X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FLeiImport.pm;h=21af28a3a79c647df2cf0b6dafa07d7e401e7efb;hb=5be0cb101bab44167a78af7a2d167f254c95bdb3;hp=c2c98030074282e0329767069d41dcd340102054;hpb=d540245cfff927692bed39b8cbe7c621452bc4b2;p=public-inbox.git diff --git a/lib/PublicInbox/LeiImport.pm b/lib/PublicInbox/LeiImport.pm index c2c98030..21af28a3 100644 --- a/lib/PublicInbox/LeiImport.pm +++ b/lib/PublicInbox/LeiImport.pm @@ -5,20 +5,32 @@ package PublicInbox::LeiImport; use strict; use v5.10.1; -use parent qw(PublicInbox::IPC); +use parent qw(PublicInbox::IPC PublicInbox::LeiInput); use PublicInbox::Eml; use PublicInbox::PktOp qw(pkt_do); -sub _import_eml { # MboxReader callback - my ($eml, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml', $eml, $set_kw ? $sto->mbox_keywords($eml) : ()); +sub eml_cb { # used by PublicInbox::LeiInput::input_fh + my ($self, $eml, $vmd) = @_; + my $xoids = $self->{lei}->{ale}->xoids_for($eml); + $self->{lei}->{sto}->ipc_do('set_eml', $eml, $vmd, $xoids); +} + +sub mbox_cb { # MboxReader callback used by PublicInbox::LeiInput::input_fh + my ($eml, $self) = @_; + my $vmd; + if ($self->{-import_kw}) { + my $kw = PublicInbox::MboxReader::mbox_keywords($eml); + $vmd = { kw => $kw } if scalar(@$kw); + } + 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 $ign = $lei->{sto}->ipc_do('done'); # PublicInbox::LeiStore::done + my $sto = delete $lei->{sto}; + my $wait = $sto->ipc_do('done') if $sto; # PublicInbox::LeiStore::done $lei->dclose; } @@ -39,6 +51,7 @@ sub net_merge_complete { # callback used by LeiAuth sub import_start { my ($lei) = @_; my $self = $lei->{imp}; + $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 @@ -55,105 +68,24 @@ sub import_start { while ($op && $op->{sock}) { $op->event_step } } -sub call { # the main "lei import" method - my ($cls, $lei, @inputs) = @_; +sub lei_import { # the main "lei import" method + my ($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 $lei->fail(<<""); ---in-format=$fmt and `$ifmt:' conflict - - } - if (-f $input_path) { - require PublicInbox::MboxLock; - require PublicInbox::MboxReader; - PublicInbox::MboxReader->can($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; - } - $self->{inputs} = \@inputs; - if ($net) { - if (my $err = $net->errors) { - return $lei->fail($err); - } - $net->{quiet} = $lei->{opt}->{quiet}; - $lei->{net} = $net; - require PublicInbox::LeiAuth; - $lei->{auth} = PublicInbox::LeiAuth->new; - } + my $self = $lei->{imp} = bless {}, __PACKAGE__; + $self->{-import_kw} = $lei->{opt}->{kw} // 1; + $self->prepare_inputs($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 $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); - } - }; - $lei->child_error(1 << 8, "$input: $@") if $@; -} - -sub _import_maildir { # maildir_each_file cb - my ($f, $sto, $set_kw) = @_; - $sto->ipc_do('set_eml_from_maildir', $f, $set_kw); +sub _import_maildir { # maildir_each_eml cb + my ($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 : ()); + $sto->ipc_do('set_eml', $eml, $set_kw ? { kw => $kw } : ()); } sub import_path_url { @@ -163,7 +95,7 @@ sub import_path_url { # TODO auto-detect? if ($input =~ m!\Aimaps?://!i) { $lei->{net}->imap_each($input, \&_import_net, $lei->{sto}, - $lei->{opt}->{kw}); + $self->{-import_kw}); return; } elsif ($input =~ m!\A(?:nntps?|s?news)://!i) { $lei->{net}->nntp_each($input, \&_import_net, $lei->{sto}, 0); @@ -175,14 +107,14 @@ sub import_path_url { 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); + $self->input_fh($ifmt, $mbl->{fh}, $input); } elsif (-d _ && (-d "$input/cur" || -d "$input/new")) { return $lei->fail(<{sto}, $lei->{opt}->{kw}); + $lei->{sto}, $self->{-import_kw}); } else { $lei->fail("$input unsupported (TODO)"); } @@ -191,9 +123,13 @@ EOM sub import_stdin { my ($self) = @_; my $lei = $self->{lei}; - _import_fh($lei, delete $self->{0}, '', $lei->{opt}->{'in-format'}); + my $in = delete $self->{0}; + $self->input_fh($lei->{opt}->{'in-format'}, $in, ''); } -no warnings 'once'; # the following works even when LeiAuth is lazy-loaded +no warnings 'once'; +*ipc_atfork_child = \&PublicInbox::LeiInput::input_only_atfork_child; + +# the following works even when LeiAuth is lazy-loaded *net_merge_all = \&PublicInbox::LeiAuth::net_merge_all; 1;