X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FLeiToMail.pm;h=da3a95d20d2e1869124f77d1e4b0f14b17d28ec6;hb=40f3f2a2c805fc37c7ed35a60948856bd962b493;hp=76a11b0efcf5d1009eda9e883b3acd51a09f777a;hpb=52a1d7a7e32340bfb9503c33740990bc5ed3d3aa;p=public-inbox.git diff --git a/lib/PublicInbox/LeiToMail.pm b/lib/PublicInbox/LeiToMail.pm index 76a11b0e..da3a95d2 100644 --- a/lib/PublicInbox/LeiToMail.pm +++ b/lib/PublicInbox/LeiToMail.pm @@ -9,12 +9,10 @@ use parent qw(PublicInbox::IPC); use PublicInbox::Eml; use PublicInbox::ProcessPipe; use PublicInbox::Spawn qw(spawn); -use PublicInbox::LeiDedupe; use PublicInbox::PktOp qw(pkt_do); use Symbol qw(gensym); use IO::Handle; # ->autoflush use Fcntl qw(SEEK_SET SEEK_END O_CREAT O_EXCL O_WRONLY); -use Digest::SHA qw(sha256_hex); my %kw2char = ( # Maildir characters draft => 'D', @@ -40,7 +38,7 @@ sub _mbox_hdr_buf ($$$) { if (my $ent = $kw2status{$k}) { push @{$hdr{$ent->[0]}}, $ent->[1]; } else { # X-Label? - warn "TODO: keyword `$k' not supported for mbox\n"; + warn "# keyword `$k' not supported for mbox\n"; } } # Messages are always 'O' (non-\Recent in IMAP), it saves @@ -139,18 +137,17 @@ sub eml2mboxcl2 { sub git_to_mail { # git->cat_async callback my ($bref, $oid, $type, $size, $arg) = @_; - if ($type ne 'blob') { - if ($type eq 'missing') { - warn "missing $oid\n"; - } else { - warn "unexpected type=$type for $oid\n"; - } - } my ($write_cb, $smsg) = @$arg; - if ($smsg->{blob} ne $oid) { - die "BUG: expected=$smsg->{blob} got=$oid"; + if ($type eq 'missing' && $smsg->{-lms_ro}) { + if ($bref = $smsg->{-lms_ro}->local_blob($oid, 1)) { + $type = 'blob'; + $size = length($$bref); + } } - $write_cb->($bref, $smsg) if $size > 0; + return warn("W: $oid is $type (!= blob)\n") if $type ne 'blob'; + return warn("E: $oid is empty\n") unless $size; + die "BUG: expected=$smsg->{blob} got=$oid" if $smsg->{blob} ne $oid; + $write_cb->($bref, $smsg); } sub reap_compress { # dwaitpid callback @@ -160,7 +157,7 @@ sub reap_compress { # dwaitpid callback $lei->fail("@$cmd failed", $? >> 8); } -sub _post_augment_mbox { # open a compressor process +sub _post_augment_mbox { # open a compressor process from top-level process my ($self, $lei) = @_; my $zsfx = $self->{zsfx} or return; my $cmd = PublicInbox::MboxReader::zsfx2cmd($zsfx, undef, $lei); @@ -174,12 +171,6 @@ sub _post_augment_mbox { # open a compressor process $lei->{1} = $pp; } -sub dup_src ($) { - my ($in) = @_; - open my $dup, '+>>&', $in or die "dup: $!"; - $dup; -} - # --augment existing output destination, with deduplication sub _augment { # MboxReader eml_cb my ($eml, $lei) = @_; @@ -208,12 +199,16 @@ sub _mbox_write_cb ($$) { sub { # for git_to_mail my ($buf, $smsg, $eml) = @_; $eml //= PublicInbox::Eml->new($buf); - return if $dedupe->is_dup($eml, $smsg->{blob}); + return if $dedupe->is_dup($eml, $smsg); $lse->xsmsg_vmd($smsg) if $lse; $buf = $eml2mbox->($eml, $smsg); - return atomic_append($lei, $buf) if $atomic_append; - my $lk = $ovv->lock_for_scope; - $lei->out($$buf); + if ($atomic_append) { + atomic_append($lei, $buf); + } else { + my $lk = $ovv->lock_for_scope; + $lei->out($$buf); + } + ++$lei->{-nr_write}; } } @@ -234,17 +229,9 @@ sub update_kw_maybe ($$$$) { } } -sub _augment_or_unlink { # maildir_each_eml cb - my ($f, $kw, $eml, $lei, $lse, $mod, $shard, $unlink) = @_; - if ($mod) { - # can't get dirent.d_ino w/ pure Perl readdir, so we extract - # the OID if it looks like one instead of doing stat(2) - my $hex = $f =~ m!\b([a-f0-9]{40,})[^/]*\z! ? - $1 : sha256_hex($f); - my $recno = hex(substr($hex, 0, 8)); - return if ($recno % $mod) != $shard; - update_kw_maybe($lei, $lse, $eml, $kw); - } +sub _md_update { # maildir_each_eml cb + my ($f, $kw, $eml, $lei, $lse, $unlink) = @_; + update_kw_maybe($lei, $lse, $eml, $kw); $unlink ? unlink($f) : _augment($eml, $lei); } @@ -261,7 +248,7 @@ sub _buf2maildir { my $kw = $smsg->{kw} // []; my $sfx = join('', sort(map { $kw2char{$_} // () } @$kw)); my $rand = ''; # chosen by die roll :P - my ($tmp, $fh, $final, $ok); + my ($tmp, $fh, $base, $ok); my $common = $smsg->{blob} // _rand; if (defined(my $pct = $smsg->{pct})) { $common .= "=$pct" } do { @@ -276,11 +263,12 @@ sub _buf2maildir { $dst .= 'cur/'; $rand = ''; do { - $final = $dst.$rand.$common.':2,'.$sfx; - } while (!($ok = link($tmp, $final)) && $!{EEXIST} && + $base = $rand.$common.':2,'.$sfx + } while (!($ok = link($tmp, $dst.$base)) && $!{EEXIST} && ($rand = _rand.',')); - die "link($tmp, $final): $!" unless $ok; + die "link($tmp, $dst$base): $!" unless $ok; unlink($tmp) or warn "W: failed to unlink $tmp: $!\n"; + \$base; } else { my $err = "Error writing $smsg->{blob} to $dst: $!\n"; $_[0] = undef; # clobber dst @@ -295,16 +283,18 @@ sub _maildir_write_cb ($$) { $dedupe->prepare_dedupe if $dedupe; my $dst = $lei->{ovv}->{dst}; my $lse = $lei->{lse}; # may be undef + my $sto = $lei->{opt}->{'mail-sync'} ? $lei->{sto} : undef; + my $out = $sto ? 'maildir:'.$lei->rel2abs($dst) : undef; sub { # for git_to_mail - my ($buf, $smsg, $eml) = @_; + my ($bref, $smsg, $eml) = @_; $dst // return $lei->fail; # dst may be undef-ed in last run - $buf //= \($eml->as_string); + return if $dedupe && $dedupe->is_dup($eml // + PublicInbox::Eml->new($$bref), + $smsg); $lse->xsmsg_vmd($smsg) if $lse; - return _buf2maildir($dst, $buf, $smsg) if !$dedupe; - $eml //= PublicInbox::Eml->new($$buf); # copy buf - return if $dedupe->is_dup($eml, $smsg->{blob}); - undef $eml; - _buf2maildir($dst, $buf, $smsg); + my $n = _buf2maildir($dst, $bref // \($eml->as_string), $smsg); + $sto->ipc_do('set_sync_info', $smsg->{blob}, $out, $n) if $sto; + ++$lei->{-nr_write}; } } @@ -312,23 +302,49 @@ sub _imap_write_cb ($$) { my ($self, $lei) = @_; my $dedupe = $lei->{dedupe}; $dedupe->prepare_dedupe if $dedupe; - my $imap_append = $lei->{net}->can('imap_append'); + my $append = $lei->{net}->can('imap_append'); my $mic = $lei->{net}->mic_get($self->{uri}); my $folder = $self->{uri}->mailbox; my $lse = $lei->{lse}; # may be undef + my $sto = $lei->{opt}->{'mail-sync'} ? $lei->{sto} : undef; + my $out = $lei->{ovv}->{dst}; sub { # for git_to_mail my ($bref, $smsg, $eml) = @_; - $mic // return $lei->fail; # dst may be undef-ed in last run - if ($dedupe) { - $eml //= PublicInbox::Eml->new($$bref); # copy bref - return if $dedupe->is_dup($eml, $smsg->{blob}); - } + $mic // return $lei->fail; # mic may be undef-ed in last run + return if $dedupe && $dedupe->is_dup($eml // + PublicInbox::Eml->new($$bref), + $smsg); $lse->xsmsg_vmd($smsg) if $lse; - eval { $imap_append->($mic, $folder, $bref, $smsg, $eml) }; + my $uid = eval { $append->($mic, $folder, $bref, $smsg, $eml) }; if (my $err = $@) { undef $mic; die $err; } + # imap_append returns UID if IMAP server has UIDPLUS extension + ($sto && $uid =~ /\A[0-9]+\z/) and + $sto->ipc_do('set_sync_info', + $smsg->{blob}, $out, $uid + 0); + ++$lei->{-nr_write}; + } +} + +sub _text_write_cb ($$) { + my ($self, $lei) = @_; + my $dedupe = $lei->{dedupe}; + $dedupe->prepare_dedupe if $dedupe; + my $lvt = $lei->{lvt}; + my $ovv = $lei->{ovv}; + $lei->{1} // die "no stdout ($ovv->{dst})"; # redirected earlier + $lei->{1}->autoflush(1); + binmode $lei->{1}, ':utf8'; + my $lse = $lei->{lse}; # may be undef + sub { # for git_to_mail + my ($bref, $smsg, $eml) = @_; + $lse->xsmsg_vmd($smsg) if $lse; + $eml //= PublicInbox::Eml->new($bref); # copy bref + return if $dedupe && $dedupe->is_dup($eml, $smsg); + my $lk = $ovv->lock_for_scope; + $lei->out(${$lvt->eml_to_text($smsg, $eml)}, "\n"); } } @@ -352,27 +368,47 @@ sub new { $lei->{ovv}->{dst} = $dst .= '/' if substr($dst, -1) ne '/'; } elsif (substr($fmt, 0, 4) eq 'mbox') { require PublicInbox::MboxReader; - (-d $dst || (-e _ && !-w _)) and die - "$dst exists and is not a writable file\n"; $self->can("eml2$fmt") or die "bad mbox format: $fmt\n"; $self->{base_type} = 'mbox'; - } elsif ($fmt =~ /\Aimaps?\z/) { # TODO .onion support + } elsif ($fmt =~ /\Aimaps?\z/) { require PublicInbox::NetWriter; + require PublicInbox::URIimap; my $net = PublicInbox::NetWriter->new; - $net->add_url($dst); $net->{quiet} = $lei->{opt}->{quiet}; - my $err = $net->errors($dst); + my $uri = PublicInbox::URIimap->new($dst)->canonical; + $net->add_url($uri); + my $err = $net->errors($lei); return $lei->fail($err) if $err; - require PublicInbox::URIimap; # TODO: URI cast early - $self->{uri} = PublicInbox::URIimap->new($dst); - $self->{uri}->mailbox or die "No mailbox: $dst"; + $uri->mailbox or return $lei->fail("No mailbox: $dst"); + $self->{uri} = $uri; + $dst = $lei->{ovv}->{dst} = $$uri; # canonicalized $lei->{net} = $net; $self->{base_type} = 'imap'; + } elsif ($fmt eq 'text') { + require PublicInbox::LeiViewText; + $lei->{lvt} = PublicInbox::LeiViewText->new($lei); + $self->{base_type} = 'text'; } else { die "bad mail --format=$fmt\n"; } + if ($self->{base_type} =~ /\A(?:text|mbox)\z/) { + (-d $dst || (-e _ && !-w _)) and die + "$dst exists and is not a writable file\n"; + } + if ($self->{base_type} eq 'text') { + my @err = map { + defined($lei->{opt}->{$_}) ? "--$_" : (); + } (qw(mua save)); + die "@err incompatible with $fmt\n" if @err; + } $self->{dst} = $dst; - $lei->{dedupe} = PublicInbox::LeiDedupe->new($lei); + $lei->{dedupe} = $lei->{lss} // do { + my $dd_cls = 'PublicInbox::'. + ($lei->{opt}->{save} ? 'LeiSavedSearch' : 'LeiDedupe'); + eval "require $dd_cls"; + die "$dd_cls: $@" if $@; + $dd_cls->new($lei); + }; $self; } @@ -386,27 +422,28 @@ sub _pre_augment_maildir { File::Path::mkpath($d); -d $d or die "$d is not a directory"; } + # for utime, so no opendir + open $self->{poke_dh}, '<', "${dst}cur" or die "open ${dst}cur: $!"; } sub _do_augment_maildir { my ($self, $lei) = @_; + return if ($lei->{opt}->{save} // 0) < 0; my $dst = $lei->{ovv}->{dst}; my $lse = $lei->{opt}->{'import-before'} ? $lei->{lse} : undef; - my ($mod, $shard) = @{$self->{shard_info} // []}; + my $mdr = PublicInbox::MdirReader->new; if ($lei->{opt}->{augment}) { my $dedupe = $lei->{dedupe}; if ($dedupe && $dedupe->prepare_dedupe) { - PublicInbox::MdirReader::maildir_each_eml($dst, - \&_augment_or_unlink, - $lei, $lse, $mod, $shard); + $mdr->{shard_info} = $self->{shard_info}; + $mdr->maildir_each_eml($dst, \&_md_update, $lei, $lse); $dedupe->pause_dedupe; } } elsif ($lse) { - PublicInbox::MdirReader::maildir_each_eml($dst, - \&_augment_or_unlink, - $lei, $lse, $mod, $shard, 1); + $mdr->{shard_info} = $self->{shard_info}; + $mdr->maildir_each_eml($dst, \&_md_update, $lei, $lse, 1); } else {# clobber existing Maildir - PublicInbox::MdirReader::maildir_each_file($dst, \&_unlink); + $mdr->maildir_each_file($dst, \&_unlink); } } @@ -422,6 +459,7 @@ sub _imap_augment_or_delete { # PublicInbox::NetReader::imap_each cb sub _do_augment_imap { my ($self, $lei) = @_; + return if ($lei->{opt}->{save} // 0) < 0; my $net = $lei->{net}; my $lse = $lei->{opt}->{'import-before'} ? $lei->{lse} : undef; if ($lei->{opt}->{augment}) { @@ -442,6 +480,29 @@ sub _do_augment_imap { } } +sub _pre_augment_text { + my ($self, $lei) = @_; + my $dst = $lei->{ovv}->{dst}; + my $out; + my $devfd = $lei->path_to_fd($dst) // die "bad $dst"; + if ($devfd >= 0) { + $out = $lei->{$devfd}; + } else { # normal-looking path + if (-p $dst) { + open $out, '>', $dst or die "open($dst): $!"; + } elsif (-f _ || !-e _) { + # text allows augment, HTML/Atom won't + my $mode = $lei->{opt}->{augment} ? '>>' : '>'; + open $out, $mode, $dst or die "open($mode, $dst): $!"; + } else { + die "$dst is not a file or FIFO\n"; + } + } + $lei->{ovv}->ovv_out_lk_init if !$lei->{ovv}->{lock_path}; + $lei->{1} = $out; + undef; +} + sub _pre_augment_mbox { my ($self, $lei) = @_; my $dst = $lei->{ovv}->{dst}; @@ -492,6 +553,7 @@ sub _do_augment_mbox { my ($self, $lei) = @_; return unless $self->{seekable}; my $opt = $lei->{opt}; + return if ($opt->{save} // 0) < 0; my $out = $lei->{1}; my ($fmt, $dst) = @{$lei->{ovv}}{qw(fmt dst)}; return unless -s $out; @@ -499,9 +561,12 @@ sub _do_augment_mbox { truncate($out, 0) or die "truncate($dst): $!"; return; } - my $zsfx = $self->{zsfx}; - my $rd = $zsfx ? PublicInbox::MboxReader::zsfxcat($out, $zsfx, $lei) - : dup_src($out); + my $rd; + if (my $zsfx = $self->{zsfx}) { + $rd = PublicInbox::MboxReader::zsfxcat($out, $zsfx, $lei); + } else { + open($rd, '+>>&', $out) or die "dup: $!"; + } my $dedupe; if ($opt->{augment}) { $dedupe = $lei->{dedupe}; @@ -532,8 +597,8 @@ sub pre_augment { # fast (1 disk seek), runs in same process as post_augment sub do_augment { # slow, runs in wq worker my ($self, $lei) = @_; # _do_augment_maildir, _do_augment_mbox, or _do_augment_imap - my $m = "_do_augment_$self->{base_type}"; - $self->$m($lei); + my $m = $self->can("_do_augment_$self->{base_type}") or return; + $m->($self, $lei); } # fast (spawn compressor or mkdir), runs in same process as pre_augment @@ -546,6 +611,7 @@ sub post_augment { $m->($self, $lei, @args); } +# called by every single l2m worker process sub do_post_auth { my ($self) = @_; my $lei = $self->{lei}; @@ -553,7 +619,7 @@ sub do_post_auth { pkt_do($lei->{pkt_op_p}, 'incr_start_query') or die "incr_start_query: $!"; my $aug; - if (lock_free($self)) { + if (lock_free($self)) { # all workers do_augment my $mod = $self->{-wq_nr_workers}; my $shard = $self->{-wq_worker_nr}; if (my $net = $lei->{net}) { @@ -562,7 +628,7 @@ sub do_post_auth { $self->{shard_info} = [ $mod, $shard ]; } $aug = '+'; # incr_post_augment - } elsif ($self->{-wq_worker_nr} == 0) { + } elsif ($self->{-wq_worker_nr} == 0) { # 1st worker do_augment $aug = '.'; # do_post_augment } if ($aug) { @@ -572,6 +638,7 @@ sub do_post_auth { pkt_do($lei->{pkt_op_p}, $aug) == 1 or die "do_post_augment trigger: $!"; } + # done augmenting, connect the compressor pipe for each worker if (my $zpipe = delete $lei->{zpipe}) { $lei->{1} = $zpipe->[1]; close $zpipe->[0]; @@ -583,6 +650,7 @@ sub ipc_atfork_child { my ($self) = @_; my $lei = $self->{lei}; $lei->_lei_atfork_child; + $self->{-lms_ro} = $lei->{lse}->lms if $lei->{lse}; $lei->{auth}->do_auth_atfork($self) if $lei->{auth}; $SIG{__WARN__} = PublicInbox::Eml::warn_ignore_cb(); $self->SUPER::ipc_atfork_child; @@ -592,16 +660,19 @@ sub lock_free { $_[0]->{base_type} =~ /\A(?:maildir|imap|jmap)\z/ ? 1 : 0; } +# wakes up the MUA when complete so it can refresh messages list sub poke_dst { my ($self) = @_; if ($self->{base_type} eq 'maildir') { my $t = time + 1; - utime($t, $t, $self->{dst} . 'cur'); + utime($t, $t, $self->{poke_dh}) or warn "futimes: $!"; } } sub write_mail { # via ->wq_io_do - my ($self, $smsg) = @_; + my ($self, $smsg, $eml) = @_; + return $self->{wcb}->(undef, $smsg, $eml) if $eml; + $smsg->{-lms_ro} = $self->{-lms_ro}; $self->{lei}->{ale}->git->cat_async($smsg->{blob}, \&git_to_mail, [$self->{wcb}, $smsg]); } @@ -609,12 +680,16 @@ sub write_mail { # via ->wq_io_do sub wq_atexit_child { my ($self) = @_; delete $self->{wcb}; - $self->{lei}->{ale}->git->async_wait_all; - $SIG{__WARN__} = 'DEFAULT'; + my $lei = $self->{lei}; + $lei->{ale}->git->async_wait_all; + my $nr = delete($lei->{-nr_write}) or return; + return if $lei->{early_mua} || !$lei->{-progress} || !$lei->{pkt_op_p}; + require PublicInbox::PktOp; + PublicInbox::PktOp::pkt_do($lei->{pkt_op_p}, 'l2m_progress', $nr); } # called in top-level lei-daemon when LeiAuth is done -sub net_merge_complete { +sub net_merge_all_done { my ($self) = @_; $self->wq_broadcast('do_post_auth'); $self->wq_close(1);