X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FLeiImport.pm;h=daaa6753f218a86bbc915f6bb718f2ae3108febf;hb=b6b86cfd238c170ea3e2c4d4179f06c7af139086;hp=e769fba828c9a85241a6956d6fffe3856dfd0b78;hpb=a00fdf39b4ec4ed045f28c29c79c8dcb9abac836;p=public-inbox.git diff --git a/lib/PublicInbox/LeiImport.pm b/lib/PublicInbox/LeiImport.pm index e769fba8..daaa6753 100644 --- a/lib/PublicInbox/LeiImport.pm +++ b/lib/PublicInbox/LeiImport.pm @@ -6,50 +6,86 @@ package PublicInbox::LeiImport; use strict; use v5.10.1; use parent qw(PublicInbox::IPC PublicInbox::LeiInput); -use PublicInbox::Eml; -use PublicInbox::PktOp qw(pkt_do); -sub _import_eml { # MboxReader callback - my ($eml, $lei, $mbox_keywords) = @_; +# /^input_/ subs are used by (or override) PublicInbox::LeiInput superclass + +sub input_eml_cb { # used by PublicInbox::LeiInput::input_fh + my ($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}->ipc_do('set_eml', $eml, $vmd, $xoids); +} + +sub input_mbox_cb { # MboxReader callback + my ($eml, $self) = @_; my $vmd; - if ($mbox_keywords) { - my $kw = $mbox_keywords->($eml); + if ($self->{-import_kw}) { + my $kw = PublicInbox::MboxReader::mbox_keywords($eml); $vmd = { kw => $kw } if scalar(@$kw); } - my $xoids = $lei->{ale}->xoids_for($eml); - $lei->{sto}->ipc_do('set_eml', $eml, $vmd, $xoids); + input_eml_cb($self, $eml, $vmd); } -sub import_done_wait { # dwaitpid callback - my ($arg, $pid) = @_; - my ($imp, $lei) = @$arg; - $lei->child_error($?, 'non-fatal errors during import') if $?; - my $sto = delete $lei->{sto}; - my $wait = $sto->ipc_do('done') if $sto; # PublicInbox::LeiStore::done - $lei->dclose; +sub input_maildir_cb { # maildir_each_eml cb + my ($f, $kw, $eml, $self) = @_; + my $vmd = $self->{-import_kw} ? { kw => $kw } : undef; + if ($self->{-mail_sync}) { + if ($f =~ m!\A(.+?)/(?:new|cur)/([^/]+)\z!) { # ugh... + $vmd->{sync_info} = [ "maildir:$1", \(my $n = $2) ]; + } else { + warn "E: $f was not from a Maildir?\n"; + } + } + input_eml_cb($self, $eml, $vmd); +} + +sub input_imap_cb { # imap_each + my ($url, $uid, $kw, $eml, $self) = @_; + my $vmd = $self->{-import_kw} ? { kw => $kw } : undef; + $vmd->{sync_info} = [ $url, $uid ] if $self->{-mail_sync}; + input_eml_cb($self, $eml, $vmd); +} + +sub input_nntp_cb { # nntp_each + my ($url, $num, $kw, $eml, $self) = @_; + input_eml_cb($self, $eml, $self->{-import_kw} ? { kw => $kw } : undef); } sub import_done { # EOF callback for main daemon my ($lei) = @_; - my $imp = delete $lei->{imp} or return; - $imp->wq_wait_old(\&import_done_wait, $lei); + my $imp = delete $lei->{imp} // return $lei->fail('BUG: {imp} gone'); + $imp->wq_wait_old($lei->can('wq_done_wait'), $lei, 'non-fatal'); } sub net_merge_complete { # callback used by LeiAuth my ($self) = @_; - for my $input (@{$self->{inputs}}) { - $self->wq_io_do('import_path_url', [], $input); - } + $self->wq_io_do('process_inputs'); $self->wq_close(1); } -sub import_start { - my ($lei) = @_; - my $self = $lei->{imp}; - $lei->ale; +sub lei_import { # the main "lei import" method + my ($lei, @inputs) = @_; + my $sto = $lei->_lei_store(1); + $sto->write_prepare($lei); + my $self = bless {}, __PACKAGE__; + $self->{-import_kw} = $lei->{opt}->{kw} // 1; + my $vmd_mod = $self->vmd_mod_extract(\@inputs); + return $lei->fail(join("\n", @{$vmd_mod->{err}})) if $vmd_mod->{err}; + $self->{all_vmd} = $vmd_mod if scalar keys %$vmd_mod; + $self->prepare_inputs($lei, \@inputs) or return; + $self->{-mail_sync} = $lei->{opt}->{sync} // 1; + + $lei->ale; # initialize for workers to read my $j = $lei->{opt}->{jobs} // scalar(@{$self->{inputs}}) || 1; if (my $net = $lei->{net}) { # $j = $net->net_concurrency($j); TODO + if ($lei->{opt}->{incremental} // 1) { + $net->{incremental} = 1; + $net->{itrk_fn} = $lei->store_path . + '/net_last.sqlite3'; + } } else { my $nproc = $self->detect_nproc; $j = $nproc if $j > $nproc; @@ -57,102 +93,15 @@ sub import_start { 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}; + (my $op_c, $ops) = $lei->workers_start($self, 'lei_import', $j, $ops); + $lei->{imp} = $self; net_merge_complete($self) unless $lei->{auth}; - while ($op && $op->{sock}) { $op->event_step } -} - -sub lei_import { # the main "lei import" method - my ($lei, @inputs) = @_; - my $sto = $lei->_lei_store(1); - $sto->write_prepare($lei); - $lei->{opt}->{kw} //= 1; - my $self = $lei->{imp} = bless {}, __PACKAGE__; - $self->prepare_inputs($lei, \@inputs) or return; - import_start($lei); -} - -sub ipc_atfork_child { - my ($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 $kw = $lei->{opt}->{kw} ? - PublicInbox::MboxReader->can('mbox_keywords') : undef; - eval { - if ($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, $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, $kw); - } - }; - $lei->child_error(1 << 8, "$input: $@") if $@; + $op_c->op_wait_event($ops); } -sub _import_maildir { # maildir_each_eml cb - my ($f, $kw, $eml, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml', $eml, $set_kw ? { kw => $kw }: ()); -} - -sub _import_net { # imap_each, nntp_each cb - my ($url, $uid, $kw, $eml, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml', $eml, $set_kw ? { kw => $kw } : ()); -} - -sub import_path_url { - my ($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; - } - if (-f $input) { - my $m = $lei->{opt}->{'lock'} // ($ifmt eq 'eml' ? ['none'] : - PublicInbox::MboxLock->defaults); - my $mbl = PublicInbox::MboxLock->acq($input, 0, $m); - _import_fh($lei, $mbl->{fh}, $input, $ifmt); - } elsif (-d _ && (-d "$input/cur" || -d "$input/new")) { - return $lei->fail(<{sto}, $lei->{opt}->{kw}); - } else { - $lei->fail("$input unsupported (TODO)"); - } -} - -sub import_stdin { - my ($self) = @_; - my $lei = $self->{lei}; - my $in = delete $self->{0}; - _import_fh($lei, $in, '', $lei->{opt}->{'in-format'}); -} +no warnings 'once'; +*ipc_atfork_child = \&PublicInbox::LeiInput::input_only_atfork_child; -no warnings 'once'; # the following works even when LeiAuth is lazy-loaded +# the following works even when LeiAuth is lazy-loaded *net_merge_all = \&PublicInbox::LeiAuth::net_merge_all; 1;