute($num, $tid);
+ $tid // do { # create a new ghost
+ my $id = mid2id($self, $mid);
+ my $num = next_ghost_num($self);
+ $num < 0 or die "ghost num is non-negative: $num\n";
+ $tid = next_tid($self);
+ my $dbh = $self->{dbh};
+ $dbh->prepare_cached(<<'')->execute($num, $tid);
INSERT INTO over (num, tid) VALUES (?,?)
- $dbh->prepare_cached(<<'')->execute($id, $num);
+ $dbh->prepare_cached(<<'')->execute($id, $num);
INSERT INTO id2num (id, num) VALUES (?,?)
- $tid;
+ $tid;
+ };
}
sub merge_threads {
$tid;
}
-sub parse_references ($$$) {
- my ($smsg, $hdr, $mids) = @_;
- my $refs = references($hdr);
- push(@$refs, @$mids) if scalar(@$mids) > 1;
- return $refs if scalar(@$refs) == 0;
-
- # prevent circular references here:
- my %seen = ( $smsg->{mid} => 1 );
- my @keep;
- foreach my $ref (@$refs) {
- if (length($ref) > PublicInbox::MID::MAX_MID_SIZE) {
- warn "References: <$ref> too long, ignoring\n";
- next;
- }
- push(@keep, $ref) unless $seen{$ref}++;
- }
- $smsg->{references} = '<'.join('> <', @keep).'>' if @keep;
- \@keep;
-}
-
# normalize subjects so they are suitable as pathnames for URLs
# XXX: consider for removal
sub subject_path ($) {
lc($subj);
}
+sub ddd_for ($) {
+ my ($smsg) = @_;
+ my $dd = $smsg->to_doc_data;
+ utf8::encode($dd);
+ compress($dd);
+}
+
sub add_overview {
my ($self, $eml, $smsg) = @_;
$smsg->{lines} = $eml->body_raw =~ tr!\n!\n!;
my $mids = mids_for_index($eml);
- my $refs = parse_references($smsg, $eml, $mids);
+ my $refs = $smsg->parse_references($eml, $mids);
+ $mids->[0] //= do {
+ $smsg->{mid} //= '';
+ $eml->{-lei_fake_mid};
+ };
my $subj = $smsg->{subject};
my $xpath;
if ($subj ne '') {
$xpath = subject_path($subj);
$xpath = id_compress($xpath);
}
- my $dd = $smsg->to_doc_data;
- utf8::encode($dd);
- $dd = compress($dd);
- add_over($self, [ @$smsg{qw(ts ds num)}, $mids, $refs, $xpath, $dd ]);
+ add_over($self, $smsg, $mids, $refs, $xpath, ddd_for($smsg));
}
sub _add_over {
}
} elsif ($n < 0) { # ghost
$$old_tid //= $cur_valid ? $cur_tid : next_tid($self);
- link_refs($self, $refs, $$old_tid);
+ $$old_tid = link_refs($self, $refs, $$old_tid);
delete_by_num($self, $n);
$$v++;
}
}
sub add_over {
- my ($self, $values) = @_;
- my ($ts, $ds, $num, $mids, $refs, $xpath, $ddd) = @$values;
+ my ($self, $smsg, $mids, $refs, $xpath, $ddd) = @_;
my $old_tid;
my $vivified = 0;
+ my $num = $smsg->{num};
begin_lazy($self);
delete_by_num($self, $num, \$old_tid);
$v > 1 and warn "BUG: vivified multiple ($v) ghosts for $mid\n";
$vivified += $v;
}
- my $tid = $vivified ? $old_tid : link_refs($self, $refs, $old_tid);
- my $sid = sid($self, $xpath);
+ $smsg->{tid} = $vivified ? $old_tid : link_refs($self, $refs, $old_tid);
+ $smsg->{sid} = sid($self, $xpath);
my $dbh = $self->{dbh};
my $sth = $dbh->prepare_cached(<<'');
INSERT INTO over (num, tid, sid, ts, ds, ddd)
VALUES (?,?,?,?,?,?)
- my $n = 0;
- my @v = ($num, $tid, $sid, $ts, $ds);
- foreach (@v) { $sth->bind_param(++$n, $_) }
- $sth->bind_param(++$n, $ddd, SQL_BLOB);
+ my $nc = 1;
+ $sth->bind_param($nc, $num);
+ $sth->bind_param(++$nc, $smsg->{$_}) for (qw(tid sid ts ds));
+ $sth->bind_param(++$nc, $ddd, SQL_BLOB);
$sth->execute;
$sth = $dbh->prepare_cached(<<'');
INSERT INTO id2num (id, num) VALUES (?,?)
$dbh->do(<<'');
CREATE TABLE IF NOT EXISTS over (
- num INTEGER NOT NULL,
- tid INTEGER NOT NULL,
- sid INTEGER,
- ts INTEGER,
- ds INTEGER,
- ddd VARBINARY, /* doc-data-deflated */
- UNIQUE (num)
+ num INTEGER PRIMARY KEY NOT NULL, /* NNTP article number == IMAP UID */
+ tid INTEGER NOT NULL, /* THREADID (IMAP REFERENCES threading, JMAP) */
+ sid INTEGER, /* Subject ID (IMAP ORDEREDSUBJECT "threading") */
+ ts INTEGER, /* IMAP INTERNALDATE (Received: header, git commit time) */
+ ds INTEGER, /* RFC-2822 sent Date: header, git author time */
+ ddd VARBINARY /* doc-data-deflated (->to_doc_data, ->load_from_data) */
)
$dbh->do('CREATE INDEX IF NOT EXISTS idx_tid ON over (tid)');
$dbh->do(<<'');
CREATE TABLE IF NOT EXISTS subject (
sid INTEGER PRIMARY KEY AUTOINCREMENT,
- path VARCHAR(40) NOT NULL,
+ path VARCHAR(40) NOT NULL, /* SHA-1 of normalized subject */
UNIQUE (path)
)
$dbh->do(<<'');
CREATE TABLE IF NOT EXISTS id2num (
- id INTEGER NOT NULL,
+ id INTEGER NOT NULL, /* <=> msgid.id */
num INTEGER NOT NULL,
UNIQUE (id, num)
)
$dbh->do(<<'');
CREATE TABLE IF NOT EXISTS msgid (
- id INTEGER PRIMARY KEY AUTOINCREMENT,
+ id INTEGER PRIMARY KEY AUTOINCREMENT, /* <=> id2num.id */
mid VARCHAR(244) NOT NULL,
UNIQUE (mid)
)
sub begin_lazy {
my ($self) = @_;
return if $self->{txn};
- my $dbh = $self->connect or return;
+ my $dbh = $self->dbh or return;
$dbh->begin_work;
# $dbh->{Profile} = 2;
$self->{txn} = 1;
$self->{dbh}->rollback;
}
-sub disconnect {
+sub dbh_close {
my ($self) = @_;
die "in transaction" if $self->{txn};
- $self->SUPER::disconnect;
+ $self->SUPER::dbh_close;
}
sub create {
my ($self) = @_;
- unless (-r $self->{filename}) {
+ my $fn = $self->{filename} // do {
+ croak('BUG: no {filename}') unless $self->{dbh};
+ return;
+ };
+ unless (-r $fn) {
require File::Path;
require File::Basename;
- File::Path::mkpath(File::Basename::dirname($self->{filename}));
+ File::Path::mkpath(File::Basename::dirname($fn));
}
# create the DB:
- PublicInbox::Over::connect($self);
- $self->disconnect;
+ PublicInbox::Over::dbh($self);
+ $self->dbh_close;
}
sub rethread_prepare {
$pr->("I: rethread culled $total ghosts\n") if $pr && $total;
}
+# used for cross-inbox search
+sub eidx_prep ($) {
+ my ($self) = @_;
+ $self->{-eidx_prep} //= do {
+ my $dbh = $self->dbh;
+ $dbh->do(<<'');
+INSERT OR IGNORE INTO counter (key) VALUES ('eidx_docid')
+
+ $dbh->do(<<'');
+CREATE TABLE IF NOT EXISTS inboxes (
+ ibx_id INTEGER PRIMARY KEY AUTOINCREMENT,
+ eidx_key VARCHAR(255) NOT NULL, /* {newsgroup} // {inboxdir} */
+ UNIQUE (eidx_key)
+)
+
+ $dbh->do(<<'');
+CREATE TABLE IF NOT EXISTS xref3 (
+ docid INTEGER NOT NULL, /* <=> over.num */
+ ibx_id INTEGER NOT NULL, /* <=> inboxes.ibx_id */
+ xnum INTEGER NOT NULL, /* NNTP article number in ibx */
+ oidbin VARBINARY NOT NULL, /* 20-byte SHA-1 or 32-byte SHA-256 */
+ UNIQUE (docid, ibx_id, xnum, oidbin)
+)
+
+ $dbh->do('CREATE INDEX IF NOT EXISTS idx_docid ON xref3 (docid)');
+
+ # performance critical, this is not UNIQUE since we may need to
+ # tolerate some old bugs from indexing mirrors
+ $dbh->do('CREATE INDEX IF NOT EXISTS idx_nntp ON '.
+ 'xref3 (oidbin,xnum,ibx_id)');
+
+ $dbh->do(<<'');
+CREATE TABLE IF NOT EXISTS eidx_meta (
+ key VARCHAR(255) PRIMARY KEY,
+ val VARCHAR(255) NOT NULL
+)
+
+ # A queue of current docids which need reindexing.
+ # eidxq persists across aborted -extindex invocations
+ # Currently used for "-extindex --reindex" for Xapian
+ # data, but may be used in more places down the line.
+ $dbh->do(<<'');
+CREATE TABLE IF NOT EXISTS eidxq (docid INTEGER PRIMARY KEY NOT NULL)
+
+ 1;
+ };
+}
+
+sub eidx_meta { # requires transaction
+ my ($self, $key, $val) = @_;
+
+ my $sql = 'SELECT val FROM eidx_meta WHERE key = ? LIMIT 1';
+ my $dbh = $self->{dbh};
+ defined($val) or return $dbh->selectrow_array($sql, undef, $key);
+
+ my $prev = $dbh->selectrow_array($sql, undef, $key);
+ if (defined $prev) {
+ $sql = 'UPDATE eidx_meta SET val = ? WHERE key = ?';
+ $dbh->do($sql, undef, $val, $key);
+ } else {
+ $sql = 'INSERT INTO eidx_meta (key,val) VALUES (?,?)';
+ $dbh->do($sql, undef, $key, $val);
+ }
+ $prev;
+}
+
+sub eidx_max {
+ my ($self) = @_;
+ get_counter($self->{dbh}, 'eidx_docid');
+}
+
+sub add_xref3 {
+ my ($self, $docid, $xnum, $oidhex, $eidx_key) = @_;
+ begin_lazy($self);
+ my $ibx_id = ibx_id($self, $eidx_key);
+ my $oidbin = pack('H*', $oidhex);
+ my $sth = $self->{dbh}->prepare_cached(<<'');
+INSERT OR IGNORE INTO xref3 (docid, ibx_id, xnum, oidbin) VALUES (?, ?, ?, ?)
+
+ $sth->bind_param(1, $docid);
+ $sth->bind_param(2, $ibx_id);
+ $sth->bind_param(3, $xnum);
+ $sth->bind_param(4, $oidbin, SQL_BLOB);
+ $sth->execute;
+}
+
+# returns remaining reference count to $docid
+sub remove_xref3 {
+ my ($self, $docid, $oidhex, $eidx_key, $rm_eidx_info) = @_;
+ begin_lazy($self);
+ my $oidbin = pack('H*', $oidhex);
+ my ($sth, $ibx_id);
+ if (defined $eidx_key) {
+ $ibx_id = ibx_id($self, $eidx_key);
+ $sth = $self->{dbh}->prepare_cached(<<'');
+DELETE FROM xref3 WHERE docid = ? AND ibx_id = ? AND oidbin = ?
+
+ $sth->bind_param(1, $docid);
+ $sth->bind_param(2, $ibx_id);
+ $sth->bind_param(3, $oidbin, SQL_BLOB);
+ } else {
+ $sth = $self->{dbh}->prepare_cached(<<'');
+DELETE FROM xref3 WHERE docid = ? AND oidbin = ?
+
+ $sth->bind_param(1, $docid);
+ $sth->bind_param(2, $oidbin, SQL_BLOB);
+ }
+ $sth->execute;
+ $sth = $self->{dbh}->prepare_cached(<<'', undef, 1);
+SELECT COUNT(*) FROM xref3 WHERE docid = ?
+
+ $sth->execute($docid);
+ my $nr = $sth->fetchrow_array;
+ if ($nr == 0) {
+ delete_by_num($self, $docid);
+ } elsif (defined($ibx_id) && $rm_eidx_info) {
+ # if deduplication rules in ContentHash change, it's
+ # possible a docid can have multiple rows with the
+ # same ibx_id. This governs whether or not we call
+ # ->shard_remove_eidx_info in ExtSearchIdx.
+ $sth = $self->{dbh}->prepare_cached(<<'', undef, 1);
+SELECT COUNT(*) FROM xref3 WHERE docid = ? AND ibx_id = ?
+
+ $sth->execute($docid, $ibx_id);
+ my $count = $sth->fetchrow_array;
+ $$rm_eidx_info = ($count == 0);
+ }
+ $nr;
+}
+
+# for when an xref3 goes missing, this does NOT update {ts}
+sub update_blob {
+ my ($self, $smsg, $oidhex) = @_;
+ my $sth = $self->{dbh}->prepare(<<'');
+UPDATE over SET ddd = ? WHERE num = ?
+
+ $smsg->{blob} = $oidhex;
+ $sth->bind_param(1, ddd_for($smsg), SQL_BLOB);
+ $sth->bind_param(2, $smsg->{num});
+ $sth->execute;
+}
+
+sub eidxq_add {
+ my ($self, $docid) = @_;
+ $self->dbh->prepare_cached(<<'')->execute($docid);
+INSERT OR IGNORE INTO eidxq (docid) VALUES (?)
+
+}
+
+sub eidxq_del {
+ my ($self, $docid) = @_;
+ $self->dbh->prepare_cached(<<'')->execute($docid);
+DELETE FROM eidxq WHERE docid = ?
+
+}
+
1;