diff options
Diffstat (limited to 'lib/PublicInbox/LeiImport.pm')
| -rw-r--r-- | lib/PublicInbox/LeiImport.pm | 293 |
1 files changed, 129 insertions, 164 deletions
diff --git a/lib/PublicInbox/LeiImport.pm b/lib/PublicInbox/LeiImport.pm index 7f247b64..5521188c 100644 --- a/lib/PublicInbox/LeiImport.pm +++ b/lib/PublicInbox/LeiImport.pm @@ -1,198 +1,163 @@ -# Copyright (C) 2021 all contributors <meta@public-inbox.org> +# Copyright (C) all contributors <meta@public-inbox.org> # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt> # front-end for the "lei import" sub-command package PublicInbox::LeiImport; use strict; use v5.10.1; -use parent qw(PublicInbox::IPC); -use PublicInbox::Eml; -use PublicInbox::PktOp qw(pkt_do); +use parent qw(PublicInbox::IPC PublicInbox::LeiInput); +use PublicInbox::InboxWritable qw(eml_from_path); +use PublicInbox::Compat qw(uniqstr); -sub _import_eml { # MboxReader callback - my ($eml, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml', $eml, $set_kw ? $sto->mbox_keywords($eml) : ()); +# /^input_/ subs are used by (or override) PublicInbox::LeiInput superclass + +sub input_eml_cb { # used by PublicInbox::LeiInput::input_fh + my ($self, $eml, $vmd) = @_; + my $xoids = $self->{lei}->{ale}->xoids_for($eml); + if (my $all_vmd = $self->{all_vmd}) { + @$vmd{keys %$all_vmd} = values %$all_vmd; + } + $self->{lei}->{sto}->wq_do('set_eml', $eml, $vmd, $xoids); } -sub import_done_wait { # dwaitpid callback - my ($arg, $pid) = @_; - my ($imp, $lei) = @$arg; - $lei->child_error($?, 'non-fatal errors during import') if $?; - my $ign = $lei->{sto}->ipc_do('done'); # PublicInbox::LeiStore::done - $lei->dclose; +sub input_mbox_cb { # MboxReader callback + my ($eml, $self) = @_; + my $vmd; + if ($self->{-import_kw}) { + my $kw = PublicInbox::MboxReader::mbox_keywords($eml); + $vmd = { kw => $kw } if scalar(@$kw); + } + input_eml_cb($self, $eml, $vmd); } -sub import_done { # EOF callback for main daemon - my ($lei) = @_; - my $imp = delete $lei->{imp} or return; - $imp->wq_wait_old(\&import_done_wait, $lei); +sub pmdir_cb { # called via wq_io_do from LeiPmdir->each_mdir_fn + my ($self, $f, $fl) = @_; + my ($folder, $bn) = ($f =~ m!\A(.+?)/(?:new|cur)/([^/]+)\z!) or + die "BUG: $f was not from a Maildir?\n"; + my $kw = PublicInbox::MdirReader::flags2kw($fl); + substr($folder, 0, 0) = 'maildir:'; # add prefix + my $lse = $self->{lse} //= $self->{lei}->{sto}->search; + my $lms = $self->{-lms_rw} //= $self->{lei}->lms; # may be 0 or undef + my @oidbin = $lms ? $lms->name_oidbin($folder, $bn) : (); + @oidbin > 1 and warn("W: $folder/*/$$bn not unique:\n", + map { "\t".unpack('H*', $_)."\n" } @oidbin); + my @docids = sort { $a <=> $b } uniqstr + map { $lse->over->oidbin_exists($_) } @oidbin; + my $vmd = $self->{-import_kw} ? { kw => $kw } : undef; + if (scalar @docids) { + $lse->kw_changed(undef, $kw, \@docids) or return; + } + if (my $eml = eml_from_path($f)) { + $vmd->{sync_info} = [ $folder, \$bn ] if $self->{-mail_sync}; + $self->input_eml_cb($eml, $vmd); + } } -sub net_merge_complete { # callback used by LeiAuth - my ($self) = @_; - for my $input (@{$self->{inputs}}) { - $self->wq_io_do('import_path_url', [], $input); +sub input_mh_cb { + my ($mhdir, $n, $kw, $eml, $self) = @_; + substr($mhdir, 0, 0) = 'mh:'; # add prefix + my $lse = $self->{lse} //= $self->{lei}->{sto}->search; + my $lms = $self->{-lms_rw} //= $self->{lei}->lms; # may be 0 or undef + my @oidbin = $lms ? $lms->num_oidbin($mhdir, $n) : (); + @oidbin > 1 and warn("W: $mhdir/$n not unique:\n", + map { "\t".unpack('H*', $_)."\n" } @oidbin); + my @docids = sort { $a <=> $b } uniqstr + map { $lse->over->oidbin_exists($_) } @oidbin; + if (scalar @docids) { + $lse->kw_changed(undef, $kw, \@docids) or return; } - $self->wq_close(1); + if (defined $eml) { + my $vmd = $self->{-import_kw} ? { kw => $kw } : undef; + $vmd->{sync_info} = [ $mhdir, $n + 0 ] if $self->{-mail_sync}; + $self->input_eml_cb($eml, $vmd); + } + # TODO: + # elsif (my $ikw = $self->{lei}->{ikw}) { # old message, kw only + # $ikw->wq_io_do('ck_update_kw', [], "mh:$dir", $uid, $kw); } -sub import_start { - my ($lei) = @_; - my $self = $lei->{imp}; - my $j = $lei->{opt}->{jobs} // scalar(@{$self->{inputs}}) || 1; - if (my $net = $lei->{net}) { - # $j = $net->net_concurrency($j); TODO - } else { - my $nproc = $self->detect_nproc; - $j = $nproc if $j > $nproc; +sub input_net_cb { # imap_each / nntp_each + my ($uri, $uid, $kw, $eml, $self) = @_; + if (defined $eml) { + my $vmd = $self->{-import_kw} ? { kw => $kw } : undef; + $vmd->{sync_info} = [ $$uri, $uid ] if $self->{-mail_sync}; + $self->input_eml_cb($eml, $vmd); + } elsif (my $ikw = $self->{lei}->{ikw}) { # old message, kw only + # we send $uri as a bare SCALAR and not a URIimap ref to + # reduce socket traffic: + $ikw->wq_io_do('ck_update_kw', [], $$uri, $uid, $kw); } - my $ops = { '' => [ \&import_done, $lei ] }; - $lei->{auth}->op_merge($ops, $self) if $lei->{auth}; - $self->{-wq_nr_workers} = $j // 1; # locked - my $op = $lei->workers_start($self, 'lei_import', undef, $ops); - $self->wq_io_do('import_stdin', []) if $self->{0}; - net_merge_complete($self) unless $lei->{auth}; - while ($op && $op->{sock}) { $op->event_step } } -sub call { # the main "lei import" method - my ($cls, $lei, @inputs) = @_; +sub do_import_index ($$@) { + my ($self, $lei, @inputs) = @_; my $sto = $lei->_lei_store(1); $sto->write_prepare($lei); - my ($net, @f, @d); - $lei->{opt}->{kw} //= 1; - my $self = $lei->{imp} = bless { inputs => \@inputs }, $cls; - if ($lei->{opt}->{stdin}) { - @inputs and return $lei->fail("--stdin and @inputs do not mix"); - $lei->check_input_format or return; - $self->{0} = $lei->{0}; - } - - my $fmt = $lei->{opt}->{'in-format'}; - # e.g. Maildir:/home/user/Mail/ or imaps://example.com/INBOX - for my $input (@inputs) { - my $input_path = $input; - if ($input =~ m!\A(?:imaps?|nntps?|s?news)://!i) { - require PublicInbox::NetReader; - $net //= PublicInbox::NetReader->new; - $net->add_url($input); - } elsif ($input_path =~ s/\A([a-z0-9]+)://is) { - my $ifmt = lc $1; - if (($fmt // $ifmt) ne $ifmt) { - return $lei->fail(<<""); ---format=$fmt and `$ifmt:' conflict - - } - if (-f $input_path) { - require PublicInbox::MboxReader; - PublicInbox::MboxReader->can($ifmt) or return - $lei->fail("$ifmt not supported"); - } elsif (-d _) { - require PublicInbox::MdirReader; - $ifmt eq 'maildir' or return - $lei->fail("$ifmt not supported"); - } else { - return $lei->fail("Unable to handle $input"); - } - } elsif (-f $input) { push @f, $input - } elsif (-d _) { push @d, $input - } else { return $lei->fail("Unable to handle $input") } - } - if (@f) { $lei->check_input_format(\@f) or return } - if (@d) { # TODO: check for MH vs Maildir, here - require PublicInbox::MdirReader; - } - $self->{inputs} = \@inputs; + $self->{-import_kw} = $lei->{opt}->{kw} // 1; + $self->{all_vmd} = $lei->{vmd_mod} if keys %{$lei->{vmd_mod}}; + $lei->ale; # initialize for workers to read (before LeiPmdir->new) + $self->{-mail_sync} = $lei->{opt}->{'mail-sync'} // 1; + $self->prepare_inputs($lei, \@inputs) or return; + + my $j = $lei->{opt}->{jobs} // 0; + $j =~ /\A([0-9]+),[0-9]+\z/ and $j = $1 + 0; + $j ||= scalar(@{$self->{inputs}}) || 1; + my $ikw; + my $net = $lei->{net}; if ($net) { - if (my $err = $net->errors) { - return $lei->fail($err); + # $j = $net->net_concurrency($j); TODO + if ($lei->{opt}->{incremental} // 1) { + $net->{incremental} = 1; + $net->{-lms_rw} = $lei->lms // 0; + if ($self->{-import_kw} && $net->{-lms_rw} && + !$lei->{opt}->{'new-only'} && + $net->{imap_order}) { + require PublicInbox::LeiImportKw; + $ikw = PublicInbox::LeiImportKw->new($lei); + $net->{each_old} = 1; + } } - $net->{quiet} = $lei->{opt}->{quiet}; - $lei->{net} = $net; - require PublicInbox::LeiAuth; - $lei->{auth} = PublicInbox::LeiAuth->new; + } else { + my $nproc = $self->detect_nproc; + $j = $nproc if $j > $nproc; } - import_start($lei); + ($lei->{opt}->{'new-only'} && (!$net || !$net->{imap_order})) and + warn "# --new-only is only for IMAP\n"; + $lei->{-eml_noisy} = 1; + $lei->{-err_type} = 'non-fatal'; + $lei->wq1_start($self, $j); } -sub ipc_atfork_child { - my ($self) = @_; - my $lei = $self->{lei}; - delete $lei->{imp}; # drop circular ref - $lei->lei_atfork_child; - $self->SUPER::ipc_atfork_child; - $lei->{auth}->do_auth_atfork($self) if $lei->{auth}; - undef; -} - -sub _import_fh { - my ($lei, $fh, $input, $ifmt) = @_; - my $set_kw = $lei->{opt}->{kw}; - eval { - if ($ifmt eq 'eml') { - my $buf = do { local $/; <$fh> } // - return $lei->child_error(1 << 8, <<""); -error reading $input: $! - - my $eml = PublicInbox::Eml->new(\$buf); - _import_eml($eml, $lei->{sto}, $set_kw); - } else { # some mbox (->can already checked in call); - my $cb = PublicInbox::MboxReader->can($ifmt) // - die "BUG: bad fmt=$ifmt"; - $cb->(undef, $fh, \&_import_eml, $lei->{sto}, $set_kw); - } - }; - $lei->child_error(1 << 8, "<stdin>: $@") if $@; +sub lei_import { # the main "lei import" method + my ($lei, @inputs) = @_; + my $self = bless {}, __PACKAGE__; + do_import_index($self, $lei, @inputs); } -sub _import_maildir { # maildir_each_file cb - my ($f, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml_from_maildir', $f, $set_kw); -} - -sub _import_net { # imap_each, nntp_each cb - my ($url, $uid, $kw, $eml, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml', $eml, $set_kw ? @$kw : ()); -} - -sub import_path_url { - my ($self, $input) = @_; - my $lei = $self->{lei}; - my $ifmt = lc($lei->{opt}->{'in-format'} // ''); - # TODO auto-detect? - if ($input =~ m!\Aimaps?://!i) { - $lei->{net}->imap_each($input, \&_import_net, $lei->{sto}, - $lei->{opt}->{kw}); - return; - } elsif ($input =~ m!\A(?:nntps?|s?news)://!i) { - $lei->{net}->nntp_each($input, \&_import_net, $lei->{sto}, 0); - return; - } elsif ($input =~ s!\A([a-z0-9]+):!!i) { - $ifmt = lc $1; +sub _complete_import { + my ($lei, @argv) = @_; + my $has_arg = @argv; + my ($pfx, $cur, $match_cb) = $lei->complete_url_prepare(\@argv); + my @try = $has_arg ? ($pfx.$cur, $argv[-1]) : ($argv[-1]); + push(@try, undef) if defined $try[-1]; + my (@f, @k); + for (@try) { + @k = $lei->url_folder_cache->keys($_, 1) and last; } - if (-f $input) { - open my $fh, '<', $input or return $lei->child_error(1 << 8, <<""); -unable to open $input: $! - - _import_fh($lei, $fh, $input, $ifmt); - } elsif (-d _ && (-d "$input/cur" || -d "$input/new")) { - return $lei->fail(<<EOM) if $ifmt && $ifmt ne 'maildir'; -$input appears to a be a maildir, not $ifmt -EOM - PublicInbox::MdirReader::maildir_each_file($input, - \&_import_maildir, - $lei->{sto}, $lei->{opt}->{kw}); - } else { - $lei->fail("$input unsupported (TODO)"); + my @L = eval { $lei->_lei_store->search->all_terms('L') }; + push(@k, map { "+L:$_" } @L); + if (my $lms = $lei->lms) { + for (@try) { + @f = $lms->folders($_, 1) and last; + } + push @k, @f; } + my @m = map { $match_cb->($_) } @k; + @m ? @m : @k; } -sub import_stdin { - my ($self) = @_; - my $lei = $self->{lei}; - _import_fh($lei, delete $self->{0}, '<stdin>', $lei->{opt}->{'in-format'}); -} +no warnings 'once'; +*ipc_atfork_child = \&PublicInbox::LeiInput::input_only_atfork_child; +*net_merge_all_done = \&PublicInbox::LeiInput::input_only_net_merge_all_done; -no warnings 'once'; # the following works even when LeiAuth is lazy-loaded -*net_merge_all = \&PublicInbox::LeiAuth::net_merge_all; 1; |