-# Copyright (C) 2016-2020 all contributors <meta@public-inbox.org>
+# Copyright (C) all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
#
# git fast-import-based ssoma-mda MDA replacement
use strict;
use parent qw(PublicInbox::Lock);
use v5.10.1;
-use PublicInbox::Spawn qw(spawn popen_rd);
+use PublicInbox::Spawn qw(run_die popen_rd);
use PublicInbox::MID qw(mids mid2path);
use PublicInbox::Address;
use PublicInbox::Smsg;
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';
+ $ref = $ibx->{ref_head};
$name //= $ibx->{name};
$email //= $ibx->{-primary_address};
$git //= $ibx->git;
git => $git,
ident => "$name <$email>",
mark => 1,
- ref => $ref,
+ ref => $ref // default_branch,
ibx => $ibx,
path_type => '2/38', # or 'v2'
lock_path => "$git->{git_dir}/ssoma.lock", # v2 changes this
sub gfi_start {
my ($self) = @_;
- return ($self->{in}, $self->{out}) if $self->{pid};
+ return ($self->{in}, $self->{out}) if $self->{in};
- my (@ret, $out_r, $out_w);
+ my ($in_r, $out_r, $out_w);
pipe($out_r, $out_w) or die "pipe failed: $!";
$self->lock_acquire;
my ($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}) {
- local $/ = "\0";
- my @t = $git->qx(qw(ls-tree -r -z --name-only), $ref);
- chomp @t;
- $self->{-tree} = { map { $_ => 1 } @t };
+ 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/, $t) };
}
- my @cmd = ('git', "--git-dir=$git->{git_dir}",
- qw(fast-import --quiet --done --date-format=raw));
- my ($in_r, $pid) = popen_rd(\@cmd, undef, { 0 => $out_r });
+ $in_r = $self->{in} = $git->popen(qw(fast-import
+ --quiet --done --date-format=raw),
+ undef, { 0 => $out_r });
$out_w->autoflush(1);
- $self->{in} = $in_r;
$self->{out} = $out_w;
- $self->{pid} = $pid;
$self->{nchg} = 0;
- @ret = ($in_r, $out_w);
};
if ($@) {
$self->lock_release;
die $@;
}
- @ret;
+ ($in_r, $out_w);
}
sub wfail () { die "write to fast-import failed: $!" }
return if $tip eq '';
print $w "ls $tip $path\n" or wfail;
local $/ = "\n";
- defined(my $info = <$r>) or die "EOF from fast-import: $!";
+ my $info = <$r> // die "EOF from fast-import: $!";
$info =~ /\Amissing / ? undef : $info;
}
my ($r, $w, $oid) = @_;
print $w "cat-blob $oid\n" or wfail;
local $/ = "\n";
- my $info = <$r>;
- defined $info or die "EOF from fast-import / cat-blob: $!";
+ my $info = <$r> // die "EOF from fast-import / cat-blob: $!";
$info =~ /\A[a-f0-9]{40,} blob ([0-9]+)\n\z/ or return;
my $left = $1;
my $offset = 0;
my $buf = '';
my $n;
while ($left > 0) {
- $n = read($r, $buf, $left, $offset);
- defined($n) or die "read cat-blob failed: $!";
+ $n = read($r, $buf, $left, $offset) //
+ die "read cat-blob failed: $!";
$n == 0 and die 'fast-export (cat-blob) died';
$left -= $n;
$offset += $n;
}
- $n = read($r, my $lf, 1);
- defined($n) or die "read final byte of cat-blob failed: $!";
+ $n = read($r, my $lf, 1) //
+ die "read final byte of cat-blob failed: $!";
die "bad read on final byte: <$lf>" if $lf ne "\n";
# fixup some bugginess in old versions:
my $oid = $1;
my $msg = _cat_blob($r, $w, $oid) or die "BUG: cat-blob $1 failed";
my $cur = PublicInbox::Eml->new($msg);
- my $cur_s = $cur->header('Subject');
- $cur_s = '' unless defined $cur_s;
- my $cur_m = $mime->header('Subject');
- $cur_m = '' unless defined $cur_m;
+ my $cur_s = $cur->header('Subject') // '';
+ my $cur_m = $mime->header('Subject') // '';
if ($cur_s ne $cur_m || norm_body($cur) ne norm_body($mime)) {
return ('MISMATCH', $cur);
}
sub checkpoint {
my ($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 $msg\n" or wfail;
readline($self->{in}) eq "progress $msg\n" or die
"progress $msg not received\n";
my $env = { GIT_INDEX_FILE => $index };
run_die([@cmd, qw(read-tree -m -v -i), $self->{ref}], $env);
}
- eval { run_die([@cmd, 'update-server-info']) };
my $ibx = $self->{ibx};
+ eval { run_die([@cmd, 'update-server-info']) } if $ibx;
if ($ibx && $ibx->version == 1 && -d "$ibx->{inboxdir}/public-inbox" &&
eval { require PublicInbox::SearchIdx }) {
eval {
};
warn "$ibx->{inboxdir} index failed: $@\n" if $@;
}
- eval { run_die([@cmd, qw(gc --auto)]) } if $do_gc;
+ if ($do_gc) {
+ my @quiet = (-t STDERR ? () : '-q');
+ eval { run_die([@cmd, qw(gc --auto), @quiet]) }
+ }
}
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";
+ my $oid = <$r> // die "get-mark failed, need git 2.6.0+\n";
chomp($oid);
$oid;
}
# v2: we need this for Xapian
if ($smsg) {
$smsg->{blob} = $self->get_mark(":$blob");
- $smsg->{raw_bytes} = $n;
+ $smsg->set_bytes($raw_email, $n);
if (my $oidx = delete $smsg->{-oidx}) { # used by LeiStore
- return if $oidx->blob_exists($smsg->{blob});
+ 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}));
}
- # XXX do we need this? it's in git at this point
- $smsg->{-raw_email} = \$raw_email;
}
my $ref = $self->{ref};
my $commit = $self->{mark}++;
$self->{tip} = ":$commit";
}
-sub run_die ($;$$) {
- my ($cmd, $env, $rdr) = @_;
- my $pid = spawn($cmd, $env, $rdr);
- waitpid($pid, 0) == $pid or die join(' ', @$cmd) .' did not finish';
- $? == 0 or die join(' ', @$cmd) . " failed: $?\n";
-}
-
-my @INIT_FILES = ('HEAD' => "ref: refs/heads/master\n",
- 'description' => <<EOD,
-Unnamed repository; edit this file 'description' to name the repository.
-EOD
+my @INIT_FILES = ('HEAD' => undef, # filled in at runtime
'config' => <<EOC);
[core]
repositoryFormatVersion = 0
EOC
sub init_bare {
- my ($dir) = @_; # or self
+ my ($dir, $head) = @_; # or self
$dir = $dir->{git}->{git_dir} if ref($dir);
require File::Path;
File::Path::mkpath([ map { "$dir/$_" } qw(objects/info refs/heads) ]);
- for (my $i = 0; $i < @INIT_FILES; $i++) {
- my $f = $dir.'/'.$INIT_FILES[$i++];
+ $INIT_FILES[1] //= 'ref: '.default_branch."\n";
+ my @fn_contents = @INIT_FILES;
+ $fn_contents[1] = "ref: refs/heads/$head\n" if defined $head;
+ while (my ($fn, $contents) = splice(@fn_contents, 0, 2)) {
+ my $f = $dir.'/'.$fn;
next if -f $f;
open my $fh, '>', $f or die "open $f: $!";
- print $fh $INIT_FILES[$i] or die "print $f: $!";
+ print $fh $contents or die "print $f: $!";
close $fh or die "close $f: $!";
}
}
eval {
my $r = delete $self->{in} or die 'BUG: missing {in} when done';
print $w "done\n" or wfail;
- my $pid = delete $self->{pid} or
- die 'BUG: missing {pid} when done';
- waitpid($pid, 0) == $pid or die 'fast-import did not finish';
- $? == 0 or die "fast-import failed: $?";
+ close $r or die "fast-import failed: $?"; # ProcessPipe::CLOSE
};
my $wait_err = $@;
my $nchg = delete $self->{nchg};
}
}
-sub digest2mid ($$) {
- 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
# 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';
}