X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FV2Writable.pm;h=37d27302565e17b66586b00e5742a6b2ef451f41;hb=2710d3105e70ff467eff9e977325628b4e9dd1c5;hp=33c0038d18ebac631c2b2a9c84f55222d5d3d5ac;hpb=96a8b8832d791d53e05f235b5bc2492daab5f60a;p=public-inbox.git diff --git a/lib/PublicInbox/V2Writable.pm b/lib/PublicInbox/V2Writable.pm index 33c0038d..37d27302 100644 --- a/lib/PublicInbox/V2Writable.pm +++ b/lib/PublicInbox/V2Writable.pm @@ -16,9 +16,10 @@ use PublicInbox::ContentId qw(content_id content_digest); use PublicInbox::Inbox; use PublicInbox::OverIdx; use PublicInbox::Msgmap; -use PublicInbox::Spawn qw(spawn); +use PublicInbox::Spawn qw(spawn popen_rd); use PublicInbox::SearchIdx; -use IO::Handle; +use IO::Handle; # ->autoflush +use File::Temp qw(tempfile); # an estimate of the post-packed size to the raw uncompressed size my $PACKING_FACTOR = 0.4; @@ -33,13 +34,8 @@ our $NPROC_MAX_DEFAULT = 4; sub nproc_shards ($) { my ($creat_opt) = @_; - if (ref($creat_opt) eq 'HASH') { - if (defined(my $n = $creat_opt->{nproc})) { - return $n - } - } - - my $n = $ENV{NPROC}; + my $n = $creat_opt->{nproc} if ref($creat_opt) eq 'HASH'; + $n //= $ENV{NPROC}; if (!$n) { chomp($n = `nproc 2>/dev/null`); # assume 2 cores if GNU nproc(1) is not available @@ -61,10 +57,13 @@ sub count_shards ($) { # Also, shard count may change while -watch is running # due to "xcpdb --reshard" if (-d $xpfx) { + require PublicInbox::Search; + PublicInbox::Search::load_xapian(); + my $XapianDatabase = $PublicInbox::Search::X{Database}; foreach my $shard (<$xpfx/*>) { -d $shard && $shard =~ m!/[0-9]+\z! or next; eval { - Search::Xapian::Database->new($shard)->close; + $XapianDatabase->new($shard)->close; $n++; }; } @@ -76,7 +75,8 @@ sub new { # $creat may be any true value, or 0/undef. A hashref is true, # and $creat->{nproc} may be set to an integer my ($class, $v2ibx, $creat) = @_; - my $dir = $v2ibx->{inboxdir} or die "no inboxdir in inbox\n"; + $v2ibx = PublicInbox::InboxWritable->new($v2ibx); + my $dir = $v2ibx->assert_usable_dir; unless (-d $dir) { if ($creat) { require File::Path; @@ -85,8 +85,6 @@ sub new { die "$dir does not exist\n"; } } - - $v2ibx = PublicInbox::InboxWritable->new($v2ibx); $v2ibx->umask_prepare; my $xpfx = "$dir/xap" . PublicInbox::Search::SCHEMA_VERSION; @@ -468,17 +466,12 @@ sub git_hash_raw ($$) { print $tmp_fh $$raw or die "print \$tmp_fh: $!"; sysseek($tmp_fh, 0, 0) or die "seek failed: $!"; - my ($r, $w); - pipe($r, $w) or die "failed to create pipe: $!"; - my $rdr = { 0 => fileno($tmp_fh), 1 => fileno($w) }; my $git_dir = $self->{-inbox}->git->{git_dir}; my $cmd = ['git', "--git-dir=$git_dir", qw(hash-object --stdin)]; - my $pid = spawn($cmd, undef, $rdr); - close $w; + my $r = popen_rd($cmd, undef, { 0 => $tmp_fh }); local $/ = "\n"; chomp(my $oid = <$r>); - waitpid($pid, 0) == $pid or die "git hash-object did not finish"; - die "git hash-object failed: $?" if $?; + close $r or die "git hash-object failed: $?"; $oid =~ /\A[a-f0-9]{40}\z/ or die "OID not expected: $oid"; $oid; } @@ -665,23 +658,43 @@ sub fill_alternates ($$) { unless (-d $all) { PublicInbox::Import::init_bare($all); } - my $alt = "$all/objects/info/alternates"; - my %alts; - my @add; + my $info_dir = "$all/objects/info"; + my $alt = "$info_dir/alternates"; + my (%alt, $new); + my $mode = 0644; if (-e $alt) { open(my $fh, '<', $alt) or die "open < $alt: $!\n"; - %alts = map { chomp; $_ => 1 } (<$fh>); + $mode = (stat($fh))[2] & 07777; + + # we assign a sort score to every alternate and favor + # the newest (highest numbered) one when we + my $score; + my $other = 0; # in case admin adds non-epoch repos + %alt = map {; + if (m!\A\Q../../\E([0-9]+)\.git/objects\z!) { + $score = $1 + 0; + } else { + $score = --$other; + } + $_ => $score; + } split(/\n+/, do { local $/; <$fh> }); } + foreach my $i (0..$epoch) { my $dir = "../../git/$i.git/objects"; - push @add, $dir if !$alts{$dir} && -d "$pfx/$i.git"; - } - return unless @add; - open my $fh, '>>', $alt or die "open >> $alt: $!\n"; - foreach my $dir (@add) { - print $fh "$dir\n" or die "print >> $alt: $!\n"; + if (!exists($alt{$dir}) && -d "$pfx/$i.git") { + $alt{$dir} = $i; + $new = 1; + } } - close $fh or die "close $alt: $!\n"; + return unless $new; + + my ($fh, $tmp) = tempfile('alt-XXXXXXXX', DIR => $info_dir); + print $fh join("\n", sort { $alt{$b} <=> $alt{$a} } keys %alt), "\n" + or die "print $tmp: $!\n"; + chmod($mode, $fh) or die "fchmod $tmp: $!\n"; + close $fh or die "close $tmp $!\n"; + rename($tmp, $alt) or die "rename $tmp => $alt: $!\n"; } sub git_init { @@ -763,7 +776,6 @@ sub import_init { # XXX experimental sub diff ($$$) { my ($mid, $cur, $new) = @_; - use File::Temp qw(tempfile); my ($ah, $an) = tempfile('email-cur-XXXXXXXX', TMPDIR => 1); print $ah $cur->as_string or die "print: $!"; @@ -775,7 +787,6 @@ sub diff ($$$) { my $cmd = [ qw(diff -u), $an, $bn ]; print STDERR "# MID conflict <$mid>\n"; my $pid = spawn($cmd, undef, { 1 => 2 }); - defined $pid or die "diff failed to spawn $!"; waitpid($pid, 0) == $pid or die "diff did not finish"; unlink($an, $bn); } @@ -859,8 +870,9 @@ sub reindex_checkpoint ($$$) { } # only for a few odd messages with multiple Message-IDs -sub reindex_oid_m ($$$$) { - my ($self, $sync, $git, $oid) = @_; +sub reindex_oid_m ($$$$;$) { + my ($self, $sync, $git, $oid, $regen_num) = @_; + $self->{current_info} = "multi_mid $oid"; my ($num, $mid0, $len); my $msgref = $git->cat_file($oid, \$len); my $mime = PublicInbox::MIME->new($$msgref); @@ -872,49 +884,51 @@ sub reindex_oid_m ($$$$) { delete($sync->{D}->{"$mid\0$cid"}) and die "BUG: reindex_oid should handle <$mid> delete"; } + my $over = $self->{over}; for my $mid (reverse @$mids) { - ($num, $mid0) = $self->{over}->num_mid0_for_oid($oid, $mid); - last if defined $num; + ($num, $mid0) = $over->num_mid0_for_oid($oid, $mid); + next unless defined $num; + if (defined($regen_num) && $regen_num != $num) { + die "BUG: regen(#$regen_num) != over(#$num)"; + } } unless (defined($num)) { for my $mid (reverse @$mids) { # is this a number we got before? - $num = $sync->{mm_tmp}->num_for($mid); - next unless defined $num; - $mid0 = $mid; + my $n = $sync->{mm_tmp}->num_for($mid); + next unless defined $n; + next if defined($regen_num) && $regen_num != $n; + ($num, $mid0) = ($n, $mid); last; } } if (defined($num)) { $sync->{mm_tmp}->num_delete($num); - } else { - $num = $sync->{regen}--; - if ($num <= 0) { - # fixup bugs in old mirrors on reindex - for my $mid (reverse @$mids) { - $num = $self->{mm}->mid_insert($mid); - next unless defined $num; - $mid0 = $mid; - last; - } - if (defined $mid0) { - if ($sync->{reindex}) { - warn "reindex added #$num <$mid0>\n"; - } - } else { - warn "E: cannot find article #\n"; - return; - } - } else { # $num > 0, use the new article number - for my $mid (reverse @$mids) { - $self->{mm}->mid_set($num, $mid) == 1 or next; - $mid0 = $mid; - last; - } - unless (defined $mid0) { - warn "E: cannot regen #$num\n"; - return; + } elsif (defined $regen_num) { + $num = $regen_num; + for my $mid (reverse @$mids) { + $self->{mm}->mid_set($num, $mid) == 1 or next; + $mid0 = $mid; + last; + } + unless (defined $mid0) { + warn "E: cannot regen #$num\n"; + return; + } + } else { # fixup bugs in old mirrors on reindex + for my $mid (reverse @$mids) { + $num = $self->{mm}->mid_insert($mid); + next unless defined $num; + $mid0 = $mid; + last; + } + if (defined $mid0) { + if ($sync->{reindex}) { + warn "reindex added #$num <$mid0>\n"; } + } else { + warn "E: cannot find article #\n"; + return; } } $sync->{nr}++; @@ -935,6 +949,30 @@ sub check_unindexed ($$$) { } } +# reuse Msgmap to store num => oid mapping (rather than num => mid) +sub multi_mid_q_new () { + my ($fh, $fn) = tempfile('multi_mid-XXXXXXX', EXLOCK => 0, TMPDIR => 1); + my $multi_mid = PublicInbox::Msgmap->new_file($fn, 1); + $multi_mid->{dbh}->do('PRAGMA synchronous = OFF'); + # for Msgmap->DESTROY: + $multi_mid->{tmp_name} = $fn; + $multi_mid->{pid} = $$; + close $fh or die "failed to close $fn: $!"; + $multi_mid +} + +sub multi_mid_q_push ($$) { + my ($sync, $oid) = @_; + my $multi_mid = $sync->{multi_mid} //= multi_mid_q_new(); + if ($sync->{reindex}) { # no regen on reindex + $multi_mid->mid_insert($oid); + } else { + my $num = $sync->{regen}--; + die "BUG: ran out of article numbers" if $num <= 0; + $multi_mid->mid_set($num, $oid); + } +} + sub reindex_oid ($$$$) { my ($self, $sync, $git, $oid) = @_; my ($num, $mid0, $len); @@ -966,7 +1004,7 @@ sub reindex_oid ($$$$) { check_unindexed($self, $num, $mid0); } else { $num = $sync->{regen}--; - die "BUG: ran out of article numbers\n" if $num <= 0; + die "BUG: ran out of article numbers" if $num <= 0; if ($self->{mm}->mid_set($num, $mid) != 1) { warn "E: unable to assign $num => <$mid>\n"; return; @@ -983,7 +1021,7 @@ sub reindex_oid ($$$$) { # do not delete from {mm_tmp}, since another # single-MID message may use it. } else { # handle them at the end: - push @{$sync->{multi_mid} //= []}, $oid; + multi_mid_q_push($sync, $oid); } return; } @@ -1102,6 +1140,7 @@ sub sync_prepare ($$$) { --no-notes --no-color --no-renames --diff-filter=AM), $range, '--', 'm'); ++$n while <$fh>; + close $fh or die "git log failed: \$?=$?"; $pr->("$n\n") if $pr; $regen_max += $n; } @@ -1169,7 +1208,7 @@ sub unindex ($$$$) { unindex_oid($self, $git, $1, $unindexed); } delete $self->{reindex_pipe}; - $fh = undef; + close $fh or die "git log failed: \$?=$?"; return unless $sync->{-opt}->{prune}; my $after = scalar keys %$unindexed; @@ -1177,7 +1216,7 @@ sub unindex ($$$$) { # ensure any blob can not longer be accessed via dumb HTTP PublicInbox::Import::run_die(['git', "--git-dir=$git->{git_dir}", - qw(-c gc.reflogExpire=now gc --prune=all)]); + qw(-c gc.reflogExpire=now gc --prune=all --quiet)]); } sub sync_ranges ($$$) { @@ -1225,7 +1264,7 @@ sub index_epoch ($$$) { mark_deleted($self, $sync, $git, $1); } } - $fh = undef; + close $fh or die "git log failed: \$?=$?"; delete $self->{reindex_pipe}; update_last_commit($self, $git, $i, $cmt) if defined $cmt; } @@ -1275,10 +1314,21 @@ sub index_sync { } if (my $multi_mid = delete $sync->{multi_mid}) { $git //= $self->{-inbox}->git; - - while (defined(my $oid = pop(@$multi_mid))) { - $self->{current_info} = "multi_mid $oid"; - reindex_oid_m($self, $sync, $git, $oid); + my ($min, $max) = $multi_mid->minmax; + if ($sync->{reindex}) { + # we may need to create new Message-IDs if mirrors + # were initially indexed with old versions + for (my $i = $max; $i >= $min; $i--) { + my $oid = $multi_mid->mid_for($i); + next unless defined $oid; + reindex_oid_m($self, $sync, $git, $oid); + } + } else { # regen on initial index + for my $num ($min..$max) { + my $oid = $multi_mid->mid_for($num); + next unless defined $oid; + reindex_oid_m($self, $sync, $git, $oid, $num); + } } } $git->cleanup if $git;