X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FImport.pm;h=60ce7b66419b3ee0e9ad2dca4b93644497539378;hb=23af251dd607c4e75ab1e68063f2c885c48cc035;hp=de8ff55f4abd9c23c7b0be5bfe888a15f348fd40;hpb=129c630a7ec3e100bba272631c3880b9beac9cef;p=public-inbox.git diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm index de8ff55f..60ce7b66 100644 --- a/lib/PublicInbox/Import.pm +++ b/lib/PublicInbox/Import.pm @@ -1,4 +1,4 @@ -# Copyright (C) 2016-2020 all contributors +# Copyright (C) 2016-2021 all contributors # License: AGPL-3.0+ # # git fast-import-based ssoma-mda MDA replacement @@ -7,33 +7,45 @@ # requires read-only access. package PublicInbox::Import; use strict; -use warnings; -use base qw(PublicInbox::Lock); -use PublicInbox::Spawn qw(spawn popen_rd); +use parent qw(PublicInbox::Lock); +use v5.10.1; +use PublicInbox::Spawn qw(run_die popen_rd); use PublicInbox::MID qw(mids mid2path); use PublicInbox::Address; -use PublicInbox::MsgTime qw(msg_timestamp msg_datestamp); -use PublicInbox::ContentId qw(content_digest); +use PublicInbox::Smsg; +use PublicInbox::MsgTime qw(msg_datestamp); +use PublicInbox::ContentHash qw(content_digest); use PublicInbox::MDA; +use PublicInbox::Eml; use POSIX qw(strftime); +sub default_branch () { + state $default_branch = do { + my $r = popen_rd([qw(git config --global init.defaultBranch)], + { GIT_CONFIG => undef }); + chomp(my $h = <$r> // ''); + close $r; + $h eq '' ? 'refs/heads/master' : "refs/heads/$h"; + } +} + sub new { # we can't change arg order, this is documented in POD # and external projects may rely on it: my ($class, $git, $name, $email, $ibx) = @_; - my $ref = 'refs/heads/master'; + my $ref; if ($ibx) { - $ref = $ibx->{ref_head} || 'refs/heads/master'; - $name ||= $ibx->{name}; - $email ||= $ibx->{-primary_address}; - $git ||= $ibx->git; + $ref = $ibx->{ref_head}; + $name //= $ibx->{name}; + $email //= $ibx->{-primary_address}; + $git //= $ibx->git; } bless { git => $git, ident => "$name <$email>", mark => 1, - ref => $ref, - -inbox => $ibx, + ref => $ref // default_branch, + ibx => $ibx, path_type => '2/38', # or 'v2' lock_path => "$git->{git_dir}/ssoma.lock", # v2 changes this bytes_added => 0, @@ -44,35 +56,33 @@ sub new { sub gfi_start { my ($self) = @_; - return ($self->{in}, $self->{out}) if $self->{pid}; + return ($self->{in}, $self->{out}) if $self->{in}; - my ($out_r, $out_w); + my ($in_r, $out_r, $out_w); pipe($out_r, $out_w) or die "pipe failed: $!"; - my $git = $self->{git}; $self->lock_acquire; - - local $/ = "\n"; - my $ref = $self->{ref}; - chomp($self->{tip} = $git->qx(qw(rev-parse --revs-only), $ref)); - if ($self->{path_type} ne '2/38' && $self->{tip}) { - local $/ = "\0"; - my @tree = $git->qx(qw(ls-tree -r -z --name-only), $ref); - chomp @tree; - $self->{-tree} = { map { $_ => 1 } @tree }; - } - - my $git_dir = $git->{git_dir}; - my @cmd = ('git', "--git-dir=$git_dir", qw(fast-import - --quiet --done --date-format=raw)); - my ($in_r, $pid) = popen_rd(\@cmd, undef, { 0 => $out_r }); - $out_w->autoflush(1); - $self->{in} = $in_r; - $self->{out} = $out_w; - $self->{pid} = $pid; - $self->{nchg} = 0; - binmode $out_w, ':raw' or die "binmode :raw failed: $!"; - binmode $in_r, ':raw' or die "binmode :raw failed: $!"; + eval { + my ($git, $ref) = @$self{qw(git ref)}; + local $/ = "\n"; + chomp($self->{tip} = $git->qx(qw(rev-parse --revs-only), $ref)); + die "fatal: rev-parse --revs-only $ref: \$?=$?" if $?; + if ($self->{path_type} ne '2/38' && $self->{tip}) { + my $t = $git->qx(qw(ls-tree -r -z --name-only), $ref); + die "fatal: ls-tree -r -z --name-only $ref: \$?=$?" if $?; + $self->{-tree} = { map { $_ => 1 } split(/\0/, $t) }; + } + $in_r = $self->{in} = $git->popen(qw(fast-import + --quiet --done --date-format=raw), + undef, { 0 => $out_r }); + $out_w->autoflush(1); + $self->{out} = $out_w; + $self->{nchg} = 0; + }; + if ($@) { + $self->lock_release; + die $@; + } ($in_r, $out_w); } @@ -103,7 +113,7 @@ sub _cat_blob ($$$) { local $/ = "\n"; my $info = <$r>; defined $info or die "EOF from fast-import / cat-blob: $!"; - $info =~ /\A[a-f0-9]{40} blob ([0-9]+)\n\z/ or return; + $info =~ /\A[a-f0-9]{40,} blob ([0-9]+)\n\z/ or return; my $left = $1; my $offset = 0; my $buf = ''; @@ -134,10 +144,10 @@ sub check_remove_v1 { my ($r, $w, $tip, $path, $mime) = @_; my $info = _check_path($r, $w, $tip, $path) or return ('MISSING',undef); - $info =~ m!\A100644 blob ([a-f0-9]{40})\t!s or die "not blob: $info"; + $info =~ m!\A100644 blob ([a-f0-9]{40,})\t!s or die "not blob: $info"; my $oid = $1; my $msg = _cat_blob($r, $w, $oid) or die "BUG: cat-blob $1 failed"; - my $cur = PublicInbox::MIME->new($msg); + my $cur = PublicInbox::Eml->new($msg); my $cur_s = $cur->header('Subject'); $cur_s = '' unless defined $cur_s; my $cur_m = $mime->header('Subject'); @@ -150,16 +160,16 @@ sub check_remove_v1 { sub checkpoint { my ($self) = @_; - return unless $self->{pid}; + return unless $self->{in}; print { $self->{out} } "checkpoint\n" or wfail; undef; } sub progress { my ($self, $msg) = @_; - return unless $self->{pid}; + return unless $self->{in}; print { $self->{out} } "progress $msg\n" or wfail; - $self->{in}->getline eq "progress $msg\n" or die + readline($self->{in}) eq "progress $msg\n" or die "progress $msg not received\n"; undef; } @@ -175,13 +185,16 @@ sub _update_git_info ($$) { my $env = { GIT_INDEX_FILE => $index }; run_die([@cmd, qw(read-tree -m -v -i), $self->{ref}], $env); } - run_die([@cmd, 'update-server-info']); - my $ibx = $self->{-inbox}; - ($ibx && $self->{path_type} eq '2/38') and eval { - require PublicInbox::SearchIdx; - my $s = PublicInbox::SearchIdx->new($ibx); - $s->index_sync({ ref => $self->{ref} }); - }; + eval { run_die([@cmd, 'update-server-info']) }; + my $ibx = $self->{ibx}; + if ($ibx && $ibx->version == 1 && -d "$ibx->{inboxdir}/public-inbox" && + eval { require PublicInbox::SearchIdx }) { + eval { + my $s = PublicInbox::SearchIdx->new($ibx); + $s->index_sync({ ref => $self->{ref} }); + }; + warn "$ibx->{inboxdir} index failed: $@\n" if $@; + } eval { run_die([@cmd, qw(gc --auto)]) } if $do_gc; } @@ -203,7 +216,7 @@ sub barrier { # used for v2 sub get_mark { my ($self, $mark) = @_; - die "not active\n" unless $self->{pid}; + die "not active\n" unless $self->{in}; my ($r, $w) = $self->gfi_start; print $w "get-mark $mark\n" or wfail; defined(my $oid = <$r>) or die "get-mark failed, need git 2.6.0+\n"; @@ -212,13 +225,13 @@ sub get_mark { } # returns undef on non-existent -# ('MISMATCH', Email::MIME) on mismatch -# (:MARK, Email::MIME) on success +# ('MISMATCH', PublicInbox::Eml) on mismatch +# (:MARK, PublicInbox::Eml) on success # # v2 callers should check with Xapian before calling this as # it is not idempotent. sub remove { - my ($self, $mime, $msg) = @_; # mime = Email::MIME + my ($self, $mime, $msg) = @_; # mime = PublicInbox::Eml or Email::MIME my $path_type = $self->{path_type}; my ($path, $err, $cur, $blob); @@ -251,7 +264,7 @@ sub remove { } my $ident = $self->{ident}; my $now = now_raw(); - $msg ||= 'rm'; + $msg //= 'rm'; my $len = length($msg) + 1; print $w "commit $ref\nmark :$commit\n", "author $ident $now\n", @@ -268,29 +281,27 @@ sub remove { (($self->{tip} = ":$commit"), $cur); } -sub git_timestamp { - my ($ts, $zone) = @_; +sub git_timestamp ($) { + my ($ts, $zone) = @{$_[0]}; $ts = 0 if $ts < 0; # git uses unsigned times "$ts $zone"; } sub extract_cmt_info ($;$) { my ($mime, $smsg) = @_; + # $mime is PublicInbox::Eml, but remains Email::MIME-compatible + $smsg //= bless {}, 'PublicInbox::Smsg'; + + $smsg->populate($mime); my $sender = ''; - my $from = $mime->header('From'); - $from ||= ''; + my $from = delete($smsg->{From}) // ''; my ($email) = PublicInbox::Address::emails($from); my ($name) = PublicInbox::Address::names($from); if (!defined($name) || !defined($email)) { - $sender = $mime->header('Sender'); - $sender ||= ''; - if (!defined($name)) { - ($name) = PublicInbox::Address::names($sender); - } - if (!defined($email)) { - ($email) = PublicInbox::Address::emails($sender); - } + $sender = $mime->header('Sender') // ''; + $name //= (PublicInbox::Address::names($sender))[0]; + $email //= (PublicInbox::Address::emails($sender))[0]; } if (defined $email) { # Email::Address::XS may leave quoted '<' in addresses, @@ -316,28 +327,21 @@ sub extract_cmt_info ($;$) { warn "no name in From: $from or Sender: $sender\n"; } - my $hdr = $mime->header_obj; - - my $subject = $hdr->header('Subject'); - $subject = '(no subject)' unless defined $subject; - # Mime decoding can create nulls replace them with spaces to protect git - $subject =~ tr/\0/ /; + my $subject = delete($smsg->{Subject}) // '(no subject)'; utf8::encode($subject); - my $at = git_timestamp(my @at = msg_datestamp($hdr)); - my $ct = git_timestamp(my @ct = msg_timestamp($hdr)); - if ($smsg) { - $smsg->{ds} = $at[0]; - $smsg->{ts} = $ct[0]; - } - ($name, $email, $at, $ct, $subject); + my $at = git_timestamp(delete $smsg->{-ds}); + my $ct = git_timestamp(delete $smsg->{-ts}); + ("$name <$email>", $at, $ct, $subject); } # kill potentially confusing/misleading headers +our @UNWANTED_HEADERS = (qw(Bytes Lines Content-Length), + qw(Status X-Status)); sub drop_unwanted_headers ($) { - my ($mime) = @_; - - $mime->header_set($_) for qw(bytes lines content-length status); - $mime->header_set($_) for @PublicInbox::MDA::BAD_HEADERS; + my ($eml) = @_; + for (@UNWANTED_HEADERS, @PublicInbox::MDA::BAD_HEADERS) { + $eml->header_set($_); + } } # used by V2Writable, too @@ -350,13 +354,12 @@ sub append_mid ($$) { } sub v1_mid0 ($) { - my ($mime) = @_; - my $hdr = $mime->header_obj; - my $mids = mids($hdr); + my ($eml) = @_; + my $mids = mids($eml); if (!scalar(@$mids)) { # spam often has no Message-ID - my $mid0 = digest2mid(content_digest($mime), $hdr); - append_mid($hdr, $mid0); + my $mid0 = digest2mid(content_digest($eml), $eml); + append_mid($eml, $mid0); return $mid0; } $mids->[0]; @@ -374,9 +377,9 @@ sub clean_tree_v2 ($$$) { # returns undef on duplicate # returns the :MARK of the most recent commit sub add { - my ($self, $mime, $check_cb, $smsg) = @_; # mime = Email::MIME + my ($self, $mime, $check_cb, $smsg) = @_; - my ($name, $email, $at, $ct, $subject) = extract_cmt_info($mime, $smsg); + my ($author, $at, $ct, $subject) = extract_cmt_info($mime, $smsg); my $path_type = $self->{path_type}; my $path; if ($path_type eq '2/38') { @@ -395,7 +398,7 @@ sub add { # spam check: if ($check_cb) { - $mime = $check_cb->($mime) or return; + $mime = $check_cb->($mime, $self->{ibx}) or return; } my $blob = $self->{mark}++; @@ -408,8 +411,22 @@ sub add { # v2: we need this for Xapian if ($smsg) { $smsg->{blob} = $self->get_mark(":$blob"); - $smsg->{bytes} = $n; - $smsg->{-raw_email} = \$raw_email; + $smsg->set_bytes($raw_email, $n); + if (my $oidx = delete $smsg->{-oidx}) { # used by LeiStore + my $eidx_git = delete $smsg->{-eidx_git}; + + # we need this sharedkv to dedupe blobs added in the + # same fast-import transaction + my $u = $self->{uniq_skv} //= do { + require PublicInbox::SharedKV; + my $x = PublicInbox::SharedKV->new; + $x->dbh; + $x; + }; + return if !$u->set_maybe($smsg->oidbin, 1); + return if (!$oidx->vivify_xvmd($smsg) && + $eidx_git->check($smsg->{blob})); + } } my $ref = $self->{ref}; my $commit = $self->{mark}++; @@ -420,7 +437,7 @@ sub add { } print $w "commit $ref\nmark :$commit\n", - "author $name <$email> $at\n", + "author $author $at\n", "committer $self->{ident} $ct\n" or wfail; print $w "data ", (length($subject) + 1), "\n", $subject, "\n\n" or wfail; @@ -433,17 +450,7 @@ sub add { $self->{tip} = ":$commit"; } -sub run_die ($;$$) { - my ($cmd, $env, $rdr) = @_; - my $pid = spawn($cmd, $env, $rdr); - waitpid($pid, 0) == $pid or die join(' ', @$cmd) .' did not finish'; - $? == 0 or die join(' ', @$cmd) . " failed: $?\n"; -} - -my @INIT_FILES = ('HEAD' => "ref: refs/heads/master\n", - 'description' => < undef, # filled in at runtime 'config' => <{git}->{git_dir} if ref($dir); require File::Path; File::Path::mkpath([ map { "$dir/$_" } qw(objects/info refs/heads) ]); - for (my $i = 0; $i < @INIT_FILES; $i++) { - my $f = $dir.'/'.$INIT_FILES[$i++]; + $INIT_FILES[1] //= 'ref: '.default_branch."\n"; + my @fn_contents = @INIT_FILES; + $fn_contents[1] = "ref: refs/heads/$head\n" if defined $head; + while (my ($fn, $contents) = splice(@fn_contents, 0, 2)) { + my $f = $dir.'/'.$fn; next if -f $f; open my $fh, '>', $f or die "open $f: $!"; - print $fh $INIT_FILES[$i] or die "print $f: $!"; + print $fh $contents or die "print $f: $!"; close $fh or die "close $f: $!"; } } +# true if locked and active +sub active { !!$_[0]->{out} } + sub done { my ($self) = @_; my $w = delete $self->{out} or return; - my $r = delete $self->{in} or die 'BUG: missing {in} when done'; - print $w "done\n" or wfail; - my $pid = delete $self->{pid} or die 'BUG: missing {pid} when done'; - waitpid($pid, 0) == $pid or die 'fast-import did not finish'; - $? == 0 or die "fast-import failed: $?"; - - _update_git_info($self, 1) if delete $self->{nchg}; - - $self->lock_release; - + eval { + my $r = delete $self->{in} or die 'BUG: missing {in} when done'; + print $w "done\n" or wfail; + close $r or die "fast-import failed: $?"; # ProcessPipe::CLOSE + }; + my $wait_err = $@; + my $nchg = delete $self->{nchg}; + if ($nchg && !$wait_err) { + eval { _update_git_info($self, 1) }; + warn "E: $self->{git}->{git_dir} update info: $@\n" if $@; + } + $self->lock_release(!!$nchg); $self->{git}->cleanup; + die $wait_err if $wait_err; } sub atfork_child { @@ -491,8 +507,8 @@ sub atfork_child { } } -sub digest2mid ($$) { - my ($dig, $hdr) = @_; +sub digest2mid ($$;$) { + my ($dig, $hdr, $fallback_time) = @_; my $b64 = $dig->clone->b64digest; # Make our own URLs nicer: # See "Base 64 Encoding with URL and Filename Safe Alphabet" in RFC4648 @@ -501,18 +517,18 @@ sub digest2mid ($$) { # Add a date prefix to prevent a leading '-' in case that trips # up some tools (e.g. if a Message-ID were a expected as a # command-line arg) - my $dt = msg_datestamp($hdr); + my $dt = msg_datestamp($hdr, $fallback_time); $dt = POSIX::strftime('%Y%m%d%H%M%S', gmtime($dt)); "$dt.$b64" . '@z'; } sub rewrite_commit ($$$$) { my ($self, $oids, $buf, $mime) = @_; - my ($name, $email, $at, $ct, $subject); + my ($author, $at, $ct, $subject); if ($mime) { - ($name, $email, $at, $ct, $subject) = extract_cmt_info($mime); + ($author, $at, $ct, $subject) = extract_cmt_info($mime); } else { - $name = $email = ''; + $author = '<>'; $subject = 'purged '.join(' ', @$oids); } @$oids = (); @@ -521,7 +537,7 @@ sub rewrite_commit ($$$$) { my $l = $buf->[$i]; if ($l =~ /^author .* ([0-9]+ [\+-]?[0-9]+)$/) { $at //= $1; - $buf->[$i] = "author $name <$email> $at\n"; + $buf->[$i] = "author $author $at\n"; } elsif ($l =~ /^committer .* ([0-9]+ [\+-]?[0-9]+)$/) { $ct //= $1; $buf->[$i] = "committer $self->{ident} $ct\n"; @@ -553,7 +569,7 @@ sub replace_oids { push @buf, "reset $tmp\n"; } elsif (/^commit (?:.+)/) { if (@buf) { - $w->print(@buf) or wfail; + print $w @buf or wfail; @buf = (); } push @buf, "commit $tmp\n"; @@ -590,7 +606,7 @@ sub replace_oids { rewrite_commit($self, \@oids, \@buf, $mime); $nreplace++; } - $w->print(@buf, "\n") or wfail; + print $w @buf, "\n" or wfail; @buf = (); } elsif ($_ eq "done\n") { $done = 1; @@ -603,7 +619,7 @@ sub replace_oids { } close $rd or die "close fast-export failed: $?"; if (@buf) { - $w->print(@buf) or wfail; + print $w @buf or wfail; } die 'done\n not seen from fast-export' unless $done; chomp(my $cmt = $self->get_mark(":$mark")) if $nreplace; @@ -647,7 +663,10 @@ version 1.0 =head1 SYNOPSIS - use Email::MIME; + use PublicInbox::Eml; + # PublicInbox::Eml exists as of public-inbox 1.5.0, + # Email::MIME was used in older versions + use PublicInbox::Git; use PublicInbox::Import; @@ -663,18 +682,17 @@ version 1.0 "Date: Thu, 01 Jan 1970 00:00:00 +0000\n" . "Message-ID: \n". "\ntest message"; - my $parsed = Email::MIME->new($message); + my $parsed = PublicInbox::Eml->new($message); my $ret = $im->add($parsed); if (!defined $ret) { - warn "duplicate: ", - $parsed->header_obj->header_raw('Message-ID'), "\n"; + warn "duplicate: ", $parsed->header_raw('Message-ID'), "\n"; } else { print "imported at mark $ret\n"; } $im->done; # to remove a message - my $junk = Email::MIME->new($message); + my $junk = PublicInbox::Eml->new($message); my ($mark, $orig) = $im->remove($junk); if ($mark eq 'MISSING') { print "not found\n"; @@ -689,8 +707,8 @@ version 1.0 =head1 DESCRIPTION -An importer and remover for public-inboxes which takes L -messages as input and stores them in a git repository as +An importer and remover for public-inboxes which takes C +or L messages as input and stores them in a git repository as documented in L, except it does not allow duplicate Message-IDs. @@ -708,7 +726,7 @@ Initialize a new PublicInbox::Import object. =head2 add - my $parsed = Email::MIME->new($message); + my $parsed = PublicInbox::Eml->new($message); $im->add($parsed); Adds a message to to the git repository. This will acquire @@ -719,12 +737,13 @@ is called, but L may be called on them. =head2 remove - my $junk = Email::MIME->new($message); + my $junk = PublicInbox::Eml->new($message); my ($code, $orig) = $im->remove($junk); Removes a message from the repository. On success, it returns a ':'-prefixed numeric code representing the git-fast-import -mark and the original messages as an Email::MIME object. +mark and the original messages as a PublicInbox::Eml +(or Email::MIME) object. If the message could not be found, the code is "MISSING" and the original message is undef. If there is a mismatch where the "Message-ID" is matched but the subject and body do not match, @@ -748,7 +767,7 @@ The mail archives are hosted at L =head1 COPYRIGHT -Copyright (C) 2016 all contributors L +Copyright (C) 2016-2020 all contributors L License: AGPL-3.0+ L