# Copyright (C) 2021 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
-# 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);
# JMAP RFC 8621 4.1.1
# https://www.iana.org/assignments/imap-jmap-keywords/imap-jmap-keywords.xhtml
# 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
}
}
+# handles mboxrd endpoints described in Documentation/design_notes.txt
+sub handle_http_input ($$@) {
+ my ($self, $url, @args) = @_;
+ my $lei = $self->{lei} or die 'BUG: {lei} missing';
+ my $curl_opt = delete $self->{"-curl-$url"} or
+ die("BUG: $url curl options not prepared");
+ my $uri = pop @$curl_opt;
+ my $curl = PublicInbox::LeiCurl->new($lei, $self->{curl}) or return;
+ push @$curl, '-s', @$curl_opt;
+ my $cmd = $curl->for_uri($lei, $uri);
+ $lei->qerr("# $cmd");
+ my $rdr = { 2 => $lei->{2}, pgid => 0 };
+ my ($fh, $pid) = popen_rd($cmd, undef, $rdr);
+ grep(/\A--compressed\z/, @$curl) or
+ $fh = IO::Uncompress::Gunzip->new($fh, MultiStream => 1);
+ eval { $self->input_fh('mboxrd', $fh, $url, @args) };
+ my $err = $@;
+ waitpid($pid, 0);
+ $? || $err and
+ $lei->child_error($? || 1, "@$cmd failed".$err ? " $err" : '');
+}
+
sub input_path_url {
my ($self, $input, @args) = @_;
my $lei = $self->{lei};
my $ifmt = lc($lei->{opt}->{'in-format'} // '');
# TODO auto-detect?
if ($input =~ m!\Aimaps?://!i) {
- $lei->{net}->imap_each($input, $self->can('input_imap_cb') //
- $self->can('input_net_cb'),
- $self, @args);
+ $lei->{net}->imap_each($input, $self->can('input_net_cb'),
+ $self, @args);
return;
} elsif ($input =~ m!\A(?:nntps?|s?news)://!i) {
- $lei->{net}->nntp_each($input, $self->can('input_nntp_cb') //
- $self->can('input_net_cb'),
- $self, @args);
+ $lei->{net}->nntp_each($input, $self->can('input_net_cb'),
+ $self, @args);
+ return;
+ } elsif ($input =~ m!\Ahttps?://!i) {
+ handle_http_input($self, $input, @args);
return;
}
if ($input =~ s!\A([a-z0-9]+):!!i) {
}
}
+sub bad_http ($$;$) {
+ my ($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 {
+ 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 = $lei->{opt}->{'mail-sync'} ? {} : undef; # using LeiMailSync
if ($lei->{opt}->{stdin}) {
@$inputs and return
$lei->fail("--stdin and @$inputs do not mix");
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}}, $input;
- }
- }
+ push @{$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 $ifmt) {
--in-format=$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 _)) {
} else {
return $lei->fail("Unable to handle $input");
}
- } elsif ($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
}
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}}");
+ $lei->err("# --mail-sync will only be used for @{$sync->{ok}}");
+ $lei->err("# --mail-sync is not supported for: @{$sync->{no}}");
}
if ($net) {
- if (my $err = $net->errors) {
+ $net->{-can_die} = 1;
+ if (my $err = $net->errors($lei)) {
return $lei->fail($err);
}
$net->{quiet} = $lei->{opt}->{quiet};
sub process_inputs {
my ($self) = @_;
+ my $err;
for my $input (@{$self->{inputs}}) {
- $self->input_path_url($input);
+ eval { $self->input_path_url($input) };
+ next unless $@;
+ $err = "$input: $@";
+ last;
}
+ # always commit first, even on error partial work is acceptable for
+ # lei <import|tag|convert>
my $wait = $self->{lei}->{sto}->ipc_do('done') if $self->{lei}->{sto};
+ $self->{lei}->fail($err) if $err;
}
sub input_only_atfork_child {
undef;
}
+# alias this as "net_merge_all_done" to use as an LeiAuth callback
+sub input_only_net_merge_all_done {
+ my ($self) = @_;
+ $self->wq_io_do('process_inputs');
+ $self->wq_close(1);
+}
+
# like Getopt::Long, but for +kw:FOO and -kw:FOO to prepare
# for update_xvmd -> update_vmd
sub vmd_mod_extract {