-# 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>
# This interface wraps and mimics PublicInbox::Import
use v5.10.1;
use parent qw(PublicInbox::Lock);
use PublicInbox::SearchIdxShard;
-use PublicInbox::IdxStack;
+use PublicInbox::IPC;
use PublicInbox::Eml;
use PublicInbox::Git;
use PublicInbox::Import;
use PublicInbox::InboxWritable;
use PublicInbox::OverIdx;
use PublicInbox::Msgmap;
-use PublicInbox::Spawn qw(spawn popen_rd);
-use PublicInbox::SearchIdx;
+use PublicInbox::Spawn qw(spawn popen_rd run_die);
+use PublicInbox::Search;
+use PublicInbox::SearchIdx qw(log2stack is_ancestor check_size is_bad_blob);
use IO::Handle; # ->autoflush
-use File::Temp qw(tempfile);
+use File::Temp ();
+use POSIX ();
my $OID = qr/[a-f0-9]{40,}/;
# an estimate of the post-packed size to the raw uncompressed size
-my $PACKING_FACTOR = 0.4;
+our $PACKING_FACTOR = 0.4;
# SATA storage lags behind what CPUs are capable of, so relying on
# nproc(1) can be misleading and having extra Xapian shards is a
# to increase Xapian shards
our $NPROC_MAX_DEFAULT = 4;
-sub detect_nproc () {
- for my $nproc (qw(nproc gnproc)) { # GNU coreutils nproc
- `$nproc 2>/dev/null` =~ /^(\d+)$/ and return $1;
- }
-
- # getconf(1) is POSIX, but *NPROCESSORS* vars are not
- for (qw(_NPROCESSORS_ONLN NPROCESSORS_ONLN)) {
- `getconf $_ 2>/dev/null` =~ /^(\d+)$/ and return $1;
- }
-
- # should we bother with `sysctl hw.ncpu`? Those only give
- # us total processor count, not online processor count.
- undef
-}
-
sub nproc_shards ($) {
my ($creat_opt) = @_;
my $n = $creat_opt->{nproc} if ref($creat_opt) eq 'HASH';
$n //= $ENV{NPROC};
if (!$n) {
# assume 2 cores if not detectable or zero
- state $NPROC_DETECTED = detect_nproc() || 2;
+ state $NPROC_DETECTED = PublicInbox::IPC::detect_nproc() || 2;
$n = $NPROC_DETECTED;
$n = $NPROC_MAX_DEFAULT if $n > $NPROC_MAX_DEFAULT;
}
sub count_shards ($) {
my ($self) = @_;
- my $n = 0;
- my $xpfx = $self->{xpfx};
-
- # always load existing shards in case core count changes:
- # Also, shard count may change while -watch is running
- # due to "xcpdb --reshard"
- if (-d $xpfx) {
- my $XapianDatabase;
- foreach my $shard (<$xpfx/*>) {
- -d $shard && $shard =~ m!/[0-9]+\z! or next;
- $XapianDatabase //= do {
- require PublicInbox::Search;
- PublicInbox::Search::load_xapian();
- $PublicInbox::Search::X{Database};
- };
- eval {
- $XapianDatabase->new($shard)->close;
- $n++;
- };
- }
+ if (my $ibx = $self->{ibx}) {
+ # always load existing shards in case core count changes:
+ # Also, shard count may change while -watch is running
+ my $srch = $ibx->search or return 0;
+ delete $ibx->{search};
+ $srch->{nshard} // 0
+ } else { # ExtSearchIdx
+ $self->{nshard} ||= scalar($self->xdb_shards_flat);
}
- $n;
}
sub new {
die "$dir does not exist\n";
}
}
- $v2ibx->umask_prepare;
-
my $xpfx = "$dir/xap" . PublicInbox::Search::SCHEMA_VERSION;
my $self = {
- -inbox => $v2ibx,
+ ibx => $v2ibx,
im => undef, # PublicInbox::Import
parallel => 1,
transact_bytes => 0,
total_bytes => 0,
current_info => '',
xpfx => $xpfx,
- over => PublicInbox::OverIdx->new("$xpfx/over.sqlite3", 1),
+ oidx => PublicInbox::OverIdx->new("$xpfx/over.sqlite3"),
lock_path => "$dir/inbox.lock",
# limit each git repo (epoch) to 1GB or so
rotate_bytes => int((1024 * 1024 * 1024) / $PACKING_FACTOR),
- last_commit => [], # git repo -> commit
+ last_commit => [], # git epoch -> commit
};
+ $self->{oidx}->{-no_fsync} = 1 if $v2ibx->{-no_fsync};
$self->{shards} = count_shards($self) || nproc_shards($creat);
- $self->{index_max_size} = $v2ibx->{index_max_size};
bless $self, $class;
}
}
$self->idx_init;
$self->{mm}->skip_artnum($skip_artnum) if defined $skip_artnum;
- my $epoch_max = -1;
- git_dir_latest($self, \$epoch_max);
- if (defined $skip_epoch && $epoch_max == -1) {
- $epoch_max = $skip_epoch;
- }
- $self->git_init($epoch_max >= 0 ? $epoch_max : 0);
+ my $max = $self->{ibx}->max_git_epoch;
+ $max = $skip_epoch if (defined($skip_epoch) && !defined($max));
+ $self->git_init($max // 0);
$self->done;
}
# mimics Import::add and wraps it for v2
sub add {
my ($self, $eml, $check_cb) = @_;
- $self->{-inbox}->with_umask(\&_add, $self, $eml, $check_cb);
+ $self->{ibx}->with_umask(\&_add, $self, $eml, $check_cb);
+}
+
+sub idx_shard ($$) {
+ my ($self, $num) = @_;
+ $self->{idx_shards}->[$num % scalar(@{$self->{idx_shards}})];
}
# indexes a message, returns true if checkpointing is needed
-sub do_idx ($$$$) {
- my ($self, $msgref, $mime, $smsg) = @_;
- $smsg->{bytes} = $smsg->{raw_bytes} +
- PublicInbox::SearchIdx::crlf_adjust($$msgref);
- $self->{over}->add_overview($mime, $smsg);
- my $idx = idx_shard($self, $smsg->{num} % $self->{shards});
- $idx->index_raw($msgref, $mime, $smsg);
- my $n = $self->{transact_bytes} += $smsg->{raw_bytes};
- $n >= ($PublicInbox::SearchIdx::BATCH_BYTES * $self->{shards});
+sub do_idx ($$$) {
+ my ($self, $eml, $smsg) = @_;
+ $self->{oidx}->add_overview($eml, $smsg);
+ if ($self->{-need_xapian}) {
+ my $idx = idx_shard($self, $smsg->{num});
+ $idx->index_eml($eml, $smsg);
+ }
+ my $n = $self->{transact_bytes} += $smsg->{bytes};
+ $n >= $self->{batch_bytes};
}
sub _add {
# spam check:
if ($check_cb) {
- $mime = $check_cb->($mime, $self->{-inbox}) or return;
+ $mime = $check_cb->($mime, $self->{ibx}) or return;
}
# All pipes (> $^F) known to Perl 5.6+ have FD_CLOEXEC set,
$cmt = $im->get_mark($cmt);
$self->{last_commit}->[$self->{epoch_max}] = $cmt;
- my $msgref = delete $smsg->{-raw_email};
- if (do_idx($self, $msgref, $mime, $smsg)) {
+ if (do_idx($self, $mime, $smsg)) {
$self->checkpoint;
}
sub v2_num_for {
my ($self, $mime) = @_;
- my $mids = mids($mime->header_obj);
+ my $mids = mids($mime);
if (@$mids) {
my $mid = $mids->[0];
my $num = $self->{mm}->mid_insert($mid);
# AltId may pre-populate article numbers (e.g. X-Mail-Count
# or NNTP article number), use that article number if it's
# not in Over.
- my $altid = $self->{-inbox}->{altid};
+ my $altid = $self->{ibx}->{altid};
if ($altid && grep(/:file=msgmap\.sqlite3\z/, @$altid)) {
my $num = $self->{mm}->num_for($mid);
- if (defined $num && !$self->{over}->get_art($num)) {
+ if (defined $num && !$self->{oidx}->get_art($num)) {
return ($num, $mid);
}
}
}
sub v2_num_for_harder {
- my ($self, $mime) = @_;
+ my ($self, $eml) = @_;
- my $hdr = $mime->header_obj;
- my $dig = content_digest($mime);
- my $mid0 = PublicInbox::Import::digest2mid($dig, $hdr);
+ my $dig = content_digest($eml);
+ my $mid0 = PublicInbox::Import::digest2mid($dig, $eml);
my $num = $self->{mm}->mid_insert($mid0);
unless (defined $num) {
# it's hard to spoof the last Received: header
- my @recvd = $hdr->header_raw('Received');
+ my @recvd = $eml->header_raw('Received');
$dig->add("Received: $_") foreach (@recvd);
- $mid0 = PublicInbox::Import::digest2mid($dig, $hdr);
+ $mid0 = PublicInbox::Import::digest2mid($dig, $eml);
$num = $self->{mm}->mid_insert($mid0);
# fall back to a random Message-ID and give up determinism:
until (defined($num)) {
$dig->add(rand);
- $mid0 = PublicInbox::Import::digest2mid($dig, $hdr);
+ $mid0 = PublicInbox::Import::digest2mid($dig, $eml);
warn "using random Message-ID <$mid0> as fallback\n";
$num = $self->{mm}->mid_insert($mid0);
}
}
- PublicInbox::Import::append_mid($hdr, $mid0);
+ PublicInbox::Import::append_mid($eml, $mid0);
($num, $mid0);
}
-sub idx_shard {
- my ($self, $shard_i) = @_;
- $self->{idx_shards}->[$shard_i];
-}
-
sub _idx_init { # with_umask callback
my ($self, $opt) = @_;
$self->lock_acquire unless $opt && $opt->{-skip_lock};
- $self->{over}->create;
+ $self->{oidx}->create;
# xcpdb can change shard count while -watch is idle
my $nshards = count_shards($self);
$self->{shards} = $nshards if $nshards && $nshards != $self->{shards};
+ $self->{batch_bytes} = $opt->{batch_size} //
+ $PublicInbox::SearchIdx::BATCH_BYTES;
# need to create all shards before initializing msgmap FD
# idx_shards must be visible to all forked processes
my $max = $self->{shards} - 1;
my $idx = $self->{idx_shards} = [];
push @$idx, PublicInbox::SearchIdxShard->new($self, $_) for (0..$max);
+ $self->{-need_xapian} = $idx->[0]->need_xapian;
+
+ # SearchIdxShard may do their own flushing, so don't scale
+ # until after forking
+ $self->{batch_bytes} *= $self->{shards} if $self->{parallel};
+
+ my $ibx = $self->{ibx} or return; # ExtIdxSearch
# Now that all subprocesses are up, we can open the FDs
# for SQLite:
my $mm = $self->{mm} = PublicInbox::Msgmap->new_file(
- "$self->{-inbox}->{inboxdir}/msgmap.sqlite3", 1);
+ "$ibx->{inboxdir}/msgmap.sqlite3",
+ $ibx->{-no_fsync} ? 2 : 1);
$mm->{dbh}->begin_work;
}
+sub parallel_init ($$) {
+ my ($self, $indexlevel) = @_;
+ $self->{parallel} = 0 if ($indexlevel // 'full') eq 'basic';
+}
+
# idempotent
sub idx_init {
my ($self, $opt) = @_;
return if $self->{idx_shards};
- my $ibx = $self->{-inbox};
+ my $ibx = $self->{ibx};
# do not leak read-only FDs to child processes, we only have these
# FDs for duplicate detection so they should not be
# frequently activated.
- # delete @$ibx{qw(git mm search)};
- delete $ibx->{$_} foreach (qw(git mm search));
-
- $self->{parallel} = 0 if ($ibx->{indexlevel}//'') eq 'basic';
- if ($self->{parallel}) {
- pipe(my ($r, $w)) or die "pipe failed: $!";
- # pipe for barrier notifications doesn't need to be big,
- # 1031: F_SETPIPE_SZ
- fcntl($w, 1031, 4096) if $^O eq 'linux';
- $self->{bnote} = [ $r, $w ];
- $w->autoflush(1);
- }
+ delete @$ibx{qw(mm search)};
+ $ibx->git->cleanup;
- $ibx->umask_prepare;
+ parallel_init($self, $ibx->{indexlevel});
$ibx->with_umask(\&_idx_init, $self, $opt);
}
sub _replace_oids ($$$) {
my ($self, $mime, $replace_map) = @_;
$self->done;
- my $pfx = "$self->{-inbox}->{inboxdir}/git";
+ my $ibx = $self->{ibx};
+ my $pfx = "$ibx->{inboxdir}/git";
my $rewrites = []; # epoch => commit
- my $max = $self->{epoch_max};
-
- unless (defined($max)) {
- defined(my $latest = git_dir_latest($self, \$max)) or return;
- $self->{epoch_max} = $max;
- }
+ my $max = $self->{epoch_max} //= $ibx->max_git_epoch // return;
foreach my $i (0..$max) {
my $git_dir = "$pfx/$i.git";
} else {
$im = $self->importer;
}
- my $over = $self->{over};
+ my $oidx = $self->{oidx};
my $chashes = content_hashes($old_eml);
my $removed = [];
- my $mids = mids($old_eml->header_obj);
+ my $mids = mids($old_eml);
# We avoid introducing new blobs into git since the raw content
# can be slightly different, so we do not need the user-supplied
foreach my $mid (@$mids) {
my %gone; # num => [ smsg, $mime, raw ]
my ($id, $prev);
- while (my $smsg = $over->next_by_mid($mid, \$id, \$prev)) {
+ while (my $smsg = $oidx->next_by_mid($mid, \$id, \$prev)) {
my $msg = get_blob($self, $smsg);
if (!defined($msg)) {
warn "broken smsg for $mid\n";
} else { # ->purge or ->remove
$self->{mm}->num_delete($num);
}
- unindex_oid_remote($self, $oid, $mid);
+ unindex_oid_aux($self, $oid, $mid);
}
}
# (retval[2]) is not part of the stable API shared with Import->remove
sub remove {
my ($self, $eml, $cmt_msg) = @_;
- my $r = $self->{-inbox}->with_umask(\&rewrite_internal,
+ my $r = $self->{ibx}->with_umask(\&rewrite_internal,
$self, $eml, $cmt_msg);
defined($r) && defined($r->[0]) ? @$r: undef;
}
sub _replace ($$;$$) {
my ($self, $old_eml, $new_eml, $sref) = @_;
my $arg = [ $self, $old_eml, undef, $new_eml, $sref ];
- my $rewritten = $self->{-inbox}->with_umask(\&rewrite_internal,
+ my $rewritten = $self->{ibx}->with_umask(\&rewrite_internal,
$self, $old_eml, undef, $new_eml, $sref) or return;
my $rewrites = $rewritten->{rewrites};
my ($self, $raw) = @_;
# grab the expected OID we have to reindex:
pipe(my($in, $w)) or die "pipe: $!";
- my $git_dir = $self->{-inbox}->git->{git_dir};
+ my $git_dir = $self->git->{git_dir};
my $cmd = ['git', "--git-dir=$git_dir", qw(hash-object --stdin)];
my $r = popen_rd($cmd, undef, { 0 => $in });
print $w $$raw or die "print \$w: $!";
# Message-IDs are pretty complex and rethreading hasn't been fully
# implemented, yet.
sub check_mids_match ($$) {
- my ($old_mime, $new_mime) = @_;
- my $old = $old_mime->header_obj;
- my $new = $new_mime->header_obj;
+ my ($old, $new) = @_;
_check_mids_match(mids($old), mids($new), 'Message-ID(s)');
_check_mids_match(references($old), references($new),
'References/In-Reply-To');
}
# make sure we really got the OID:
- my ($blob, $type, $bytes) = $self->{-inbox}->git->check($expect_oid);
+ my ($blob, $type, $bytes) = $self->git->check($expect_oid);
$blob eq $expect_oid or die "BUG: $expect_oid not found after replace";
# don't leak FDs to Xapian:
- $self->{-inbox}->git->cleanup;
+ $self->git->cleanup;
# reindex modified messages:
for my $smsg (@$need_reindex) {
my $new_smsg = bless {
blob => $blob,
- raw_bytes => $bytes,
num => $smsg->{num},
mid => $smsg->{mid},
}, 'PublicInbox::Smsg';
my $sync = { autime => $smsg->{ds}, cotime => $smsg->{ts} };
$new_smsg->populate($new_mime, $sync);
- do_idx($self, \$raw, $new_mime, $new_smsg);
+ $new_smsg->set_bytes($raw, $bytes);
+ do_idx($self, $new_mime, $new_smsg);
}
$rewritten->{rewrites};
}
$self->{mm}->last_commit_xap($v, $i, $cmt);
}
-sub set_last_commits ($) {
+sub set_last_commits ($) { # this is NOT for ExtSearchIdx
my ($self) = @_;
defined(my $epoch_max = $self->{epoch_max}) or return;
my $last_commit = $self->{last_commit};
}
}
-sub barrier_init {
- my ($self, $n) = @_;
- $self->{bnote} or return;
- --$n;
- my $barrier = { map { $_ => 1 } (0..$n) };
-}
-
-sub barrier_wait {
- my ($self, $barrier) = @_;
- my $bnote = $self->{bnote} or return;
- my $r = $bnote->[0];
- while (scalar keys %$barrier) {
- defined(my $l = readline($r)) or die "EOF on barrier_wait: $!";
- $l =~ /\Abarrier (\d+)/ or die "bad line on barrier_wait: $l";
- delete $barrier->{$1} or die "bad shard[$1] on barrier wait";
- }
-}
-
# public
sub checkpoint ($;$) {
my ($self, $wait) = @_;
}
my $shards = $self->{idx_shards};
if ($shards) {
- my $dbh = $self->{mm}->{dbh};
+ my $mm = $self->{mm};
+ my $dbh = $mm->{dbh} if $mm;
# SQLite msgmap data is second in importance
- $dbh->commit;
+ $dbh->commit if $dbh;
# SQLite overview is third
- $self->{over}->commit_lazy;
+ $self->{oidx}->commit_lazy;
# Now deal with Xapian
- if ($wait) {
- my $barrier = $self->barrier_init(scalar @$shards);
- # each shard needs to issue a barrier command
- $_->remote_barrier for @$shards;
+ # start commit_txn_lazy asynchronously on all parallel shards
+ # (non-parallel waits here)
+ $_->ipc_do('commit_txn_lazy') for @$shards;
+
+ # transactions started on parallel shards,
+ # wait for them by issuing an echo command (echo can only
+ # run after commit_txn_lazy is done)
+ if ($wait && $self->{parallel}) {
+ my $i = 0;
+ for my $shard (@$shards) {
+ my $echo = $shard->ipc_do('echo', $i);
+ $echo == $i or die <<"";
+shard[$i] bad echo:$echo != $i waiting for txn commit
+
+ ++$i;
+ }
+ }
- # wait for each Xapian shard
- $self->barrier_wait($barrier);
- } else {
- $_->remote_commit for @$shards;
+ my $midx = $self->{midx}; # misc index
+ if ($midx) {
+ $midx->commit_txn;
+ $PublicInbox::Search::X{CLOEXEC_UNSET} and
+ $self->git->cleanup;
}
# last_commit is special, don't commit these until
- # remote shards are done:
- $dbh->begin_work;
+ # Xapian shards are done:
+ $dbh->begin_work if $dbh;
set_last_commits($self);
- $dbh->commit;
-
- $dbh->begin_work;
+ if ($dbh) {
+ $dbh->commit;
+ $dbh->begin_work;
+ }
}
$self->{total_bytes} += $self->{transact_bytes};
$self->{transact_bytes} = 0;
# public
sub barrier { checkpoint($_[0], 1) };
+# true if locked and active
+sub active { !!$_[0]->{im} }
+
# public
sub done {
my ($self) = @_;
- my $im = delete $self->{im};
- $im->done if $im; # PublicInbox::Import::done
- checkpoint($self);
- my $mm = delete $self->{mm};
- $mm->{dbh}->commit if $mm;
+ my $err = '';
+ if (my $im = delete $self->{im}) {
+ eval { $im->done }; # PublicInbox::Import::done
+ $err .= "import done: $@\n" if $@;
+ }
+ if (!$err) {
+ eval { checkpoint($self) };
+ $err .= "checkpoint: $@\n" if $@;
+ }
+ if (my $mm = delete $self->{mm}) {
+ my $m = $err ? 'rollback' : 'commit';
+ eval { $mm->{dbh}->$m };
+ $err .= "msgmap $m: $@\n" if $@;
+ }
my $shards = delete $self->{idx_shards};
if ($shards) {
- $_->remote_close for @$shards;
+ for (@$shards) {
+ eval { $_->shard_close };
+ $err .= "shard close: $@\n" if $@;
+ }
}
- $self->{over}->disconnect;
- delete $self->{bnote};
+ eval { $self->{oidx}->dbh_close };
+ $err .= "over close: $@\n" if $@;
+ delete $self->{midx};
my $nbytes = $self->{total_bytes};
$self->{total_bytes} = 0;
$self->lock_release(!!$nbytes) if $shards;
- $self->{-inbox}->git->cleanup;
+ $self->git->cleanup;
+ die $err if $err;
+}
+
+sub write_alternates ($$$) {
+ my ($info_dir, $mode, $out) = @_;
+ my $fh = File::Temp->new(TEMPLATE => 'alt-XXXX', DIR => $info_dir);
+ my $tmp = $fh->filename;
+ print $fh @$out or die "print $tmp: $!\n";
+ chmod($mode, $fh) or die "fchmod $tmp: $!\n";
+ close $fh or die "close $tmp $!\n";
+ my $alt = "$info_dir/alternates";
+ rename($tmp, $alt) or die "rename $tmp => $alt: $!\n";
+ $fh->unlink_on_destroy(0);
}
sub fill_alternates ($$) {
my ($self, $epoch) = @_;
- my $pfx = "$self->{-inbox}->{inboxdir}/git";
- my $all = "$self->{-inbox}->{inboxdir}/all.git";
+ my $pfx = "$self->{ibx}->{inboxdir}/git";
+ my $all = "$self->{ibx}->{inboxdir}/all.git";
PublicInbox::Import::init_bare($all) unless -d $all;
my $info_dir = "$all/objects/info";
my $alt = "$info_dir/alternates";
}
}
return unless $new;
-
- my ($fh, $tmp) = tempfile('alt-XXXXXXXX', DIR => $info_dir);
- print $fh join("\n", sort { $alt{$b} <=> $alt{$a} } keys %alt), "\n"
- or die "print $tmp: $!\n";
- chmod($mode, $fh) or die "fchmod $tmp: $!\n";
- close $fh or die "close $tmp $!\n";
- rename($tmp, $alt) or die "rename $tmp => $alt: $!\n";
+ write_alternates($info_dir, $mode,
+ [join("\n", sort { $alt{$b} <=> $alt{$a} } keys %alt), "\n"]);
}
sub git_init {
my ($self, $epoch) = @_;
- my $git_dir = "$self->{-inbox}->{inboxdir}/git/$epoch.git";
+ my $git_dir = "$self->{ibx}->{inboxdir}/git/$epoch.git";
PublicInbox::Import::init_bare($git_dir);
- my @cmd = (qw/git config/, "--file=$git_dir/config",
- 'include.path', '../../all.git/config');
- PublicInbox::Import::run_die(\@cmd);
+ run_die([qw(git config), "--file=$git_dir/config",
+ qw(include.path ../../all.git/config)]);
fill_alternates($self, $epoch);
$git_dir
}
-sub git_dir_latest {
- my ($self, $max) = @_;
- $$max = -1;
- my $pfx = "$self->{-inbox}->{inboxdir}/git";
- return unless -d $pfx;
- my $latest;
- opendir my $dh, $pfx or die "opendir $pfx: $!\n";
- while (defined(my $git_dir = readdir($dh))) {
- $git_dir =~ m!\A([0-9]+)\.git\z! or next;
- if ($1 > $$max) {
- $$max = $1;
- $latest = "$pfx/$git_dir";
- }
- }
- $latest;
-}
-
sub importer {
my ($self) = @_;
my $im = $self->{im};
}
my $epoch = 0;
my $max;
- my $latest = git_dir_latest($self, \$max);
+ my $latest = $self->{ibx}->git_dir_latest(\$max);
if (defined $latest) {
my $git = PublicInbox::Git->new($latest);
my $packed_bytes = $git->packed_bytes;
sub import_init {
my ($self, $git, $packed_bytes, $tmp) = @_;
- my $im = PublicInbox::Import->new($git, undef, undef, $self->{-inbox});
+ my $im = PublicInbox::Import->new($git, undef, undef, $self->{ibx});
$im->{bytes_added} = int($packed_bytes / $PACKING_FACTOR);
$im->{lock_path} = undef;
$im->{path_type} = 'v2';
sub diff ($$$) {
my ($mid, $cur, $new) = @_;
- my ($ah, $an) = tempfile('email-cur-XXXXXXXX', TMPDIR => 1);
+ my $ah = File::Temp->new(TEMPLATE => 'email-cur-XXXX', TMPDIR => 1);
print $ah $cur->as_string or die "print: $!";
- close $ah or die "close: $!";
- my ($bh, $bn) = tempfile('email-new-XXXXXXXX', TMPDIR => 1);
+ $ah->flush or die "flush: $!";
PublicInbox::Import::drop_unwanted_headers($new);
+ my $bh = File::Temp->new(TEMPLATE => 'email-new-XXXX', TMPDIR => 1);
print $bh $new->as_string or die "print: $!";
- close $bh or die "close: $!";
- my $cmd = [ qw(diff -u), $an, $bn ];
+ $bh->flush or die "flush: $!";
+ my $cmd = [ qw(diff -u), $ah->filename, $bh->filename ];
print STDERR "# MID conflict <$mid>\n";
my $pid = spawn($cmd, undef, { 1 => 2 });
waitpid($pid, 0) == $pid or die "diff did not finish";
- unlink($an, $bn);
}
sub get_blob ($$) {
return $msg if $msg;
}
# older message, should be in alternates
- my $ibx = $self->{-inbox};
- $ibx->msg_by_smsg($smsg);
+ $self->{ibx}->msg_by_smsg($smsg);
}
sub content_exists ($$$) {
my ($self, $mime, $mid) = @_;
- my $over = $self->{over};
+ my $oidx = $self->{oidx};
my $chashes = content_hashes($mime);
my ($id, $prev);
- while (my $smsg = $over->next_by_mid($mid, \$id, \$prev)) {
+ while (my $smsg = $oidx->next_by_mid($mid, \$id, \$prev)) {
my $msg = get_blob($self, $smsg);
if (!defined($msg)) {
warn "broken smsg for $mid\n";
sub atfork_child {
my ($self) = @_;
- my $fh = delete $self->{reindex_pipe};
- close $fh if $fh;
- if (my $shards = $self->{idx_shards}) {
- $_->atfork_child foreach @$shards;
+ if (my $older_siblings = $self->{idx_shards}) {
+ $_->ipc_sibling_atfork_child for @$older_siblings;
}
if (my $im = $self->{im}) {
$im->atfork_child;
}
- die "unexpected mm" if $self->{mm};
- close $self->{bnote}->[0] or die "close bnote[0]: $!\n";
- $self->{bnote}->[1];
+ die "BUG: unexpected mm" if $self->{mm};
}
-sub reindex_checkpoint ($$$) {
- my ($self, $sync, $git) = @_;
-
- $git->cleanup;
- $sync->{mm_tmp}->atfork_prepare;
- $self->done; # release lock
+sub reindex_checkpoint ($$) {
+ my ($self, $sync) = @_;
+
+ $self->git->async_wait_all;
+ $self->update_last_commit($sync);
+ ${$sync->{need_checkpoint}} = 0;
+ my $mm_tmp = $sync->{mm_tmp};
+ $mm_tmp->atfork_prepare if $mm_tmp;
+ die 'BUG: {im} during reindex' if $self->{im};
+ if ($self->{ibx_map} && !$sync->{checkpoint_unlocks}) {
+ checkpoint($self, 1); # no need to release lock on pure index
+ } else {
+ $self->done; # release lock
+ }
- if (my $pr = $sync->{-opt}->{-progress}) {
- my ($bn) = (split('/', $git->{git_dir}))[-1];
- $pr->("$bn ".sprintf($sync->{-regen_fmt}, $sync->{nr}));
+ if (my $pr = $sync->{-regen_fmt} ? $sync->{-opt}->{-progress} : undef) {
+ $pr->(sprintf($sync->{-regen_fmt}, ${$sync->{nr}}));
}
# allow -watch or -mda to write...
- $self->idx_init; # reacquire lock
- $sync->{mm_tmp}->atfork_parent;
+ $self->idx_init($sync->{-opt}); # reacquire lock
+ if (my $intvl = $sync->{check_intvl}) { # eidx
+ $sync->{next_check} = PublicInbox::DS::now() + $intvl;
+ }
+ $mm_tmp->atfork_parent if $mm_tmp;
}
-sub reindex_oid ($$$$) {
- my ($self, $sync, $git, $oid) = @_;
- return if PublicInbox::SearchIdx::too_big($self, $git, $oid);
- my ($num, $mid0, $len);
- my $msgref = $git->cat_file($oid, \$len);
- return if $len == 0; # purged
- my $mime = PublicInbox::Eml->new($$msgref);
- my $mids = mids($mime->header_obj);
- my $chash = content_hash($mime);
+sub index_finalize ($$) {
+ my ($arg, $index) = @_;
+ ++$arg->{self}->{nidx};
+ if (defined(my $cur = $arg->{cur_cmt})) {
+ ${$arg->{latest_cmt}} = $cur;
+ } elsif ($index) {
+ die 'BUG: {cur_cmt} missing';
+ } # else { unindexing @leftovers doesn't set {cur_cmt}
+}
+
+sub index_oid { # cat_async callback
+ my ($bref, $oid, $type, $size, $arg) = @_;
+ is_bad_blob($oid, $type, $size, $arg->{oid}) and
+ return index_finalize($arg, 1); # size == 0 purged returns here
+ my $self = $arg->{self};
+ local $self->{current_info} = "$self->{current_info} $oid";
+ my ($num, $mid0);
+ my $eml = PublicInbox::Eml->new($$bref);
+ my $mids = mids($eml);
+ my $chash = content_hash($eml);
if (scalar(@$mids) == 0) {
warn "E: $oid has no Message-ID, skipping\n";
}
# {unindexed} is unlikely
- if ((my $unindexed = $self->{unindexed}) && scalar(@$mids) == 1) {
- $num = delete($unindexed->{$mids->[0]});
+ if (my $unindexed = $arg->{unindexed}) {
+ my $oidbin = pack('H*', $oid);
+ my $u = $unindexed->{$oidbin};
+ ($num, $mid0) = splice(@$u, 0, 2) if $u;
if (defined $num) {
- $mid0 = $mids->[0];
$self->{mm}->mid_set($num, $mid0);
- delete($self->{unindexed}) if !keys(%$unindexed);
+ if (scalar(@$u) == 0) { # done with current OID
+ delete $unindexed->{$oidbin};
+ delete($arg->{unindexed}) if !keys(%$unindexed);
+ }
}
}
if (!defined($num)) { # reuse if reindexing (or duplicates)
- my $over = $self->{over};
+ my $oidx = $self->{oidx};
for my $mid (@$mids) {
- ($num, $mid0) = $over->num_mid0_for_oid($oid, $mid);
+ ($num, $mid0) = $oidx->num_mid0_for_oid($oid, $mid);
last if defined $num;
}
}
$mid0 //= do { # is this a number we got before?
- $num = $sync->{mm_tmp}->num_for($mids->[0]);
+ $num = $arg->{mm_tmp}->num_for($mids->[0]);
defined($num) ? $mids->[0] : undef;
};
if (!defined($num)) {
for (my $i = $#$mids; $i >= 1; $i--) {
- $num = $sync->{mm_tmp}->num_for($mids->[$i]);
+ $num = $arg->{mm_tmp}->num_for($mids->[$i]);
if (defined($num)) {
$mid0 = $mids->[$i];
last;
}
}
if (defined($num)) {
- $sync->{mm_tmp}->num_delete($num);
+ $arg->{mm_tmp}->num_delete($num);
} else { # never seen
$num = $self->{mm}->mid_insert($mids->[0]);
if (defined($num)) {
warn "E: $oid <", join('> <', @$mids), "> is a duplicate\n";
return;
}
- $sync->{nr}++;
+ ++${$arg->{nr}};
my $smsg = bless {
- raw_bytes => $len,
num => $num,
blob => $oid,
mid => $mid0,
}, 'PublicInbox::Smsg';
- $smsg->populate($mime, $sync);
- if (do_idx($self, $msgref, $mime, $smsg)) {
- reindex_checkpoint($self, $sync, $git);
+ $smsg->populate($eml, $arg);
+ $smsg->set_bytes($$bref, $size);
+ if (do_idx($self, $eml, $smsg)) {
+ ${$arg->{need_checkpoint}} = 1;
}
+ index_finalize($arg, 1);
}
# only update last_commit for $i on reindex iff newer than current
-sub update_last_commit ($$$$) {
- my ($self, $git, $i, $cmt) = @_;
- my $last = last_epoch_commit($self, $i);
- if (defined $last && is_ancestor($git, $last, $cmt)) {
- my @cmd = (qw(rev-list --count), "$last..$cmt");
- chomp(my $n = $git->qx(@cmd));
+sub update_last_commit {
+ my ($self, $sync, $stk) = @_;
+ my $unit = $sync->{unit} // return;
+ my $latest_cmt = $stk ? $stk->{latest_cmt} : ${$sync->{latest_cmt}};
+ defined($latest_cmt) or return;
+ my $last = last_epoch_commit($self, $unit->{epoch});
+ if (defined $last && is_ancestor($self->git, $last, $latest_cmt)) {
+ my @cmd = (qw(rev-list --count), "$last..$latest_cmt");
+ chomp(my $n = $unit->{git}->qx(@cmd));
return if $n ne '' && $n == 0;
}
- last_epoch_commit($self, $i, $cmt);
+ last_epoch_commit($self, $unit->{epoch}, $latest_cmt);
}
-sub git_dir_n ($$) { "$_[0]->{-inbox}->{inboxdir}/git/$_[1].git" }
-
-sub last_commits ($$) {
- my ($self, $epoch_max) = @_;
+sub last_commits {
+ my ($self, $sync) = @_;
my $heads = [];
- for (my $i = $epoch_max; $i >= 0; $i--) {
+ for (my $i = $sync->{epoch_max}; $i >= 0; $i--) {
$heads->[$i] = last_epoch_commit($self, $i);
}
$heads;
}
-*is_ancestor = *PublicInbox::SearchIdx::is_ancestor;
-
# returns a revision range for git-log(1)
-sub log_range ($$$$$) {
- my ($self, $sync, $git, $i, $tip) = @_;
+sub log_range ($$$) {
+ my ($sync, $unit, $tip) = @_;
my $opt = $sync->{-opt};
my $pr = $opt->{-progress} if (($opt->{verbose} || 0) > 1);
+ my $i = $unit->{epoch};
my $cur = $sync->{ranges}->[$i] or do {
- $pr->("$i.git indexing all of $tip") if $pr;
+ $pr->("$i.git indexing all of $tip\n") if $pr;
return $tip; # all of it
};
my $range = "$cur..$tip";
$pr->("$i.git checking contiguity... ") if $pr;
- if (is_ancestor($git, $cur, $tip)) { # common case
+ my $git = $unit->{git};
+ if (is_ancestor($sync->{self}->git, $cur, $tip)) { # common case
$pr->("OK\n") if $pr;
my $n = $git->qx(qw(rev-list --count), $range);
chomp($n);
warn "discarding history at $cur\n";
}
warn <<"";
-reindexing $git->{git_dir} starting at
-$range
-
- $sync->{unindex_range}->{$i} = "$base..$cur";
+reindexing $git->{git_dir}
+starting at $range
+
+ # $cur^0 may no longer exist if pruned by git
+ if ($git->qx(qw(rev-parse -q --verify), "$cur^0")) {
+ $unit->{unindex_range} = "$base..$cur";
+ } elsif ($base && $git->qx(qw(rev-parse -q --verify), $base)) {
+ $unit->{unindex_range} = "$base..";
+ } else {
+ warn "W: unable to unindex before $range\n";
+ }
}
$range;
}
-sub prepare_range_stack {
- my ($git, $sync, $range) = @_;
- # Don't bump num_highwater on --reindex by using {D}.
- # We intentionally do NOT use {D} in the non-reindex case because
- # we want NNTP article number gaps from unindexed messages to
- # show up in mirrors, too.
- my $D = $sync->{D} //= $sync->{reindex} ? {} : undef; # OID_BIN => NR
-
- my $fh = $git->popen(qw(log --raw -r --pretty=tformat:%at-%ct-%H
- --no-notes --no-color --no-renames --no-abbrev),
- $range);
- my ($at, $ct, $stk);
- while (<$fh>) {
- if (/\A([0-9]+)-([0-9]+)-($OID)$/o) {
- ($at, $ct) = ($1 + 0, $2 + 0);
- $stk //= PublicInbox::IdxStack->new($3);
- } elsif (/\A:\d{6} 100644 $OID ($OID) [AM]\td$/o) {
- my $oid = $1;
- if ($D) { # reindex case
- $D->{pack('H*', $oid)}++;
- } else { # non-reindex case:
- $stk->push_rec('d', $at, $ct, $oid);
- }
- } elsif (/\A:\d{6} 100644 $OID ($OID) [AM]\tm$/o) {
- my $oid = $1;
- if ($D) {
- my $oid_bin = pack('H*', $oid);
- my $nr = --$D->{$oid_bin};
- delete($D->{$oid_bin}) if $nr <= 0;
-
- # nr < 0 (-1) means it never existed
- $stk->push_rec('m', $at, $ct, $oid) if $nr < 0;
- } else {
- $stk->push_rec('m', $at, $ct, $oid);
- }
- }
- }
- close $fh or die "git log failed: \$?=$?";
- $stk ? $stk->read_prepare : undef;
-}
+# overridden by ExtSearchIdx
+sub artnum_max { $_[0]->{mm}->num_highwater }
-sub sync_prepare ($$$) {
- my ($self, $sync, $epoch_max) = @_;
+sub sync_prepare ($$) {
+ my ($self, $sync) = @_;
+ $sync->{ranges} = sync_ranges($self, $sync);
my $pr = $sync->{-opt}->{-progress};
my $regen_max = 0;
- my $head = $self->{-inbox}->{ref_head} || 'refs/heads/master';
-
- # reindex stops at the current heads and we later rerun index_sync
- # without {reindex}
- my $reindex_heads = last_commits($self, $epoch_max) if $sync->{reindex};
+ my $head = $sync->{ibx}->{ref_head} || 'HEAD';
+ my $pfx;
+ if ($pr) {
+ ($pfx) = ($sync->{ibx}->{inboxdir} =~ m!([^/]+)\z!g);
+ $pfx //= $sync->{ibx}->{inboxdir};
+ }
- for (my $i = $epoch_max; $i >= 0; $i--) {
- die 'BUG: already indexing!' if $self->{reindex_pipe};
- my $git_dir = git_dir_n($self, $i);
+ my $reindex_heads;
+ if ($self->{ibx_map}) {
+ # ExtSearchIdx won't index messages unless they're in
+ # over.sqlite3 for a given inbox, so don't read beyond
+ # what's in the per-inbox index.
+ $reindex_heads = [];
+ my $v = PublicInbox::Search::SCHEMA_VERSION;
+ my $mm = $sync->{ibx}->mm;
+ for my $i (0..$sync->{epoch_max}) {
+ $reindex_heads->[$i] = $mm->last_commit_xap($v, $i);
+ }
+ } elsif ($sync->{reindex}) { # V2 inbox
+ # reindex stops at the current heads and we later
+ # rerun index_sync without {reindex}
+ $reindex_heads = $self->last_commits($sync);
+ }
+ if ($sync->{max_size} = $sync->{-opt}->{max_size}) {
+ $sync->{index_oid} = $self->can('index_oid');
+ }
+ my $git_pfx = "$sync->{ibx}->{inboxdir}/git";
+ for (my $i = $sync->{epoch_max}; $i >= 0; $i--) {
+ my $git_dir = "$git_pfx/$i.git";
-d $git_dir or next; # missing epochs are fine
my $git = PublicInbox::Git->new($git_dir);
+ my $unit = { git => $git, epoch => $i };
+ my $tip;
if ($reindex_heads) {
- $head = $reindex_heads->[$i] or next;
+ $tip = $head = $reindex_heads->[$i] or next;
+ } else {
+ $tip = $git->qx(qw(rev-parse -q --verify), $head);
+ next if $?; # new repo
+ chomp $tip;
}
- chomp(my $tip = $git->qx(qw(rev-parse -q --verify), $head));
-
- next if $?; # new repo
- my $range = log_range($self, $sync, $git, $i, $tip) or next;
+ my $range = log_range($sync, $unit, $tip) or next;
# can't use 'rev-list --count' if we use --diff-filter
- $pr->("$i.git counting $range ... ") if $pr;
- my $stk = prepare_range_stack($git, $sync, $range);
+ $pr->("$pfx $i.git counting $range ... ") if $pr;
+ # Don't bump num_highwater on --reindex by using {D}.
+ # We intentionally do NOT use {D} in the non-reindex case
+ # because we want NNTP article number gaps from unindexed
+ # messages to show up in mirrors, too.
+ $sync->{D} //= $sync->{reindex} ? {} : undef; # OID_BIN => NR
+ my $stk = log2stack($sync, $git, $range);
+ return 0 if $sync->{quit};
my $nr = $stk ? $stk->num_records : 0;
$pr->("$nr\n") if $pr;
- $sync->{stacks}->[$i] = $stk if $stk;
+ $unit->{stack} = $stk; # may be undef
+ unshift @{$sync->{todo}}, $unit;
$regen_max += $nr;
}
+ return 0 if $sync->{quit};
# XXX this should not happen unless somebody bypasses checks in
# our code and blindly injects "d" file history into git repos
if (my @leftovers = keys %{delete($sync->{D}) // {}}) {
warn('W: unindexing '.scalar(@leftovers)." leftovers\n");
- my $git = $self->{-inbox}->git;
+ local $self->{current_info} = 'leftover ';
+ my $unindex_oid = $self->can('unindex_oid');
for my $oid (@leftovers) {
+ last if $sync->{quit};
$oid = unpack('H*', $oid);
- $self->{current_info} = "leftover $oid";
- unindex_oid($self, $git, $oid);
+ my $req = { %$sync, oid => $oid };
+ $self->git->cat_async($oid, $unindex_oid, $req);
}
- $git->cleanup;
+ $self->git->cat_async_wait;
+ }
+ return 0 if $sync->{quit};
+ if (!$regen_max) {
+ $sync->{-regen_fmt} = "%u/?\n";
+ return 0;
}
- return 0 if (!$regen_max && !keys(%{$self->{unindex_range}}));
# reindex should NOT see new commits anymore, if we do,
# it's a problem and we need to notice it via die()
my $pad = length($regen_max) + 1;
$sync->{-regen_fmt} = "% ${pad}u/$regen_max\n";
- $sync->{nr} = 0;
+ $sync->{nr} = \(my $nr = 0);
return -1 if $sync->{reindex};
- $regen_max + $self->{mm}->num_highwater() || 0;
+ $regen_max + $self->artnum_max || 0;
}
-sub unindex_oid_remote ($$$) {
+sub unindex_oid_aux ($$$) {
my ($self, $oid, $mid) = @_;
- my @removed = $self->{over}->remove_oid($oid, $mid);
+ my @removed = $self->{oidx}->remove_oid($oid, $mid);
+ return unless $self->{-need_xapian};
for my $num (@removed) {
- my $idx = idx_shard($self, $num % $self->{shards});
- $idx->remote_remove($oid, $num);
+ idx_shard($self, $num)->ipc_do('xdb_remove', $num);
}
}
-sub unindex_oid ($$$;$) {
- my ($self, $git, $oid, $unindexed) = @_;
+sub unindex_oid ($$;$) { # git->cat_async callback
+ my ($bref, $oid, $type, $size, $arg) = @_;
+ is_bad_blob($oid, $type, $size, $arg->{oid}) and
+ return index_finalize($arg, 0);
+ my $self = $arg->{self};
+ local $self->{current_info} = "$self->{current_info} $oid";
+ my $unindexed = $arg->{in_unindex} ? $arg->{unindexed} : undef;
my $mm = $self->{mm};
- my $msgref = $git->cat_file($oid);
- my $mime = PublicInbox::Eml->new($msgref);
- my $mids = mids($mime->header_obj);
- $mime = $msgref = undef;
- my $over = $self->{over};
+ my $mids = mids(PublicInbox::Eml->new($bref));
+ undef $$bref;
+ my $oidx = $self->{oidx};
foreach my $mid (@$mids) {
my %gone;
my ($id, $prev);
- while (my $smsg = $over->next_by_mid($mid, \$id, \$prev)) {
+ while (my $smsg = $oidx->next_by_mid($mid, \$id, \$prev)) {
$gone{$smsg->{num}} = 1 if $oid eq $smsg->{blob};
}
my $n = scalar(keys(%gone)) or next;
warn "BUG: multiple articles linked to $oid\n",
join(',',sort keys %gone), "\n";
}
- foreach my $num (keys %gone) {
+ # reuse (num => mid) mapping in ascending numeric order
+ for my $num (sort { $a <=> $b } keys %gone) {
+ $num += 0;
if ($unindexed) {
my $mid0 = $mm->mid_for($num);
- $unindexed->{$mid0} = $num;
+ my $oidbin = pack('H*', $oid);
+ push @{$unindexed->{$oidbin}}, $num, $mid0;
}
$mm->num_delete($num);
}
- unindex_oid_remote($self, $oid, $mid);
+ unindex_oid_aux($self, $oid, $mid);
}
+ index_finalize($arg, 0);
}
-sub unindex ($$$$) {
- my ($self, $sync, $git, $unindex_range) = @_;
- my $unindexed = $self->{unindexed} ||= {}; # $mid0 => $num
+sub git { $_[0]->{ibx}->git }
+
+# this is rare, it only happens when we get discontiguous history in
+# a mirror because the source used -purge or -edit
+sub unindex_todo ($$$) {
+ my ($self, $sync, $unit) = @_;
+ my $unindex_range = delete($unit->{unindex_range}) // return;
+ my $unindexed = $sync->{unindexed} //= {}; # $oidbin => [$num, $mid0]
my $before = scalar keys %$unindexed;
# order does not matter, here:
- my @cmd = qw(log --raw -r
- --no-notes --no-color --no-abbrev --no-renames);
- my $fh = $self->{reindex_pipe} = $git->popen(@cmd, $unindex_range);
+ my $fh = $unit->{git}->popen(qw(log --raw -r --no-notes --no-color
+ --no-abbrev --no-renames), $unindex_range);
+ local $sync->{in_unindex} = 1;
+ my $unindex_oid = $self->can('unindex_oid');
while (<$fh>) {
/\A:\d{6} 100644 $OID ($OID) [AM]\tm$/o or next;
- unindex_oid($self, $git, $1, $unindexed);
+ $self->git->cat_async($1, $unindex_oid, { %$sync, oid => $1 });
}
- delete $self->{reindex_pipe};
close $fh or die "git log failed: \$?=$?";
+ $self->git->cat_async_wait;
return unless $sync->{-opt}->{prune};
my $after = scalar keys %$unindexed;
return if $before == $after;
# ensure any blob can not longer be accessed via dumb HTTP
- PublicInbox::Import::run_die(['git', "--git-dir=$git->{git_dir}",
+ run_die(['git', "--git-dir=$unit->{git}->{git_dir}",
qw(-c gc.reflogExpire=now gc --prune=all --quiet)]);
}
-sub sync_ranges ($$$) {
- my ($self, $sync, $epoch_max) = @_;
+sub sync_ranges ($$) {
+ my ($self, $sync) = @_;
my $reindex = $sync->{reindex};
-
- return last_commits($self, $epoch_max) unless $reindex;
+ return $self->last_commits($sync) unless $reindex;
return [] if ref($reindex) ne 'HASH';
my $ranges = $reindex->{from}; # arrayref;
$ranges;
}
-sub index_epoch ($$$) {
- my ($self, $sync, $i) = @_;
+sub index_xap_only { # git->cat_async callback
+ my ($bref, $oid, $type, $size, $smsg) = @_;
+ my $self = delete $smsg->{self};
+ my $idx = idx_shard($self, $smsg->{num});
+ $idx->index_eml(PublicInbox::Eml->new($bref), $smsg);
+ $self->{transact_bytes} += $smsg->{bytes};
+}
+
+sub index_xap_step ($$$;$) {
+ my ($self, $sync, $beg, $step) = @_;
+ my $end = $sync->{art_end};
+ return if $beg > $end; # nothing to do
+
+ $step //= $self->{shards};
+ my $ibx = $self->{ibx};
+ if (my $pr = $sync->{-opt}->{-progress}) {
+ $pr->("Xapian indexlevel=$ibx->{indexlevel} ".
+ "$beg..$end (% $step)\n");
+ }
+ for (my $num = $beg; $num <= $end; $num += $step) {
+ last if $sync->{quit};
+ my $smsg = $ibx->over->get_art($num) or next;
+ $smsg->{self} = $self;
+ $ibx->git->cat_async($smsg->{blob}, \&index_xap_only, $smsg);
+ if ($self->{transact_bytes} >= $self->{batch_bytes}) {
+ ${$sync->{nr}} = $num;
+ reindex_checkpoint($self, $sync);
+ }
+ }
+}
- my $git_dir = git_dir_n($self, $i);
- die 'BUG: already reindexing!' if $self->{reindex_pipe};
- -d $git_dir or return; # missing epochs are fine
- my $git = PublicInbox::Git->new($git_dir);
- if (my $unindex_range = delete $sync->{unindex_range}->{$i}) {
- unindex($self, $sync, $git, $unindex_range);
+sub index_todo ($$$) {
+ my ($self, $sync, $unit) = @_;
+ return if $sync->{quit};
+ unindex_todo($self, $sync, $unit);
+ my $stk = delete($unit->{stack}) or return;
+ my $all = $self->git;
+ my $index_oid = $self->can('index_oid');
+ my $unindex_oid = $self->can('unindex_oid');
+ my $pfx;
+ if ($unit->{git}->{git_dir} =~ m!/([^/]+)/git/([0-9]+\.git)\z!) {
+ $pfx = "$1 $2"; # v2
+ } else { # v1
+ ($pfx) = ($unit->{git}->{git_dir} =~ m!/([^/]+)\z!g);
+ $pfx //= $unit->{git}->{git_dir};
}
- defined(my $stk = $sync->{stacks}->[$i]) or return;
- $sync->{stacks}->[$i] = undef;
- while (my ($f, $at, $ct, $oid) = $stk->pop_rec) {
- $self->{current_info} = "$i.git $oid";
+ local $self->{current_info} = "$pfx ";
+ local $sync->{latest_cmt} = \(my $latest_cmt);
+ local $sync->{unit} = $unit;
+ while (my ($f, $at, $ct, $oid, $cmt) = $stk->pop_rec) {
+ if ($sync->{quit}) {
+ warn "waiting to quit...\n";
+ $all->async_wait_all;
+ $self->update_last_commit($sync);
+ return;
+ }
+ my $req = {
+ %$sync,
+ autime => $at,
+ cotime => $ct,
+ oid => $oid,
+ cur_cmt => $cmt
+ };
if ($f eq 'm') {
- $sync->{autime} = $at;
- $sync->{cotime} = $ct;
- reindex_oid($self, $sync, $git, $oid);
+ if ($sync->{max_size}) {
+ $all->check_async($oid, \&check_size, $req);
+ } else {
+ $all->cat_async($oid, $index_oid, $req);
+ }
} elsif ($f eq 'd') {
- unindex_oid($self, $git, $oid);
+ $all->cat_async($oid, $unindex_oid, $req);
+ }
+ if (${$sync->{need_checkpoint}}) {
+ reindex_checkpoint($self, $sync);
+ }
+ }
+ $all->async_wait_all;
+ $self->update_last_commit($sync, $stk);
+}
+
+sub xapian_only {
+ my ($self, $opt, $sync, $art_beg) = @_;
+ my $seq = $opt->{sequential_shard};
+ $art_beg //= 0;
+ local $self->{parallel} = 0 if $seq;
+ $self->idx_init($opt); # acquire lock
+ if (my $art_end = $self->{ibx}->mm->max) {
+ $sync //= {
+ need_checkpoint => \(my $bool = 0),
+ -opt => $opt,
+ self => $self,
+ nr => \(my $nr = 0),
+ -regen_fmt => "%u/?\n",
+ };
+ $sync->{art_end} = $art_end;
+ if ($seq || !$self->{parallel}) {
+ my $shard_end = $self->{shards} - 1;
+ for my $i (0..$shard_end) {
+ last if $sync->{quit};
+ index_xap_step($self, $sync, $art_beg + $i);
+ if ($i != $shard_end) {
+ reindex_checkpoint($self, $sync);
+ }
+ }
+ } else { # parallel (maybe)
+ index_xap_step($self, $sync, $art_beg, 1);
}
}
- delete @$sync{qw(autime cotime)};
- update_last_commit($self, $git, $i, $stk->{latest_cmt});
+ $self->git->cat_async_wait;
+ $self->{ibx}->cleanup;
+ $self->done;
}
# public, called by public-inbox-index
sub index_sync {
my ($self, $opt) = @_;
- $opt ||= {};
- my $pr = $opt->{-progress};
+ $opt //= {};
+ return xapian_only($self, $opt) if $opt->{xapian_only};
+
my $epoch_max;
- my $latest = git_dir_latest($self, \$epoch_max);
- return unless defined $latest;
+ my $latest = $self->{ibx}->git_dir_latest(\$epoch_max) // return;
+ if ($opt->{'fast-noop'}) { # nanosecond (st_ctim) comparison
+ use Time::HiRes qw(stat);
+ if (my @mm = stat("$self->{ibx}->{inboxdir}/msgmap.sqlite3")) {
+ my $c = $mm[10]; # 10 = ctime (nsec NV)
+ my @hd = stat("$latest/refs/heads");
+ my @pr = stat("$latest/packed-refs");
+ return if $c > ($hd[10] // 0) && $c > ($pr[10] // 0);
+ }
+ }
+
+ my $pr = $opt->{-progress};
+ my $seq = $opt->{sequential_shard};
+ my $art_beg; # the NNTP article number we start xapian_only at
+ my $idxlevel = $self->{ibx}->{indexlevel};
+ local $self->{ibx}->{indexlevel} = 'basic' if $seq;
+
$self->idx_init($opt); # acquire lock
fill_alternates($self, $epoch_max);
- $self->{over}->rethread_prepare($opt);
+ $self->{oidx}->rethread_prepare($opt);
my $sync = {
- unindex_range => {}, # EPOCH => oid_old..oid_new
+ need_checkpoint => \(my $bool = 0),
reindex => $opt->{reindex},
- -opt => $opt
+ -opt => $opt,
+ self => $self,
+ ibx => $self->{ibx},
+ epoch_max => $epoch_max,
};
- $sync->{ranges} = sync_ranges($self, $sync, $epoch_max);
- if (sync_prepare($self, $sync, $epoch_max)) {
+ my $quit = PublicInbox::SearchIdx::quit_cb($sync);
+ local $SIG{QUIT} = $quit;
+ local $SIG{INT} = $quit;
+ local $SIG{TERM} = $quit;
+
+ if (sync_prepare($self, $sync)) {
# tmp_clone seems to fail if inside a transaction, so
# we rollback here (because we opened {mm} for reading)
# Note: we do NOT rely on DBI transactions for atomicity;
# only for batch performance.
$self->{mm}->{dbh}->rollback;
$self->{mm}->{dbh}->begin_work;
- $sync->{mm_tmp} = $self->{mm}->tmp_clone;
- }
+ $sync->{mm_tmp} =
+ $self->{mm}->tmp_clone($self->{ibx}->{inboxdir});
+ # xapian_only works incrementally w/o --reindex
+ if ($seq && !$opt->{reindex}) {
+ $art_beg = $sync->{mm_tmp}->max || -1;
+ $art_beg++;
+ }
+ }
# work forwards through history
- index_epoch($self, $sync, $_) for (0..$epoch_max);
+ index_todo($self, $sync, $_) for @{delete($sync->{todo}) // []};
+ $self->{oidx}->rethread_done($opt) unless $sync->{quit};
$self->done;
if (my $nr = $sync->{nr}) {
my $pr = $sync->{-opt}->{-progress};
- $pr->('all.git '.sprintf($sync->{-regen_fmt}, $nr)) if $pr;
+ $pr->('all.git '.sprintf($sync->{-regen_fmt}, $$nr)) if $pr;
+ }
+
+ my $quit_warn;
+ # deal with Xapian shards sequentially
+ if ($seq && delete($sync->{mm_tmp})) {
+ if ($sync->{quit}) {
+ $quit_warn = 1;
+ } else {
+ $self->{ibx}->{indexlevel} = $idxlevel;
+ xapian_only($self, $opt, $sync, $art_beg);
+ $quit_warn = 1 if $sync->{quit};
+ }
+ }
+
+ # --reindex on the command-line
+ if (!$sync->{quit} && $opt->{reindex} &&
+ !ref($opt->{reindex}) && $idxlevel ne 'basic') {
+ $self->lock_acquire;
+ my $s0 = PublicInbox::SearchIdx->new($self->{ibx}, 0, 0);
+ if (my $xdb = $s0->idx_acquire) {
+ my $n = $xdb->get_metadata('has_threadid');
+ $xdb->set_metadata('has_threadid', '1') if $n ne '1';
+ }
+ $s0->idx_release;
+ $self->lock_release;
}
- $self->{over}->rethread_done($opt);
# reindex does not pick up new changes, so we rerun w/o it:
- if ($opt->{reindex}) {
+ if ($opt->{reindex} && !$sync->{quit}) {
my %again = %$opt;
$sync = undef;
delete @again{qw(rethread reindex -skip_lock)};
index_sync($self, \%again);
+ $opt->{quit} = $again{quit}; # propagate to caller
}
+ warn <<EOF if $quit_warn;
+W: interrupted, --xapian-only --reindex required upon restart
+EOF
}
1;