X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FLeiInput.pm;h=b6c2b6bbe8b77b30af2e273be04a548049a15937;hb=HEAD;hp=d11d23d4064d374789b8200cab8c23101377e804;hpb=b6b86cfd238c170ea3e2c4d4179f06c7af139086;p=public-inbox.git diff --git a/lib/PublicInbox/LeiInput.pm b/lib/PublicInbox/LeiInput.pm index d11d23d4..b6c2b6bb 100644 --- a/lib/PublicInbox/LeiInput.pm +++ b/lib/PublicInbox/LeiInput.pm @@ -1,11 +1,14 @@ # Copyright (C) 2021 all contributors # License: AGPL-3.0+ -# parent class for LeiImport, LeiConvert +# parent class for LeiImport, LeiConvert, LeiIndex package PublicInbox::LeiInput; use strict; use v5.10.1; use PublicInbox::DS; +use PublicInbox::Spawn qw(which popen_rd); +use PublicInbox::InboxWritable qw(eml_from_path); +use PublicInbox::AutoReap; # JMAP RFC 8621 4.1.1 # https://www.iana.org/assignments/imap-jmap-keywords/imap-jmap-keywords.xhtml @@ -27,7 +30,9 @@ my %ERR = ( my ($label) = @_; length($label) >= $L_MAX and return "`$label' too long (must be <= $L_MAX)"; - $label =~ m{\A[a-z0-9_](?:[a-z0-9_\-\./\@,]*[a-z0-9])?\z}i ? + $label =~ /[A-Z]/ and + return "`$label' must be lowercase"; + $label =~ m{\A[a-z0-9_](?:[a-z0-9_\-\./\@,]*[a-z0-9])?\z} ? undef : "`$label' is invalid"; }, kw => sub { @@ -55,19 +60,41 @@ sub check_input_format ($;$) { 1; } +sub input_mbox_cb { # base MboxReader callback + my ($eml, $self) = @_; + $eml->header_set($_) for (qw(Status X-Status)); + $self->input_eml_cb($eml); +} + +sub input_maildir_cb { + my ($fn, $kw, $eml, $self) = @_; + $self->input_eml_cb($eml); +} + +sub input_net_cb { # imap_each, nntp_each cb + my ($url, $uid, $kw, $eml, $self) = @_; + $self->input_eml_cb($eml); +} + # import a single file handle of $name # Subclass must define ->input_eml_cb and ->input_mbox_cb sub input_fh { my ($self, $ifmt, $fh, $name, @args) = @_; if ($ifmt eq 'eml') { my $buf = do { local $/; <$fh> } // - return $self->{lei}->child_error(1 << 8, <<""); + return $self->{lei}->child_error(0, <<""); error reading $name: $! # mutt pipes single RFC822 messages with a "From " line, # but no Content-Length or "From " escaping. # "git format-patch" also generates such files by default. $buf =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s; + + # a user may feed just a body: git diff | lei rediff -U9 + if ($self->{-force_eml}) { + my $eml = PublicInbox::Eml->new($buf); + substr($buf, 0, 0) = "\n\n" if !$eml->{bdy}; + } $self->input_eml_cb(PublicInbox::Eml->new(\$buf), @args); } else { # prepare_inputs already validated $ifmt @@ -77,26 +104,63 @@ error reading $name: $! } } +# handles mboxrd endpoints described in Documentation/design_notes.txt +sub handle_http_input ($$@) { + my ($self, $url, @args) = @_; + my $lei = $self->{lei} or die 'BUG: {lei} missing'; + my $curl_opt = delete $self->{"-curl-$url"} or + die("BUG: $url curl options not prepared"); + my $uri = pop @$curl_opt; + my $curl = PublicInbox::LeiCurl->new($lei, $self->{curl}) or return; + push @$curl, '-s', @$curl_opt; + my $cmd = $curl->for_uri($lei, $uri); + $lei->qerr("# $cmd"); + my ($fh, $pid) = popen_rd($cmd, undef, { 2 => $lei->{2} }); + my $ar = PublicInbox::AutoReap->new($pid); + grep(/\A--compressed\z/, @$curl) or + $fh = IO::Uncompress::Gunzip->new($fh, MultiStream => 1); + eval { $self->input_fh('mboxrd', $fh, $url, @args) }; + my @err = ($@ ? $@ : ()); + $ar->join; + push(@err, "\$?=$?") if $?; + $lei->child_error($?, "@$cmd failed: @err") if @err; +} + sub input_path_url { my ($self, $input, @args) = @_; my $lei = $self->{lei}; my $ifmt = lc($lei->{opt}->{'in-format'} // ''); # TODO auto-detect? if ($input =~ m!\Aimaps?://!i) { - $lei->{net}->imap_each($input, $self->can('input_imap_cb') // - $self->can('input_net_cb'), - $self, @args); + $lei->{net}->imap_each($input, $self->can('input_net_cb'), + $self, @args); return; } elsif ($input =~ m!\A(?:nntps?|s?news)://!i) { - $lei->{net}->nntp_each($input, $self->can('input_nntp_cb') // - $self->can('input_net_cb'), - $self, @args); + $lei->{net}->nntp_each($input, $self->can('input_net_cb'), + $self, @args); + return; + } elsif ($input =~ m!\Ahttps?://!i) { + handle_http_input($self, $input, @args); return; } + + # local-only below + my $ifmt_pfx = ''; if ($input =~ s!\A([a-z0-9]+):!!i) { + $ifmt_pfx = "$1:"; $ifmt = lc($1); } elsif ($input =~ /\.(?:patch|eml)\z/i) { $ifmt = 'eml'; + } elsif (-f $input && $input =~ m{\A(?:.+)/(?:new|cur)/([^/]+)\z}) { + my $bn = $1; + my $fl = PublicInbox::MdirReader::maildir_basename_flags($bn); + return if index($fl, 'T') >= 0; + return $self->pmdir_cb($input, $fl) if $self->can('pmdir_cb'); + my $eml = eml_from_path($input) or return + $lei->qerr("# $input not readable"); + my $kw = PublicInbox::MdirReader::flags2kw($fl); + $self->can('input_maildir_cb')->($input, $kw, $eml, $self); + return; } my $devfd = $lei->path_to_fd($input) // return; if ($devfd >= 0) { @@ -115,24 +179,91 @@ sub input_path_url { $mbl->{fh} = PublicInbox::MboxReader::zsfxcat($in, $zsfx, $lei); } - local $PublicInbox::DS::in_loop = 0 if $zsfx; # dwaitpid + local $PublicInbox::DS::in_loop = 0 if $zsfx; # awaitpid $self->input_fh($ifmt, $mbl->{fh}, $input, @args); } elsif (-d _ && (-d "$input/cur" || -d "$input/new")) { return $lei->fail(<new->maildir_each_eml($input, - $self->can('input_maildir_cb'), - $self, @args); + my $mdr = PublicInbox::MdirReader->new; + if (my $pmd = $self->{pmd}) { + $mdr->maildir_each_file($input, + $pmd->can('each_mdir_fn'), + $pmd, @args); + } else { + $mdr->maildir_each_eml($input, + $self->can('input_maildir_cb'), + $self, @args); + } + } elsif ($self->{missing_ok} && !-e $input) { # don't ->fail + if ($lei->{cmd} eq 'p2q') { + my $fp = [ qw(git format-patch --stdout -1), $input ]; + my $rdr = { 2 => $lei->{2} }; + my $fh = popen_rd($fp, undef, $rdr); + eval { $self->input_fh('eml', $fh, $input, @args) }; + my @err = ($@ ? $@ : ()); + close($fh) or push @err, "\$?=$?"; + $lei->child_error($?, "@$fp failed: @err") if @err; + } else { + $self->folder_missing("$ifmt:$input"); + } + } else { + $lei->fail("$ifmt_pfx$input unsupported (TODO)"); + } +} + +# subclasses should overrride this (see LeiRefreshMailSync) +sub folder_missing { die "BUG: ->folder_missing undefined for $_[0]" } + +sub bad_http ($$;$) { + my ($lei, $url, $alt) = @_; + my $x = $alt ? "did you mean <$alt>?" : 'download and import manually'; + $lei->fail("E: <$url> not recognized, $x"); +} + +sub prepare_http_input ($$$) { + my ($self, $lei, $url) = @_; + require URI; + require PublicInbox::MboxReader; + require PublicInbox::LeiCurl; + require IO::Uncompress::Gunzip; + $self->{curl} //= which('curl') or + return $lei->fail("curl missing for <$url>"); + my $uri = URI->new($url); + my $path = $uri->path; + my %qf = $uri->query_form; + my @curl_opt; + if ($path =~ m!/(?:t\.mbox\.gz|all\.mbox\.gz)\z!) { + # OK + } elsif ($path =~ m!/raw\z!) { + push @curl_opt, '--compressed'; + # convert search query to mboxrd request since they require POST + # this is only intended for PublicInbox::WWW, and will false-positive + # on many other search engines... oh well + } elsif (defined $qf{'q'}) { + $qf{x} = 'm'; + $uri->query_form(\%qf); + push @curl_opt, '-d', ''; + $$uri ne $url and $lei->qerr(<<""); +# <$url> rewritten to <$$uri> with HTTP POST + + # try to provide hints for /$INBOX/$MSGID/T/ and /$INBOX/ + } elsif ($path =~ s!/[tT]/\z!/t.mbox.gz! || + $path =~ s!/t\.atom\z!/t.mbox.gz! || + $path =~ s!/([^/]+\@[^/]+)/\z!/$1/raw!) { + $uri->path($path); + return bad_http($lei, $url, $$uri); } else { - $lei->fail("$input unsupported (TODO)"); + return bad_http($lei, $url); } + $self->{"-curl-$url"} = [ @curl_opt, $uri ]; # for handle_http_input } sub prepare_inputs { # returns undef on error my ($self, $lei, $inputs) = @_; my $in_fmt = $lei->{opt}->{'in-format'}; - my $sync = $lei->{opt}->{sync} ? {} : undef; # using LeiMailSync + my $sync = $lei->{opt}->{'mail-sync'} ? {} : undef; # using LeiMailSync + my $may_sync = $sync || $self->{-mail_sync}; if ($lei->{opt}->{stdin}) { @$inputs and return $lei->fail("--stdin and @$inputs do not mix"); @@ -141,21 +272,19 @@ sub prepare_inputs { # returns undef on error push @{$sync->{no}}, '/dev/stdin' if $sync; } my $net = $lei->{net}; # NetWriter may be created by l2m - my (@f, @d); + my (@f, @md); # 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); - if ($sync) { - if ($input =~ m!\Aimaps?://!) { - push @{$sync->{ok}}, $input; - } else { - push @{$sync->{no}}, $input; - } - } + $net->add_url($input, $self->{-ls_ok}); + push @{$sync->{ok}}, $input if $sync; + } elsif ($input_path =~ m!\Ahttps?://!i) { # mboxrd.gz + # TODO: how would we detect r/w JMAP? + push @{$sync->{no}}, $input if $sync; + prepare_http_input($self, $lei, $input_path) or return; } elsif ($input_path =~ s/\A([a-z0-9]+)://is) { my $ifmt = lc $1; if (($in_fmt // $ifmt) ne $ifmt) { @@ -163,12 +292,10 @@ sub prepare_inputs { # returns undef on error --in-format=$in_fmt and `$ifmt:' conflict } - if ($sync) { - if ($ifmt =~ /\A(?:maildir|mh)\z/i) { - push @{$sync->{ok}}, $input; - } else { - push @{$sync->{no}}, $input; - } + if ($ifmt =~ /\A(?:maildir|mh)\z/i) { + push @{$sync->{ok}}, $input if $sync; + } else { + push @{$sync->{no}}, $input if $sync; } my $devfd = $lei->path_to_fd($input_path) // return; if ($devfd >= 0 || (-f $input_path || -p _)) { @@ -177,49 +304,77 @@ sub prepare_inputs { # returns undef on error PublicInbox::MboxReader->reads($ifmt) or return $lei->fail("$ifmt not supported"); } elsif (-d $input_path) { - require PublicInbox::MdirReader; $ifmt eq 'maildir' or return $lei->fail("$ifmt not supported"); - $input = $lei->abs_path($input) if $sync; + $may_sync and $input = 'maildir:'. + $lei->abs_path($input_path); + push @md, $input; + } elsif ($self->{missing_ok} && !-e _) { + # for "lei rm-watch" on missing Maildir + $may_sync and $input = 'maildir:'. + $lei->abs_path($input_path); } else { - return $lei->fail("Unable to handle $input"); + my $m = "Unable to handle $input"; + $input =~ /\A(?:L|kw):/ and + $m .= ", did you mean +$input?"; + return $lei->fail($m); } - } elsif ($input =~ /\.(eml|patch)\z/i && -f $input) { + } elsif ($input =~ /\.(?:eml|patch)\z/i && -f $input) { lc($in_fmt//'eml') eq 'eml' or return $lei->fail(<<""); $input is `eml', not --in-format=$in_fmt - require PublicInbox::Eml; push @{$sync->{no}}, $input if $sync; + } elsif (-f $input && $input =~ m{\A(.+)/(new|cur)/([^/]+)\z}) { + # single file in a Maildir + my ($mdir, $nc, $bn) = ($1, $2, $3); + my $other = $mdir . ($nc eq 'new' ? '/cur' : '/new'); + return $lei->fail(<fail(<<""); +$input is `eml', not --in-format=$in_fmt + + if ($sync) { + $input = $lei->abs_path($mdir) . "/$nc/$bn"; + push @{$sync->{ok}}, $input if $sync; + } + require PublicInbox::MdirReader; } else { my $devfd = $lei->path_to_fd($input) // return; if ($devfd >= 0 || -f $input || -p _) { push @{$sync->{no}}, $input if $sync; push @f, $input; - } elsif (-d $input) { - if ($sync) { - $input = $lei->abs_path($input); - push @{$sync->{ok}}, $input; + } elsif (-d "$input/new" && -d "$input/cur") { + if ($may_sync) { + $input = 'maildir:'. + $lei->abs_path($input); + push @{$sync->{ok}}, $input if $sync; + } + push @md, $input; + } elsif ($self->{missing_ok} && !-e $input) { + if ($lei->{cmd} eq 'p2q') { + # will run "git format-patch" + } elsif ($may_sync) { # for lei rm-watch + $input = 'maildir:'. + $lei->abs_path($input); } - push @d, $input; } else { return $lei->fail("Unable to handle $input") } } } if (@f) { check_input_format($lei, \@f) or return } - if (@d) { # TODO: check for MH vs Maildir, here - require PublicInbox::MdirReader; - } if ($sync && $sync->{no}) { return $lei->fail(<<"") if !$sync->{ok}; ---sync specified but no inputs support it +--mail-sync specified but no inputs support it # non-fatal if some inputs support support sync - $lei->err("# --sync will only be used for @{$sync->{ok}}"); - $lei->err("# --sync is not supported for: @{$sync->{no}}"); + warn("# --mail-sync will only be used for @{$sync->{ok}}\n"); + warn("# --mail-sync is not supported for: @{$sync->{no}}\n"); } if ($net) { - if (my $err = $net->errors) { + $net->{-can_die} = 1; + if (my $err = $net->errors($lei)) { return $lei->fail($err); } $net->{quiet} = $lei->{opt}->{quiet}; @@ -227,15 +382,36 @@ $input is `eml', not --in-format=$in_fmt $lei->{auth} //= PublicInbox::LeiAuth->new; $lei->{net} //= $net; } + if (scalar(@md)) { + require PublicInbox::MdirReader; + if ($self->can('pmdir_cb')) { + require PublicInbox::LeiPmdir; + $self->{pmd} = PublicInbox::LeiPmdir->new($lei, $self); + } + + # start watching Maildirs ASAP + if ($may_sync && $lei->{sto}) { + grep(!m!\Amaildir:/!i, @md) and die "BUG: @md (no pfx)"; + $lei->lms(1)->lms_write_prepare->add_folders(@md); + $lei->refresh_watches; + } + } $self->{inputs} = $inputs; } sub process_inputs { my ($self) = @_; + my $err; for my $input (@{$self->{inputs}}) { - $self->input_path_url($input); + eval { $self->input_path_url($input) }; + next unless $@; + $err = "$input: $@"; + last; } - my $wait = $self->{lei}->{sto}->ipc_do('done') if $self->{lei}->{sto}; + # always commit first, even on error partial work is acceptable for + # lei + my $wait = $self->{lei}->{sto}->wq_do('done') if $self->{lei}->{sto}; + $self->{lei}->fail($err) if $err; } sub input_only_atfork_child { @@ -247,8 +423,16 @@ sub input_only_atfork_child { undef; } +# alias this as "net_merge_all_done" to use as an LeiAuth callback +sub input_only_net_merge_all_done { + my ($self) = @_; + $self->wq_io_do('process_inputs'); + $self->wq_close; +} + # like Getopt::Long, but for +kw:FOO and -kw:FOO to prepare # for update_xvmd -> update_vmd +# returns something like { "+L" => [ @Labels ], ... } sub vmd_mod_extract { my $argv = $_[-1]; my $vmd_mod = {};