-# 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::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 qw(log2stack crlf_adjust is_ancestor check_size);
+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 ();
+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 () {
- # getconf(1) is POSIX, but *NPROCESSORS* vars are not
- for (qw(_NPROCESSORS_ONLN NPROCESSORS_ONLN)) {
- `getconf $_ 2>/dev/null` =~ /^(\d+)$/ and return $1;
- }
- for my $nproc (qw(nproc gnproc)) { # GNU coreutils nproc
- `$nproc 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) = @_;
- $self->{ibx} ? do {
+ 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 = $self->{ibx}->search or return 0;
- delete $self->{ibx}->{search};
+ my $srch = $ibx->search or return 0;
+ delete $ibx->{search};
$srch->{nshard} // 0
- } : $self->{nshard}; # self->{nshard} is for ExtSearchIdx
+ } else { # ExtSearchIdx
+ $self->{nshard} ||= scalar($self->xdb_shards_flat);
+ }
}
sub new {
die "$dir does not exist\n";
}
}
- $v2ibx->umask_prepare;
-
my $xpfx = "$dir/xap" . PublicInbox::Search::SCHEMA_VERSION;
my $self = {
ibx => $v2ibx,
}
$self->idx_init;
$self->{mm}->skip_artnum($skip_artnum) if defined $skip_artnum;
- my $epoch_max = -1;
- $self->{ibx}->git_dir_latest(\$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;
}
}
# indexes a message, returns true if checkpointing is needed
-sub do_idx ($$$$) {
- my ($self, $msgref, $mime, $smsg) = @_;
- $smsg->{bytes} = $smsg->{raw_bytes} + crlf_adjust($$msgref);
- $self->{oidx}->add_overview($mime, $smsg);
- my $idx = idx_shard($self, $smsg->{num});
- $idx->index_raw($msgref, $mime, $smsg);
- my $n = $self->{transact_bytes} += $smsg->{raw_bytes};
+sub do_idx ($$$) {
+ my ($self, $eml, $smsg) = @_;
+ $self->{oidx}->add_overview($eml, $smsg);
+ if ($self->{-need_xapian}) {
+ my $idx = idx_shard($self, $smsg->{num});
+ $idx->index_eml($eml, $smsg);
+ }
+ my $n = $self->{transact_bytes} += $smsg->{bytes};
$n >= $self->{batch_bytes};
}
$cmt = $im->get_mark($cmt);
$self->{last_commit}->[$self->{epoch_max}] = $cmt;
- my $msgref = delete $smsg->{-raw_email};
- if (do_idx($self, $msgref, $mime, $smsg)) {
+ if (do_idx($self, $mime, $smsg)) {
$self->checkpoint;
}
$self->{shards} = $nshards if $nshards && $nshards != $self->{shards};
$self->{batch_bytes} = $opt->{batch_size} //
$PublicInbox::SearchIdx::BATCH_BYTES;
- $self->{batch_bytes} *= $self->{shards} if $self->{parallel};
# need to create all shards before initializing msgmap FD
# idx_shards must be visible to all forked processes
my $max = $self->{shards} - 1;
my $idx = $self->{idx_shards} = [];
push @$idx, PublicInbox::SearchIdxShard->new($self, $_) for (0..$max);
+ $self->{-need_xapian} = $idx->[0]->need_xapian;
+
+ # SearchIdxShard may do their own flushing, so don't scale
+ # until after forking
+ $self->{batch_bytes} *= $self->{shards} if $self->{parallel};
+
my $ibx = $self->{ibx} or return; # ExtIdxSearch
# Now that all subprocesses are up, we can open the FDs
sub parallel_init ($$) {
my ($self, $indexlevel) = @_;
- if (($indexlevel // 'full') eq 'basic') {
- $self->{parallel} = 0;
- } else {
- pipe(my ($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);
- }
+ $self->{parallel} = 0 if ($indexlevel // 'full') eq 'basic';
}
# idempotent
$ibx->git->cleanup;
parallel_init($self, $ibx->{indexlevel});
- $ibx->umask_prepare;
$ibx->with_umask(\&_idx_init, $self, $opt);
}
my $ibx = $self->{ibx};
my $pfx = "$ibx->{inboxdir}/git";
my $rewrites = []; # epoch => commit
- my $max = $self->{epoch_max};
-
- unless (defined($max)) {
- defined(my $latest = $ibx->git_dir_latest(\$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";
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);
}
$rewritten->{rewrites};
}
}
}
-sub barrier_init {
- my ($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) = @_;
$self->{oidx}->commit_lazy;
# Now deal with Xapian
- if ($wait) {
- my $barrier = barrier_init($self, scalar @$shards);
- # each shard needs to issue a barrier command
- $_->shard_barrier for @$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
- barrier_wait($self, $barrier);
- } else {
- $_->shard_commit for @$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
}
eval { $self->{oidx}->dbh_close };
$err .= "over close: $@\n" if $@;
- delete $self->{bnote};
+ delete $self->{midx};
my $nbytes = $self->{total_bytes};
$self->{total_bytes} = 0;
$self->lock_release(!!$nbytes) if $shards;
sub write_alternates ($$$) {
my ($info_dir, $mode, $out) = @_;
- my $fh = File::Temp->new(TEMPLATE => 'alt-XXXXXXXX', DIR => $info_dir);
+ my $fh = File::Temp->new(TEMPLATE => 'alt-XXXX', DIR => $info_dir);
my $tmp = $fh->filename;
print $fh @$out or die "print $tmp: $!\n";
chmod($mode, $fh) or die "fchmod $tmp: $!\n";
my ($self, $epoch) = @_;
my $git_dir = "$self->{ibx}->{inboxdir}/git/$epoch.git";
PublicInbox::Import::init_bare($git_dir);
- my @cmd = (qw/git config/, "--file=$git_dir/config",
- 'include.path', '../../all.git/config');
- PublicInbox::Import::run_die(\@cmd);
+ run_die([qw(git config), "--file=$git_dir/config",
+ qw(include.path ../../all.git/config)]);
fill_alternates($self, $epoch);
$git_dir
}
sub diff ($$$) {
my ($mid, $cur, $new) = @_;
- my $ah = File::Temp->new(TEMPLATE => 'email-cur-XXXXXXXX', TMPDIR => 1);
+ my $ah = File::Temp->new(TEMPLATE => 'email-cur-XXXX', TMPDIR => 1);
print $ah $cur->as_string or die "print: $!";
$ah->flush or die "flush: $!";
PublicInbox::Import::drop_unwanted_headers($new);
- my $bh = File::Temp->new(TEMPLATE => 'email-new-XXXXXXXX', TMPDIR => 1);
+ my $bh = File::Temp->new(TEMPLATE => 'email-new-XXXX', TMPDIR => 1);
print $bh $new->as_string or die "print: $!";
$bh->flush or die "flush: $!";
my $cmd = [ qw(diff -u), $ah->filename, $bh->filename ];
sub atfork_child {
my ($self) = @_;
- 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 ($$) {
my ($self, $sync) = @_;
$self->git->async_wait_all;
+ $self->update_last_commit($sync);
${$sync->{need_checkpoint}} = 0;
my $mm_tmp = $sync->{mm_tmp};
$mm_tmp->atfork_prepare if $mm_tmp;
die 'BUG: {im} during reindex' if $self->{im};
- if ($self->{ibx_map}) {
+ if ($self->{ibx_map} && !$sync->{checkpoint_unlocks}) {
checkpoint($self, 1); # no need to release lock on pure index
} else {
$self->done; # release lock
}
- if (my $pr = $sync->{-opt}->{-progress}) {
+ if (my $pr = $sync->{-regen_fmt} ? $sync->{-opt}->{-progress} : undef) {
$pr->(sprintf($sync->{-regen_fmt}, ${$sync->{nr}}));
}
# allow -watch or -mda to write...
$self->idx_init($sync->{-opt}); # reacquire lock
+ if (my $intvl = $sync->{check_intvl}) { # eidx
+ $sync->{next_check} = PublicInbox::DS::now() + $intvl;
+ }
$mm_tmp->atfork_parent if $mm_tmp;
}
+sub index_finalize ($$) {
+ 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";
- return if $size == 0; # purged
my ($num, $mid0);
my $eml = PublicInbox::Eml->new($$bref);
my $mids = mids($eml);
}
# {unindexed} is unlikely
- if ((my $unindexed = $arg->{unindexed}) && scalar(@$mids) == 1) {
- $num = delete($unindexed->{$mids->[0]});
+ if (my $unindexed = $arg->{unindexed}) {
+ my $oidbin = pack('H*', $oid);
+ my $u = $unindexed->{$oidbin};
+ ($num, $mid0) = splice(@$u, 0, 2) if $u;
if (defined $num) {
- $mid0 = $mids->[0];
$self->{mm}->mid_set($num, $mid0);
- delete($arg->{unindexed}) if !keys(%$unindexed);
+ if (scalar(@$u) == 0) { # done with current OID
+ delete $unindexed->{$oidbin};
+ delete($arg->{unindexed}) if !keys(%$unindexed);
+ }
}
}
if (!defined($num)) { # reuse if reindexing (or duplicates)
}
++${$arg->{nr}};
my $smsg = bless {
- raw_bytes => $size,
num => $num,
blob => $oid,
mid => $mid0,
}, 'PublicInbox::Smsg';
$smsg->populate($eml, $arg);
- if (do_idx($self, $bref, $eml, $smsg)) {
+ $smsg->set_bytes($$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
-# $sync will be used by subclasses
sub update_last_commit {
- my ($self, $sync, $unit, $cmt) = @_;
+ my ($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 $last && is_ancestor($unit->{git}, $last, $cmt)) {
- my @cmd = (qw(rev-list --count), "$last..$cmt");
+ 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, $unit->{epoch}, $cmt);
+ last_epoch_commit($self, $unit->{epoch}, $latest_cmt);
}
sub last_commits {
my $range = "$cur..$tip";
$pr->("$i.git checking contiguity... ") if $pr;
my $git = $unit->{git};
- if (is_ancestor($git, $cur, $tip)) { # common case
+ if (is_ancestor($sync->{self}->git, $cur, $tip)) { # common case
$pr->("OK\n") if $pr;
my $n = $git->qx(qw(rev-list --count), $range);
chomp($n);
$pfx //= $sync->{ibx}->{inboxdir};
}
- # reindex stops at the current heads and we later rerun index_sync
- # without {reindex}
- my $reindex_heads = $self->last_commits($sync) if $sync->{reindex};
-
+ 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 = $sync->{ibx}->git_dir_n($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 $range = log_range($sync, $unit, $tip) or next;
# can't use 'rev-list --count' if we use --diff-filter
$pr->("$pfx $i.git counting $range ... ") if $pr;
# 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;
$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
local $self->{current_info} = 'leftover ';
my $unindex_oid = $self->can('unindex_oid');
for my $oid (@leftovers) {
+ last if $sync->{quit};
$oid = unpack('H*', $oid);
my $req = { %$sync, oid => $oid };
$self->git->cat_async($oid, $unindex_oid, $req);
}
$self->git->cat_async_wait;
}
+ return 0 if $sync->{quit};
if (!$regen_max) {
$sync->{-regen_fmt} = "%u/?\n";
return 0;
sub unindex_oid_aux ($$$) {
my ($self, $oid, $mid) = @_;
my @removed = $self->{oidx}->remove_oid($oid, $mid);
+ return unless $self->{-need_xapian};
for my $num (@removed) {
- my $idx = idx_shard($self, $num);
- $idx->shard_remove($oid, $num);
+ idx_shard($self, $num)->ipc_do('xdb_remove', $num);
}
}
sub unindex_oid ($$;$) { # git->cat_async callback
- my ($bref, $oid, $type, $size, $sync) = @_;
- my $self = $sync->{self};
+ my ($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 = $sync->{in_unindex} ? $sync->{unindexed} : undef;
+ my $unindexed = $arg->{in_unindex} ? $arg->{unindexed} : undef;
my $mm = $self->{mm};
my $mids = mids(PublicInbox::Eml->new($bref));
undef $$bref;
warn "BUG: multiple articles linked to $oid\n",
join(',',sort keys %gone), "\n";
}
- foreach my $num (keys %gone) {
+ # reuse (num => mid) mapping in ascending numeric order
+ for my $num (sort { $a <=> $b } keys %gone) {
+ $num += 0;
if ($unindexed) {
my $mid0 = $mm->mid_for($num);
- $unindexed->{$mid0} = $num;
+ my $oidbin = pack('H*', $oid);
+ push @{$unindexed->{$oidbin}}, $num, $mid0;
}
$mm->num_delete($num);
}
unindex_oid_aux($self, $oid, $mid);
}
+ index_finalize($arg, 0);
}
sub git { $_[0]->{ibx}->git }
sub unindex_todo ($$$) {
my ($self, $sync, $unit) = @_;
my $unindex_range = delete($unit->{unindex_range}) // return;
- my $unindexed = $sync->{unindexed} //= {}; # $mid0 => $num
+ my $unindexed = $sync->{unindexed} //= {}; # $oidbin => [$num, $mid0]
my $before = scalar keys %$unindexed;
# order does not matter, here:
my $fh = $unit->{git}->popen(qw(log --raw -r --no-notes --no-color
return if $before == $after;
# ensure any blob can not longer be accessed via dumb HTTP
- PublicInbox::Import::run_die(['git',
- "--git-dir=$unit->{git}->{git_dir}",
+ run_die(['git', "--git-dir=$unit->{git}->{git_dir}",
qw(-c gc.reflogExpire=now gc --prune=all --quiet)]);
}
sub index_xap_only { # git->cat_async callback
my ($bref, $oid, $type, $size, $smsg) = @_;
- my $self = $smsg->{self};
+ my $self = delete $smsg->{self};
my $idx = idx_shard($self, $smsg->{num});
- $smsg->{raw_bytes} = $size;
- $idx->index_raw($bref, undef, $smsg);
- $self->{transact_bytes} += $size;
+ $idx->index_eml(PublicInbox::Eml->new($bref), $smsg);
+ $self->{transact_bytes} += $smsg->{bytes};
}
sub index_xap_step ($$$;$) {
"$beg..$end (% $step)\n");
}
for (my $num = $beg; $num <= $end; $num += $step) {
+ last if $sync->{quit};
my $smsg = $ibx->over->get_art($num) or next;
$smsg->{self} = $self;
$ibx->git->cat_async($smsg->{blob}, \&index_xap_only, $smsg);
sub index_todo ($$$) {
my ($self, $sync, $unit) = @_;
+ return if $sync->{quit};
unindex_todo($self, $sync, $unit);
my $stk = delete($unit->{stack}) or return;
my $all = $self->git;
$pfx //= $unit->{git}->{git_dir};
}
local $self->{current_info} = "$pfx ";
- while (my ($f, $at, $ct, $oid) = $stk->pop_rec) {
- my $req = { %$sync, autime => $at, cotime => $ct, oid => $oid };
+ 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') {
if ($sync->{max_size}) {
$all->check_async($oid, \&check_size, $req);
}
}
$all->async_wait_all;
- $self->update_last_commit($sync, $unit, $stk->{latest_cmt});
+ $self->update_last_commit($sync, $stk);
}
sub xapian_only {
if ($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);
}
}
$self->git->cat_async_wait;
+ $self->{ibx}->cleanup;
$self->done;
}
$opt //= {};
return xapian_only($self, $opt) if $opt->{xapian_only};
- my $pr = $opt->{-progress};
my $epoch_max;
- my $latest = $self->{ibx}->git_dir_latest(\$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};
ibx => $self->{ibx},
epoch_max => $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, $sync)) {
# tmp_clone seems to fail if inside a transaction, so
# we rollback here (because we opened {mm} for reading)
# xapian_only works incrementally w/o --reindex
if ($seq && !$opt->{reindex}) {
- $art_beg = $sync->{mm_tmp}->max;
- $art_beg++ if defined($art_beg);
+ $art_beg = $sync->{mm_tmp}->max || -1;
+ $art_beg++;
}
}
# work forwards through history
- index_todo($self, $sync, $_) for @{$sync->{todo}};
- $self->{oidx}->rethread_done($opt);
+ index_todo($self, $sync, $_) for @{delete($sync->{todo}) // []};
+ $self->{oidx}->rethread_done($opt) unless $sync->{quit};
$self->done;
if (my $nr = $sync->{nr}) {
$pr->('all.git '.sprintf($sync->{-regen_fmt}, $$nr)) if $pr;
}
+ my $quit_warn;
# deal with Xapian shards sequentially
if ($seq && delete($sync->{mm_tmp})) {
- $self->{ibx}->{indexlevel} = $idxlevel;
- xapian_only($self, $opt, $sync, $art_beg);
+ if ($sync->{quit}) {
+ $quit_warn = 1;
+ } else {
+ $self->{ibx}->{indexlevel} = $idxlevel;
+ xapian_only($self, $opt, $sync, $art_beg);
+ $quit_warn = 1 if $sync->{quit};
+ }
}
# --reindex on the command-line
- if ($opt->{reindex} && !ref($opt->{reindex}) && $idxlevel ne 'basic') {
+ if (!$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 = $s0->idx_acquire) {
}
# reindex does not pick up new changes, so we rerun w/o it:
- if ($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;