search: support "patchid:" prefix (git patch-id --stable)

[public-inbox.git] / lib / PublicInbox / SearchIdx.pm

diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm
index 95b14c3a55de4c7adfa6bf58d700a98534850324..53ec23a5c50c4facfcf1977ea04f2fa4498d8805 100644 (file)
--- a/lib/PublicInbox/SearchIdx.pm
+++ b/lib/PublicInbox/SearchIdx.pm
@@ -18,6 +18,7 @@ use PublicInbox::MsgIter;
 use PublicInbox::IdxStack;
 use Carp qw(croak carp);
 use POSIX qw(strftime);
+use Fcntl qw(SEEK_SET);
 use Time::Local qw(timegm);
 use PublicInbox::OverIdx;
 use PublicInbox::Spawn qw(spawn);
@@ -30,6 +31,7 @@ our @EXPORT_OK = qw(log2stack is_ancestor check_size prepare_stack
 my $X = \%PublicInbox::Search::X;
 our ($DB_CREATE_OR_OPEN, $DB_OPEN);
 our $DB_NO_SYNC = 0;
+our $DB_DANGEROUS = 0;
 our $BATCH_BYTES = $ENV{XAPIAN_FLUSH_THRESHOLD} ? 0x7fffffff :
        # assume a typical 64-bit system has 8x more RAM than a
        # typical 32-bit system:
@@ -115,7 +117,10 @@ sub load_xapian_writable () {
        my $ver = (eval($xap.'::major_version()') << 16) |
                (eval($xap.'::minor_version()') << 8) |
                eval($xap.'::revision()');
-       $DB_NO_SYNC = 0x4 if $ver >= 0x10400;
+       if ($ver >= 0x10400) {
+               $DB_NO_SYNC = 0x4;
+               $DB_DANGEROUS = 0x10;
+       }
        # Xapian v1.2.21..v1.2.24 were missing close-on-exec on OFD locks
        $X->{CLOEXEC_UNSET} = 1 if $ver >= 0x010215 && $ver <= 0x010218;
        1;
@@ -142,6 +147,9 @@ sub idx_acquire {
                        require PublicInbox::Syscall;
                        PublicInbox::Syscall::nodatacow_dir($dir);
                        $self->{-set_has_threadid_once} = 1;
+                       if (($self->{ibx} // $self->{eidx})->{-dangerous}) {
+                               $flag |= $DB_DANGEROUS;
+                       }
                }
        }
        return unless defined $flag;
@@ -229,8 +237,8 @@ sub index_old_diff_fn {
 
        # no renames or space support for traditional diffs,
        # find the number of leading common paths to strip:
-       my @fa = split('/', $fa);
-       my @fb = split('/', $fb);
+       my @fa = split(m'/', $fa);
+       my @fb = split(m'/', $fb);
        while (scalar(@fa) && scalar(@fb)) {
                $fa = join('/', @fa);
                $fb = join('/', @fb);
@@ -271,12 +279,12 @@ sub index_diff ($$$) {
                                                        $xnq);
                } elsif (m!^--- ("?[^/]+/.+)!) {
                        my $fn = $1;
-                       $fn = (split('/', git_unquote($fn), 2))[1];
+                       $fn = (split(m'/', git_unquote($fn), 2))[1];
                        $seen{$fn}++ or index_diff_inc($self, $fn, 'XDFN', $xnq);
                        $in_diff = 1;
                } elsif (m!^\+\+\+ ("?[^/]+/.+)!)  {
                        my $fn = $1;
-                       $fn = (split('/', git_unquote($fn), 2))[1];
+                       $fn = (split(m'/', git_unquote($fn), 2))[1];
                        $seen{$fn}++ or index_diff_inc($self, $fn, 'XDFN', $xnq);
                        $in_diff = 1;
                } elsif (/^--- (\S+)/) {
@@ -342,6 +350,20 @@ sub index_xapian { # msg_iter callback
        defined $s or return;
        $_[0]->[0] = $part = undef; # free memory
 
+       if ($s =~ /^(?:diff|---|\+\+\+) /ms) {
+               open(my $fh, '+>:utf8', undef) or die "open: $!";
+               open(my $eh, '+>', undef) or die "open: $!";
+               $fh->autoflush(1);
+               print $fh $s or die "print: $!";
+               sysseek($fh, 0, SEEK_SET) or die "sysseek: $!";
+               my $id = ($self->{ibx} // $self->{eidx})->git->qx(
+                                               [qw(patch-id --stable)],
+                                               {}, { 0 => $fh, 2 => $eh });
+               $id =~ /\A([a-f0-9]{40,})/ and $doc->add_term('XDFID'.$1);
+               seek($eh, 0, SEEK_SET) or die "seek: $!";
+               while (<$eh>) { warn $_ }
+       }
+
        # split off quoted and unquoted blocks:
        my @sections = PublicInbox::MsgIter::split_quotes($s);
        undef $s; # free memory